input
null | chosen
stringlengths 2.16k
251k
| rejected
stringlengths 2.13k
251k
| question
stringclasses 5
values | margin
float64 0.25
0.25
| pi_key
stringlengths 64
64
| pi_score_chosen
float64 0
0.99
| pi_score_rejected
float64 0
1
| pi_score_chosen_sfc_q_30750
float64 0
0.99
| pi_score_rejected_sfc_q_30750
float64 0
1
| pi_score_chosen_sfc_o_33000
float64 0
1
| pi_score_rejected_sfc_o_33000
float64 0
1
|
|---|---|---|---|---|---|---|---|---|---|---|---|
null |
{"patent": "while the support device of the present invention may be used in a range of different vessels , including blood vessels , it has particular application in procedures where an organ or anatomical region is undergoing localized perfusion with a therapeutic , diagnostic or other agent . for simplicity , these agents will be hereinafter referred to as therapeutic agents . however , it is to be understood that the term \u201c therapeutic \u201d is not to be construed as limiting , and that it includes , without limitation , therapeutic , diagnostic , prophylactic and other agents not specifically identified herein , but which would be considered by the relevant skilled addressee to be suitable for perfusion to an organ or anatomical region . perfusion may be total perfusion , where the entire organ is totally or substantially isolated from the systemic flow , or partial perfusion where only a portion of the organ is substantially isolated . localized perfusion of this kind presents advantages by improving efficacy and the time exposure of the therapeutic agent to the relevant cells . it also limits exposure and hence toxicity to non - target cells as described in brief above . however , it is to be understood that the present invention may also be used simply to collect or drain fluid from an organ or region . collected fluid may be removed from the subject and re - circulated into the organ , filtered and / or treated , or discarded . in some organs , it may be difficult to achieve total isolation , so partial isolation and perfusion may be performed , for example to the right or left lobe of the liver . despite partial perfusion being capable of delivering therapeutic agent to merely a part of the organ , significant therapeutic benefit may still be achieved . particular benefit may be achieved where perfusate is collected after perfusing the target organ , so as to prevent subsequent circulation of the therapeutic agent to other regions of the body where toxic effects may be observed , or the therapeutic agent wasted . the benefit may be improved further where collected perfusate is re - circulated into the target organ utilizing any therapeutic agent which remains after a first pass through the target organ . this may be achieved using the approach described in published patent application wo2005 / 082440 , the entire contents of which are herein incorporated by reference . as discussed infra , when fluid is collected from vessels draining from a target organ or region , one or more of these vessels may require cannulation with a collection catheter . when fluid is drained through these collection catheters , the vessels in which they are positioned become susceptible to collapse as the pressure inside decreases . while some vessels may be more susceptible to collapse than others , the support device of the present invention can provide advantages by supporting and stabilizing the vessel and even anchoring the collection catheter in position . the support device of the present invention may facilitate or at least improve the performance of perfusion . in some instances , the advantages of the present invention have been found to be essential to maintaining adequate positioning of collection catheters and flow rates within the vessel during perfusion . the right and left lobes of the liver have been identified as possible target regions and in this context , the support device may be deployed in one of the hepatic veins to support and maintain patency of the vein as fluid ( e . g . perfusate ) is collected from the liver . however , it is to be understood that fluid from many other organs or regions may be accessed in this way . deploying the support device may also protect the vessel wall by maintaining the tip of the catheter substantially centrally of the vessel or at least at a distance from the vessel walls to prevent aspiration or cavitation . deployment of the device may refer to partial or complete deployment . in complete deployment , the entire expandable member is released from the catheter and expanded to its full extent . in partial deployment , part of the expandable member is retained within the catheter and the amount of expansion is limited by the diameter of the catheter opening . partial deployment may be useful where , for example , during deployment it is found that the diameter of the expandable member may exceed the vessel diameter by an unsafe amount and complete deployment is likely to damage the vessel wall . limiting expansion of the device by partial deployment may avoid vessel damage . partial deployment may also stabilize the expandable member by limiting its movement relative to the catheter tip . thus by retaining part of the expandable member within the catheter , torsional , axial and lateral movement of the member , relative to the catheter is prevented or at least minimized by the struts of the expandable member being in abutment with the internal surface of the catheter . alternatively , the expandable member may be modified at the proximal end , for example by incorporating a lead , a link or other means to limit the extent of movement possible between the catheter tip and the expandable member once deployed . as a further positioning aid , markings may be provided at the proximal end of the control stem / shaft , outside the patient &# 39 ; s body . as the device is released into the vessel , the markings may be utilized to indicate the distance of device deployment , past the catheter tip . during collection of fluid from the vessel , low pressures may develop at the collection device tip , particularly where a roller / peristaltic pump or the like is used to draw fluid from the target organ out of the vessel . this may be indicated by pressures in a lumen feeding into the pump as low as , for example , \u2212 190 mmhg , although clearly these pressures are variable depending on the vessel type , health and age of the subject , characteristics of the perfusion circuit and the like . in the absence of the inventive support device , these pressures can cause the vessel to collapse . not only would vessel collapse affect the perfusion procedure , vessel collapse can also cause venous pooling in the organ and irreversible tissue damage . the advantages and benefits of the present invention will be expanded upon in the following detailed description presenting some of the preferred embodiments of the invention , and the specific examples which follow . it is to be understood that the embodiments and examples provided herein are intended to indicate how the present invention may be performed and are not intended to be limiting on the scope of protection sought as is defined in the claims appended hereto . fig1 a shows an example of an expandable member , in its expanded condition , suitable for supporting a vessel . expandable member 104 is provided in the form of an expandable framework and is adapted to be percutaneously deliverable to the blood vessel in a collapsed condition . fig1 b shows the expandable member in a collapsed condition within a catheter 110 , in which ends 105 , 107 have been drawn apart to radially reduce the member . when collapsed within catheter 110 , atraumatic tip 101 may protrude from the catheter to assist in guiding the support device into the vessel prior to deployment . when the expandable member has been guided into the target blood vessel , the catheter 110 is retracted ( or the expandable member is pushed out of the catheter ), deploying the device into the vessel where it expands . fig1 c shows the support device fully deployed from the catheter , with the expandable member in its fully expanded condition . a guidewire or stem 106 extends within the catheter 110 and is used to deliver the device from a point of entry through the peripheral vasculature to the target vessel . atraumatic tip 101 coupled to the expandable member 104 , is adapted to make atraumatic contact with vessel walls during placement of the device by deforming or deflecting off the vessel wall on contact . this can be achieved by incorporating flexibility into the tip so that it deforms upon contact with the vessel wall . alternatively or additionally , the tip may be shaped or curved to avoid trauma . the atraumatic tip may take any one of a number of forms . in the examples illustrated in fig1 to 3 , the atraumatic tip 101 , 201 is j - shaped . however , other shapes are considered to be suitable , including but not limited to those illustrated in fig4 . for example , the atraumatic tip may have a cross section which is enlarged relative to the guidewire radius , and have a smooth surface so as to avoid causing perforation when the tip comes into contact with the vessel wall . one such example is shown in fig4 a where the atraumatic tip 401 is tear - shaped . alternatively , the atraumatic tip may include a portion having a pigtail shaped curve 402 ( fig4 b ), or an angled tip ( not shown ). preferably , the expandable member is formed from a biocompatible superelastic material , or alternatively from a shape memory material or a material which exhibits both of these properties , being capable of recovery after deformation for delivery in a collapsed or compressed state within a catheter . devices manufactured using these materials can be collapsed for percutaneous delivery to a deployment site and then resume a known shape on deployment . a range of biocompatible materials may be suitable such as alloys of nickel and titanium ( e . g . nitinol ). other suitable biocompatible materials include but are not limited to polymers and plastics such as hydrophilic plastics , ceramics and the like . fig3 illustrates the support device of fig1 a to 1 c , with an occluding balloon inflated around catheter 110 . the occluding balloon 114 may be utilized during collection of fluid from an organ or region of the body in isolation , where substantially all of the fluid flowing out of the organ or region is collected by the catheter 110 . the occluding means substantially prevents blood , therapeutic agent and / or other fluids entering the vessel from flowing on to other organs or regions , and permits collection of substantially all of the fluid entering the vessel . collected fluid may then be analyzed and / or re - oxygenated and / or perfused through the organ , discarded or handled otherwise . the occlusion means may include an occluding balloon , flange , disc or other means . catheter 110 is delivered to the vessel with the balloon 114 in a deflated condition . the expandable member is delivered , through the catheter , and deployed inside the vessel . the balloon is then inflated around the catheter and substantially all the fluid in the vessel flows through the catheter and into a perfusion set or reservoir to which it is connected . a pump , syringe or other means may be incorporated into the perfusion set to draw fluid out of the vessel , through the catheter , at a rate which substantially maintains the required flow through the organ or region , or through a re - perfusion circuit . as fluid is drawn out of the vessel through the catheter , the expanded support structure supports the vessel walls , preventing collapse or cavitation which might otherwise result from the low pressures or high flow rates generated at the catheter tip , maintaining patency and ensuring flow in the circuit . the expandable member may also anchor the device in position within the vessel , substantially precluding movement of the device and ensuring that the catheter is retained in an optimal location for collection of fluid . the expandable member may take a range of different shapes when in an expanded ( or collapsed ) configuration , and may provide any number of supporting filaments or struts . the design of the expandable member may be based on a range of criteria including but not limited to the size and strength of the vessel wall and the flow rates and pressures likely to be generated near the device . some of these embodiments are illustrated in fig8 a to 8 c although these are examples only and are not intended to limit the scope of the invention as broadly described herein . fig8 a to 8 c illustrate expandable members having elongate portions in the supporting struts adapted for contact with the vessel wall . in the example in fig8 b , the supporting struts are slightly rounded to reduce trauma to the vessel walls . fig8 c provides additional struts when compared with fig8 a , as may be necessitated in particularly flaccid vessels requiring more substantial support . embodiments illustrated herein provide expandable members with a substantially elongate structure adapted for coaxial insertion into and placement within the vessel . the elongate structure supports the vessel over a length on the elongate portions of the struts substantially parallel to and in contact with the vessel wall . these elongate portions may be substantially straight , or may be curved ( e . g . fig8 b ). supporting the vessel wall over a length of the support device , compared with the point of supports of the prior art , improves the capacity of the device to maintain patency , even when very low pressures and high flow rates are generated at the catheter tip , and also reduces the likelihood of the device causing damage to the vessel wall . the elongate portions may have a length which is about the same as or greater than the diameter of the vessel being supported , or some multiple of the vessel diameter , or for example from 1 mm up to 30 mm depending on the vessel size and structure . the length of the elongate portion may be selected according to the vessel being supported , the size of the catheter being used and the flow rates and pressures likely to be generated at the catheter tip . preferably , the elongate portions of the expandable member which contact the vessel wall , are just adjacent the distal tip of the catheter when the device is fully deployed . thus , a proximal end of one or more of the elongate portions may commence , for example , within 0 . 1 to 25 mm of the catheter tip , or at least at a distance which is less than the diameter of the catheter opening . this prevents the vessel wall from being drawn into the space between the catheter tip and the start of the elongate portion of the expandable member which contacts the vessel wall . further , the device may be configured so that when it is in an expanded condition , the distance between adjacent elongate portions is sufficiently small to prevent the vessel wall from being drawn into gaps between them . for example , the distance between adjacent elongate portions may be less than the diameter of the catheter . alternatively , the distance between the adjacent elongate portions may be less than , for example , 3 , 2 . 5 , 2 , 1 . 5 , 1 or 0 . 5 mm , depending on the size and type of the target vessel , and the diameter of the collection catheter being used . preferably , the support device possesses sufficient mechanical strength to maintain patency during collection of fluid , withstanding the deformation forces which may occur in response to suction or low pressures produced at the collection catheter tip . in some embodiments however , it may also be desirable for the device to exhibit some flexibility , and conform to the shape of the vessel when deployed . thus , the support device is capable of providing support and maintaining patency along a length of the vessel , even where there is a curve in the vessel wall . an alternative embodiment of a support device 200 is illustrated in fig2 . proximal end 205 of the expandable member 204 is fixedly attached to a stem or shaft 206 , whereas distal end 203 of the expandable member is movable and able to slide over part of the shaft . this enables the member to collapse radially for delivery inside a delivery catheter , and also facilitates recapture of the device . fig5 illustrates another alternative embodiment of a support device shown at 500 in an expanded condition . in this embodiment , both the proximal end 505 and the distal end 503 of the expandable member are movable along a stem or shaft 506 used to deliver the device to the vessel . stops 508 a , 508 b are provided at fixed locations on a distal portion of the shaft , arranged between ends 503 , 505 of the expandable member . these stops may consist of a small ring , crimp or node of increased diameter , relative to the shaft diameter , and prevent the ends of the expandable member from moving across the stop . this facilitates deployment and retrieval of the expandable member from a catheter . fig1 illustrates a support device 151 consisting of an expandable framework 155 having a woven or braided , basket - like configuration when in the expanded condition . in this arrangement , the support device may also include occluding means in the form of a thin flow - proof coating 156 on the inner and / or outer surface of framework 155 to prevent flow of liquid from the vessel . thus , substantially all fluid in the vessel may be collected by catheter 160 . the flow - proof coating may be made from biocompatible silicon , elastomer or flow - proof polymer . preferably , the support device includes a radiopaque or other marker so that it can be positioned within the target vessel using an imaging system such as those generally known in the art . this enables the physician to position and deploy the expandable member into the blood vessel accurately . the marker may be incorporated into the expandable member and / or into an atraumatic guiding tip which may be incorporated into the support device . preferably , the atraumatic tip is manufactured from , includes or is coated with a lubricant and / or a material having a low coefficient of friction . many materials having low coefficient of friction properties may be used including but not limited to biocompatible high density polyethylene ( hdpe ), teflon \u00ae, polypropylene , polyethylene , microglide \u2122, low friction chromium and silicon to name a few . this improves the performance of the atraumatic tip , so that it \u201c slides \u201d along the vessel wall upon making contact , thereby substantially avoiding trauma . use of an atraumatic guiding tip improves the safety and ability to position the expandable member in the target vessel . moreover , since the atraumatic tip may exhibit greater flexibility than the rest of the device , the device is easier to manipulate into position . the atraumatic tip may be provided at a distance from the distal end of the expandable member which enables a physician to guide the expandable member into position within the target vessel . this distance may be anywhere from , for example , 0 . 25 to 5 centimeters from the distal end of the expandable member when in an expanded condition , although it is to be understood that larger or smaller distances may be utilized , depending on the location of the target vessel and the anatomy surrounding it . referring now to fig6 a and 6 b , another example of a support device 600 is shown . a lumen 602 has a control stem 601 extending therein . four loop portions 603 are provided . each loop portion is attached at a first loop end to a distal end 604 of the lumen , and at a second loop end to the control stem at 605 . the loop portions are controllably expandable by advancing the control stem within the lumen in the direction shown by arrow 606 ( fig6 b ). the support device is percutaneously deliverable with the plurality of loop portions housed substantially within the lumen 602 as illustrated in fig6 a and expandable as illustrated in fig6 b . whilst the embodiment illustrated in fig6 a and 6 b provides 4 loop portions , it is to be understood that any number of loop portions may be used . the number of loop portions incorporated into the device may depend on , for example , the anatomy of the vessel being supported , and / or the size of the catheter used to deliver the device . fig7 a and 7 b illustrate another example of a support device 700 which provides 3 loop portions 703 attached to control stem 701 at juncture 705 . the 3 loop portions are contained during delivery substantially within lumen 702 ( fig7 a ), and are controllably expandable to maintain patency within the blood vessel by advancing control stem 701 in the direction of arrow 706 ( fig7 b ). the rounded edges of the loop portions present a reduced risk of damaging the vessel walls , e . g . by perforation or bruising during delivery . the one or more loop portions may be attached to or near the distal end of the delivery lumen in any suitable manner . the point of attachment may be inside or outside the lumen . the loops may be manufactured from any suitable material such as a metal , metal alloy , plastic , polymer , or other filamentous material or composite . the one or more loop portions may be attached at a second loop end to the control stem by soldering , fusing , an adhesive , or any other suitable means . in another embodiment , the loop portions may be attached to a first and a second loop end to the control stem . the support structure of fig6 a , 6 b , 7 a and 7 b may further include an atraumatic guiding tip of the kind described above to aid in positioning the support structure within the blood vessel . alternatively , parts of the loop portions which may protrude from the lumen when the loop portions are in their collapsed state may be used to guide the support structure into the blood vessel . one or more of the loop portions may be provided with a radiopaque or other marker to assist in this regard . retention means may also be provided with the support structure to retain the expandable member in an expanded condition within the vessel . the retention means may be in the form of a clamp , clip , thumb - slide or the like accessible from outside the patient &# 39 ; s body , and may facilitate adjustment of a deployed expandable member during a procedure . retention means may also impart additional rigidity and strength to the expandable member . thus , the retention member may be used to counteract excessively low pressures which may otherwise cause the expanded member to fail . a support structure of the kind illustrated in the figures may be delivered within a multilumen catheter 900 of the kind illustrated in cross section in fig9 . using this catheter , the support device 910 can be delivered through a first internal lumen 901 without interfering with flow in a second lumen 902 . a third lumen 903 may be provided for monitoring flow rates and pressures , for blood analysis or for delivering other percutaneous tools or devices to the vessel or as an inflation lumen for an occlusion balloon . it is to be understood that in the various embodiments of the present invention , the expanded member does not require constant contact with the vessel walls to provide the required support . for example , the diameter of the expanded member may be less than the diameter of the vessel so that the expanded member only contacts the vessel wall when the vessel begins to collapse . patency is considered to be maintained as long as the support device keeps the vessel open to a degree which is sufficient to maintain continuous flow . to avoid causing turbulence or other undesirable blood flow effects within the vessel , and to optimize flow in the vessel it may be desirable to substantially match the diameter of the expanded member to the diameter of the vessel . alternatively the expandable member may be shaped , e . g . as a coil or helix , to have minimal effect on the flow in the vessel . in one embodiment , the expandable member may have a slightly larger expanded diameter than the relaxed vessel to create an anchoring effect . depending on the size of the outflow vessel from which blood and perfusate is collected from the target region , there may be a natural tendency for the collection catheter tip to move about and contact the vessel , thus increasing the risk of vessel collapse or invagination of the catheter tip into the vessel wall . this can cause pooling of fluid in the isolated target region and may cause serious and permanent damage to the organ or region of the patient being treated . use of a support structure in conjunction with the collection catheter to maintain patency of the outflow vessel , in accordance with embodiments of the present invention can minimize the risk of these complications eventuating . thus , a collection catheter associated with the expanding member can be retained in position during fluid collection . this minimizes movement of the catheter tip , ensures that it is substantially centered relative to the vessel walls and improves withdrawal of fluid out of the vessel . at completion of the procedure , it is desirable that the expanded member is collapsed or compressed and recaptured , preferably in the catheter from which it was deployed . this facilitates removal of the support device from the patient . a reinforcing tip may be provided on the catheter end to strengthen it for recapture . alternatively or additionally , the tip may be coated with a lubricant and / or material having a low coefficient of friction to facilitate smooth recapture of the expandable member . the catheter may also have an internal coating of lubricant and / or a material having a low coefficient of friction to assist translation of support device along its interior during delivery and removal of the device from the patient . although the present invention has been described in terms of the presently preferred embodiments , it is to be understood that the disclosure is not to be interpreted as limiting . various alterations and modifications will no doubt become apparent to those skilled in the art after having read the above disclosure and it is intended that the present disclosure be interpreted as covering all alterations and modifications as fall within the true spirit and scope of the invention . effect of support device on flow rates and pressures achievable during recirculation in sheep right hepatic vein , cephalic vein , coronary sinus and renal vein during recirculation a 0 . 014 \u2033 diameter superelastic nitinol wire stem of 1 . 35 m length was used , coupled to an expandable member having 6 pre - shaped elliptical loop portions welded to the stem . a 0 . 024 \u2033 od atraumatic tip of 2 cm length attached to the distal end of the expandable member was used to position the device in the blood vessel . a balloon occlusion catheter was positioned in the vessel and the expandable member deployed at the tip of the catheter . the balloon was inflated to isolate and capture flows in the vessel and the catheter was connected to a standard extracorporeal circuit for blood circulation . negative pressures were observed in perfusion lines draining the coronary sinus , renal vein , right hepatic vein and cephalic vein during recirculation both with and without a support device . these data show that cavitation is prevented at certain pressures in the vessels tested where a support device is used , but is not prevented where the support device is absent in the vessel at those pressures . although cavitation may occur even with the support device , it occurs at higher flows . also , cavitation ceases sooner where the support device was employed allowing flow to return to normal . in the coronary sinus , recovery from cavitation was not possible without the support device , emphasizing the importance of the device in the procedure . the data further demonstrates that vessel collapse can be irreversible in the absence of a support structure . however , where a support structure is present , the vessel collapse may be reversed by increasing pressure in the vessel or by slowing or reversing the flow rate of fluid through the vessel . more specifically , considering the data for the right hepatic vein , flow rates of up to 250 ml per minute may be achieved before cavitation occurs where a support device is present in the vessel . under the same conditions but where there is no support device , flow rates of only up to 180 ml per minute are possible . a more striking example of the advantages of the support device is seen for the cephalic vein where no flow is achievable without the device . when the vessel wall is supported by the device flow rates of up to 200 ml per minute are noted before cavitation occurs . when the vessel wall is supported flow rates of up to 200 ml per minute are noted before cavitation occurs .", "category": "Human Necessities"}
|
{"category": "Mechanical Engineering; Lightning; Heating; Weapons; Blasting", "patent": "while the support device of the present invention may be used in a range of different vessels , including blood vessels , it has particular application in procedures where an organ or anatomical region is undergoing localized perfusion with a therapeutic , diagnostic or other agent . for simplicity , these agents will be hereinafter referred to as therapeutic agents . however , it is to be understood that the term \u201c therapeutic \u201d is not to be construed as limiting , and that it includes , without limitation , therapeutic , diagnostic , prophylactic and other agents not specifically identified herein , but which would be considered by the relevant skilled addressee to be suitable for perfusion to an organ or anatomical region . perfusion may be total perfusion , where the entire organ is totally or substantially isolated from the systemic flow , or partial perfusion where only a portion of the organ is substantially isolated . localized perfusion of this kind presents advantages by improving efficacy and the time exposure of the therapeutic agent to the relevant cells . it also limits exposure and hence toxicity to non - target cells as described in brief above . however , it is to be understood that the present invention may also be used simply to collect or drain fluid from an organ or region . collected fluid may be removed from the subject and re - circulated into the organ , filtered and / or treated , or discarded . in some organs , it may be difficult to achieve total isolation , so partial isolation and perfusion may be performed , for example to the right or left lobe of the liver . despite partial perfusion being capable of delivering therapeutic agent to merely a part of the organ , significant therapeutic benefit may still be achieved . particular benefit may be achieved where perfusate is collected after perfusing the target organ , so as to prevent subsequent circulation of the therapeutic agent to other regions of the body where toxic effects may be observed , or the therapeutic agent wasted . the benefit may be improved further where collected perfusate is re - circulated into the target organ utilizing any therapeutic agent which remains after a first pass through the target organ . this may be achieved using the approach described in published patent application wo2005 / 082440 , the entire contents of which are herein incorporated by reference . as discussed infra , when fluid is collected from vessels draining from a target organ or region , one or more of these vessels may require cannulation with a collection catheter . when fluid is drained through these collection catheters , the vessels in which they are positioned become susceptible to collapse as the pressure inside decreases . while some vessels may be more susceptible to collapse than others , the support device of the present invention can provide advantages by supporting and stabilizing the vessel and even anchoring the collection catheter in position . the support device of the present invention may facilitate or at least improve the performance of perfusion . in some instances , the advantages of the present invention have been found to be essential to maintaining adequate positioning of collection catheters and flow rates within the vessel during perfusion . the right and left lobes of the liver have been identified as possible target regions and in this context , the support device may be deployed in one of the hepatic veins to support and maintain patency of the vein as fluid ( e . g . perfusate ) is collected from the liver . however , it is to be understood that fluid from many other organs or regions may be accessed in this way . deploying the support device may also protect the vessel wall by maintaining the tip of the catheter substantially centrally of the vessel or at least at a distance from the vessel walls to prevent aspiration or cavitation . deployment of the device may refer to partial or complete deployment . in complete deployment , the entire expandable member is released from the catheter and expanded to its full extent . in partial deployment , part of the expandable member is retained within the catheter and the amount of expansion is limited by the diameter of the catheter opening . partial deployment may be useful where , for example , during deployment it is found that the diameter of the expandable member may exceed the vessel diameter by an unsafe amount and complete deployment is likely to damage the vessel wall . limiting expansion of the device by partial deployment may avoid vessel damage . partial deployment may also stabilize the expandable member by limiting its movement relative to the catheter tip . thus by retaining part of the expandable member within the catheter , torsional , axial and lateral movement of the member , relative to the catheter is prevented or at least minimized by the struts of the expandable member being in abutment with the internal surface of the catheter . alternatively , the expandable member may be modified at the proximal end , for example by incorporating a lead , a link or other means to limit the extent of movement possible between the catheter tip and the expandable member once deployed . as a further positioning aid , markings may be provided at the proximal end of the control stem / shaft , outside the patient &# 39 ; s body . as the device is released into the vessel , the markings may be utilized to indicate the distance of device deployment , past the catheter tip . during collection of fluid from the vessel , low pressures may develop at the collection device tip , particularly where a roller / peristaltic pump or the like is used to draw fluid from the target organ out of the vessel . this may be indicated by pressures in a lumen feeding into the pump as low as , for example , \u2212 190 mmhg , although clearly these pressures are variable depending on the vessel type , health and age of the subject , characteristics of the perfusion circuit and the like . in the absence of the inventive support device , these pressures can cause the vessel to collapse . not only would vessel collapse affect the perfusion procedure , vessel collapse can also cause venous pooling in the organ and irreversible tissue damage . the advantages and benefits of the present invention will be expanded upon in the following detailed description presenting some of the preferred embodiments of the invention , and the specific examples which follow . it is to be understood that the embodiments and examples provided herein are intended to indicate how the present invention may be performed and are not intended to be limiting on the scope of protection sought as is defined in the claims appended hereto . fig1 a shows an example of an expandable member , in its expanded condition , suitable for supporting a vessel . expandable member 104 is provided in the form of an expandable framework and is adapted to be percutaneously deliverable to the blood vessel in a collapsed condition . fig1 b shows the expandable member in a collapsed condition within a catheter 110 , in which ends 105 , 107 have been drawn apart to radially reduce the member . when collapsed within catheter 110 , atraumatic tip 101 may protrude from the catheter to assist in guiding the support device into the vessel prior to deployment . when the expandable member has been guided into the target blood vessel , the catheter 110 is retracted ( or the expandable member is pushed out of the catheter ), deploying the device into the vessel where it expands . fig1 c shows the support device fully deployed from the catheter , with the expandable member in its fully expanded condition . a guidewire or stem 106 extends within the catheter 110 and is used to deliver the device from a point of entry through the peripheral vasculature to the target vessel . atraumatic tip 101 coupled to the expandable member 104 , is adapted to make atraumatic contact with vessel walls during placement of the device by deforming or deflecting off the vessel wall on contact . this can be achieved by incorporating flexibility into the tip so that it deforms upon contact with the vessel wall . alternatively or additionally , the tip may be shaped or curved to avoid trauma . the atraumatic tip may take any one of a number of forms . in the examples illustrated in fig1 to 3 , the atraumatic tip 101 , 201 is j - shaped . however , other shapes are considered to be suitable , including but not limited to those illustrated in fig4 . for example , the atraumatic tip may have a cross section which is enlarged relative to the guidewire radius , and have a smooth surface so as to avoid causing perforation when the tip comes into contact with the vessel wall . one such example is shown in fig4 a where the atraumatic tip 401 is tear - shaped . alternatively , the atraumatic tip may include a portion having a pigtail shaped curve 402 ( fig4 b ), or an angled tip ( not shown ). preferably , the expandable member is formed from a biocompatible superelastic material , or alternatively from a shape memory material or a material which exhibits both of these properties , being capable of recovery after deformation for delivery in a collapsed or compressed state within a catheter . devices manufactured using these materials can be collapsed for percutaneous delivery to a deployment site and then resume a known shape on deployment . a range of biocompatible materials may be suitable such as alloys of nickel and titanium ( e . g . nitinol ). other suitable biocompatible materials include but are not limited to polymers and plastics such as hydrophilic plastics , ceramics and the like . fig3 illustrates the support device of fig1 a to 1 c , with an occluding balloon inflated around catheter 110 . the occluding balloon 114 may be utilized during collection of fluid from an organ or region of the body in isolation , where substantially all of the fluid flowing out of the organ or region is collected by the catheter 110 . the occluding means substantially prevents blood , therapeutic agent and / or other fluids entering the vessel from flowing on to other organs or regions , and permits collection of substantially all of the fluid entering the vessel . collected fluid may then be analyzed and / or re - oxygenated and / or perfused through the organ , discarded or handled otherwise . the occlusion means may include an occluding balloon , flange , disc or other means . catheter 110 is delivered to the vessel with the balloon 114 in a deflated condition . the expandable member is delivered , through the catheter , and deployed inside the vessel . the balloon is then inflated around the catheter and substantially all the fluid in the vessel flows through the catheter and into a perfusion set or reservoir to which it is connected . a pump , syringe or other means may be incorporated into the perfusion set to draw fluid out of the vessel , through the catheter , at a rate which substantially maintains the required flow through the organ or region , or through a re - perfusion circuit . as fluid is drawn out of the vessel through the catheter , the expanded support structure supports the vessel walls , preventing collapse or cavitation which might otherwise result from the low pressures or high flow rates generated at the catheter tip , maintaining patency and ensuring flow in the circuit . the expandable member may also anchor the device in position within the vessel , substantially precluding movement of the device and ensuring that the catheter is retained in an optimal location for collection of fluid . the expandable member may take a range of different shapes when in an expanded ( or collapsed ) configuration , and may provide any number of supporting filaments or struts . the design of the expandable member may be based on a range of criteria including but not limited to the size and strength of the vessel wall and the flow rates and pressures likely to be generated near the device . some of these embodiments are illustrated in fig8 a to 8 c although these are examples only and are not intended to limit the scope of the invention as broadly described herein . fig8 a to 8 c illustrate expandable members having elongate portions in the supporting struts adapted for contact with the vessel wall . in the example in fig8 b , the supporting struts are slightly rounded to reduce trauma to the vessel walls . fig8 c provides additional struts when compared with fig8 a , as may be necessitated in particularly flaccid vessels requiring more substantial support . embodiments illustrated herein provide expandable members with a substantially elongate structure adapted for coaxial insertion into and placement within the vessel . the elongate structure supports the vessel over a length on the elongate portions of the struts substantially parallel to and in contact with the vessel wall . these elongate portions may be substantially straight , or may be curved ( e . g . fig8 b ). supporting the vessel wall over a length of the support device , compared with the point of supports of the prior art , improves the capacity of the device to maintain patency , even when very low pressures and high flow rates are generated at the catheter tip , and also reduces the likelihood of the device causing damage to the vessel wall . the elongate portions may have a length which is about the same as or greater than the diameter of the vessel being supported , or some multiple of the vessel diameter , or for example from 1 mm up to 30 mm depending on the vessel size and structure . the length of the elongate portion may be selected according to the vessel being supported , the size of the catheter being used and the flow rates and pressures likely to be generated at the catheter tip . preferably , the elongate portions of the expandable member which contact the vessel wall , are just adjacent the distal tip of the catheter when the device is fully deployed . thus , a proximal end of one or more of the elongate portions may commence , for example , within 0 . 1 to 25 mm of the catheter tip , or at least at a distance which is less than the diameter of the catheter opening . this prevents the vessel wall from being drawn into the space between the catheter tip and the start of the elongate portion of the expandable member which contacts the vessel wall . further , the device may be configured so that when it is in an expanded condition , the distance between adjacent elongate portions is sufficiently small to prevent the vessel wall from being drawn into gaps between them . for example , the distance between adjacent elongate portions may be less than the diameter of the catheter . alternatively , the distance between the adjacent elongate portions may be less than , for example , 3 , 2 . 5 , 2 , 1 . 5 , 1 or 0 . 5 mm , depending on the size and type of the target vessel , and the diameter of the collection catheter being used . preferably , the support device possesses sufficient mechanical strength to maintain patency during collection of fluid , withstanding the deformation forces which may occur in response to suction or low pressures produced at the collection catheter tip . in some embodiments however , it may also be desirable for the device to exhibit some flexibility , and conform to the shape of the vessel when deployed . thus , the support device is capable of providing support and maintaining patency along a length of the vessel , even where there is a curve in the vessel wall . an alternative embodiment of a support device 200 is illustrated in fig2 . proximal end 205 of the expandable member 204 is fixedly attached to a stem or shaft 206 , whereas distal end 203 of the expandable member is movable and able to slide over part of the shaft . this enables the member to collapse radially for delivery inside a delivery catheter , and also facilitates recapture of the device . fig5 illustrates another alternative embodiment of a support device shown at 500 in an expanded condition . in this embodiment , both the proximal end 505 and the distal end 503 of the expandable member are movable along a stem or shaft 506 used to deliver the device to the vessel . stops 508 a , 508 b are provided at fixed locations on a distal portion of the shaft , arranged between ends 503 , 505 of the expandable member . these stops may consist of a small ring , crimp or node of increased diameter , relative to the shaft diameter , and prevent the ends of the expandable member from moving across the stop . this facilitates deployment and retrieval of the expandable member from a catheter . fig1 illustrates a support device 151 consisting of an expandable framework 155 having a woven or braided , basket - like configuration when in the expanded condition . in this arrangement , the support device may also include occluding means in the form of a thin flow - proof coating 156 on the inner and / or outer surface of framework 155 to prevent flow of liquid from the vessel . thus , substantially all fluid in the vessel may be collected by catheter 160 . the flow - proof coating may be made from biocompatible silicon , elastomer or flow - proof polymer . preferably , the support device includes a radiopaque or other marker so that it can be positioned within the target vessel using an imaging system such as those generally known in the art . this enables the physician to position and deploy the expandable member into the blood vessel accurately . the marker may be incorporated into the expandable member and / or into an atraumatic guiding tip which may be incorporated into the support device . preferably , the atraumatic tip is manufactured from , includes or is coated with a lubricant and / or a material having a low coefficient of friction . many materials having low coefficient of friction properties may be used including but not limited to biocompatible high density polyethylene ( hdpe ), teflon \u00ae, polypropylene , polyethylene , microglide \u2122, low friction chromium and silicon to name a few . this improves the performance of the atraumatic tip , so that it \u201c slides \u201d along the vessel wall upon making contact , thereby substantially avoiding trauma . use of an atraumatic guiding tip improves the safety and ability to position the expandable member in the target vessel . moreover , since the atraumatic tip may exhibit greater flexibility than the rest of the device , the device is easier to manipulate into position . the atraumatic tip may be provided at a distance from the distal end of the expandable member which enables a physician to guide the expandable member into position within the target vessel . this distance may be anywhere from , for example , 0 . 25 to 5 centimeters from the distal end of the expandable member when in an expanded condition , although it is to be understood that larger or smaller distances may be utilized , depending on the location of the target vessel and the anatomy surrounding it . referring now to fig6 a and 6 b , another example of a support device 600 is shown . a lumen 602 has a control stem 601 extending therein . four loop portions 603 are provided . each loop portion is attached at a first loop end to a distal end 604 of the lumen , and at a second loop end to the control stem at 605 . the loop portions are controllably expandable by advancing the control stem within the lumen in the direction shown by arrow 606 ( fig6 b ). the support device is percutaneously deliverable with the plurality of loop portions housed substantially within the lumen 602 as illustrated in fig6 a and expandable as illustrated in fig6 b . whilst the embodiment illustrated in fig6 a and 6 b provides 4 loop portions , it is to be understood that any number of loop portions may be used . the number of loop portions incorporated into the device may depend on , for example , the anatomy of the vessel being supported , and / or the size of the catheter used to deliver the device . fig7 a and 7 b illustrate another example of a support device 700 which provides 3 loop portions 703 attached to control stem 701 at juncture 705 . the 3 loop portions are contained during delivery substantially within lumen 702 ( fig7 a ), and are controllably expandable to maintain patency within the blood vessel by advancing control stem 701 in the direction of arrow 706 ( fig7 b ). the rounded edges of the loop portions present a reduced risk of damaging the vessel walls , e . g . by perforation or bruising during delivery . the one or more loop portions may be attached to or near the distal end of the delivery lumen in any suitable manner . the point of attachment may be inside or outside the lumen . the loops may be manufactured from any suitable material such as a metal , metal alloy , plastic , polymer , or other filamentous material or composite . the one or more loop portions may be attached at a second loop end to the control stem by soldering , fusing , an adhesive , or any other suitable means . in another embodiment , the loop portions may be attached to a first and a second loop end to the control stem . the support structure of fig6 a , 6 b , 7 a and 7 b may further include an atraumatic guiding tip of the kind described above to aid in positioning the support structure within the blood vessel . alternatively , parts of the loop portions which may protrude from the lumen when the loop portions are in their collapsed state may be used to guide the support structure into the blood vessel . one or more of the loop portions may be provided with a radiopaque or other marker to assist in this regard . retention means may also be provided with the support structure to retain the expandable member in an expanded condition within the vessel . the retention means may be in the form of a clamp , clip , thumb - slide or the like accessible from outside the patient &# 39 ; s body , and may facilitate adjustment of a deployed expandable member during a procedure . retention means may also impart additional rigidity and strength to the expandable member . thus , the retention member may be used to counteract excessively low pressures which may otherwise cause the expanded member to fail . a support structure of the kind illustrated in the figures may be delivered within a multilumen catheter 900 of the kind illustrated in cross section in fig9 . using this catheter , the support device 910 can be delivered through a first internal lumen 901 without interfering with flow in a second lumen 902 . a third lumen 903 may be provided for monitoring flow rates and pressures , for blood analysis or for delivering other percutaneous tools or devices to the vessel or as an inflation lumen for an occlusion balloon . it is to be understood that in the various embodiments of the present invention , the expanded member does not require constant contact with the vessel walls to provide the required support . for example , the diameter of the expanded member may be less than the diameter of the vessel so that the expanded member only contacts the vessel wall when the vessel begins to collapse . patency is considered to be maintained as long as the support device keeps the vessel open to a degree which is sufficient to maintain continuous flow . to avoid causing turbulence or other undesirable blood flow effects within the vessel , and to optimize flow in the vessel it may be desirable to substantially match the diameter of the expanded member to the diameter of the vessel . alternatively the expandable member may be shaped , e . g . as a coil or helix , to have minimal effect on the flow in the vessel . in one embodiment , the expandable member may have a slightly larger expanded diameter than the relaxed vessel to create an anchoring effect . depending on the size of the outflow vessel from which blood and perfusate is collected from the target region , there may be a natural tendency for the collection catheter tip to move about and contact the vessel , thus increasing the risk of vessel collapse or invagination of the catheter tip into the vessel wall . this can cause pooling of fluid in the isolated target region and may cause serious and permanent damage to the organ or region of the patient being treated . use of a support structure in conjunction with the collection catheter to maintain patency of the outflow vessel , in accordance with embodiments of the present invention can minimize the risk of these complications eventuating . thus , a collection catheter associated with the expanding member can be retained in position during fluid collection . this minimizes movement of the catheter tip , ensures that it is substantially centered relative to the vessel walls and improves withdrawal of fluid out of the vessel . at completion of the procedure , it is desirable that the expanded member is collapsed or compressed and recaptured , preferably in the catheter from which it was deployed . this facilitates removal of the support device from the patient . a reinforcing tip may be provided on the catheter end to strengthen it for recapture . alternatively or additionally , the tip may be coated with a lubricant and / or material having a low coefficient of friction to facilitate smooth recapture of the expandable member . the catheter may also have an internal coating of lubricant and / or a material having a low coefficient of friction to assist translation of support device along its interior during delivery and removal of the device from the patient . although the present invention has been described in terms of the presently preferred embodiments , it is to be understood that the disclosure is not to be interpreted as limiting . various alterations and modifications will no doubt become apparent to those skilled in the art after having read the above disclosure and it is intended that the present disclosure be interpreted as covering all alterations and modifications as fall within the true spirit and scope of the invention . effect of support device on flow rates and pressures achievable during recirculation in sheep right hepatic vein , cephalic vein , coronary sinus and renal vein during recirculation a 0 . 014 \u2033 diameter superelastic nitinol wire stem of 1 . 35 m length was used , coupled to an expandable member having 6 pre - shaped elliptical loop portions welded to the stem . a 0 . 024 \u2033 od atraumatic tip of 2 cm length attached to the distal end of the expandable member was used to position the device in the blood vessel . a balloon occlusion catheter was positioned in the vessel and the expandable member deployed at the tip of the catheter . the balloon was inflated to isolate and capture flows in the vessel and the catheter was connected to a standard extracorporeal circuit for blood circulation . negative pressures were observed in perfusion lines draining the coronary sinus , renal vein , right hepatic vein and cephalic vein during recirculation both with and without a support device . these data show that cavitation is prevented at certain pressures in the vessels tested where a support device is used , but is not prevented where the support device is absent in the vessel at those pressures . although cavitation may occur even with the support device , it occurs at higher flows . also , cavitation ceases sooner where the support device was employed allowing flow to return to normal . in the coronary sinus , recovery from cavitation was not possible without the support device , emphasizing the importance of the device in the procedure . the data further demonstrates that vessel collapse can be irreversible in the absence of a support structure . however , where a support structure is present , the vessel collapse may be reversed by increasing pressure in the vessel or by slowing or reversing the flow rate of fluid through the vessel . more specifically , considering the data for the right hepatic vein , flow rates of up to 250 ml per minute may be achieved before cavitation occurs where a support device is present in the vessel . under the same conditions but where there is no support device , flow rates of only up to 180 ml per minute are possible . a more striking example of the advantages of the support device is seen for the cephalic vein where no flow is achievable without the device . when the vessel wall is supported by the device flow rates of up to 200 ml per minute are noted before cavitation occurs . when the vessel wall is supported flow rates of up to 200 ml per minute are noted before cavitation occurs ."}
|
Does the category match the content of the patent?
| 0.25 |
1d5133cde40c92eaabe8131cbdf960efd3029fddffc551de233b4ea69b05bd19
| 0.017456 | 0.003601 | 0.099609 | 0.002319 | 0.094238 | 0.010986 |
null |
{"category": "Human Necessities", "patent": "while the support device of the present invention may be used in a range of different vessels , including blood vessels , it has particular application in procedures where an organ or anatomical region is undergoing localized perfusion with a therapeutic , diagnostic or other agent . for simplicity , these agents will be hereinafter referred to as therapeutic agents . however , it is to be understood that the term \u201c therapeutic \u201d is not to be construed as limiting , and that it includes , without limitation , therapeutic , diagnostic , prophylactic and other agents not specifically identified herein , but which would be considered by the relevant skilled addressee to be suitable for perfusion to an organ or anatomical region . perfusion may be total perfusion , where the entire organ is totally or substantially isolated from the systemic flow , or partial perfusion where only a portion of the organ is substantially isolated . localized perfusion of this kind presents advantages by improving efficacy and the time exposure of the therapeutic agent to the relevant cells . it also limits exposure and hence toxicity to non - target cells as described in brief above . however , it is to be understood that the present invention may also be used simply to collect or drain fluid from an organ or region . collected fluid may be removed from the subject and re - circulated into the organ , filtered and / or treated , or discarded . in some organs , it may be difficult to achieve total isolation , so partial isolation and perfusion may be performed , for example to the right or left lobe of the liver . despite partial perfusion being capable of delivering therapeutic agent to merely a part of the organ , significant therapeutic benefit may still be achieved . particular benefit may be achieved where perfusate is collected after perfusing the target organ , so as to prevent subsequent circulation of the therapeutic agent to other regions of the body where toxic effects may be observed , or the therapeutic agent wasted . the benefit may be improved further where collected perfusate is re - circulated into the target organ utilizing any therapeutic agent which remains after a first pass through the target organ . this may be achieved using the approach described in published patent application wo2005 / 082440 , the entire contents of which are herein incorporated by reference . as discussed infra , when fluid is collected from vessels draining from a target organ or region , one or more of these vessels may require cannulation with a collection catheter . when fluid is drained through these collection catheters , the vessels in which they are positioned become susceptible to collapse as the pressure inside decreases . while some vessels may be more susceptible to collapse than others , the support device of the present invention can provide advantages by supporting and stabilizing the vessel and even anchoring the collection catheter in position . the support device of the present invention may facilitate or at least improve the performance of perfusion . in some instances , the advantages of the present invention have been found to be essential to maintaining adequate positioning of collection catheters and flow rates within the vessel during perfusion . the right and left lobes of the liver have been identified as possible target regions and in this context , the support device may be deployed in one of the hepatic veins to support and maintain patency of the vein as fluid ( e . g . perfusate ) is collected from the liver . however , it is to be understood that fluid from many other organs or regions may be accessed in this way . deploying the support device may also protect the vessel wall by maintaining the tip of the catheter substantially centrally of the vessel or at least at a distance from the vessel walls to prevent aspiration or cavitation . deployment of the device may refer to partial or complete deployment . in complete deployment , the entire expandable member is released from the catheter and expanded to its full extent . in partial deployment , part of the expandable member is retained within the catheter and the amount of expansion is limited by the diameter of the catheter opening . partial deployment may be useful where , for example , during deployment it is found that the diameter of the expandable member may exceed the vessel diameter by an unsafe amount and complete deployment is likely to damage the vessel wall . limiting expansion of the device by partial deployment may avoid vessel damage . partial deployment may also stabilize the expandable member by limiting its movement relative to the catheter tip . thus by retaining part of the expandable member within the catheter , torsional , axial and lateral movement of the member , relative to the catheter is prevented or at least minimized by the struts of the expandable member being in abutment with the internal surface of the catheter . alternatively , the expandable member may be modified at the proximal end , for example by incorporating a lead , a link or other means to limit the extent of movement possible between the catheter tip and the expandable member once deployed . as a further positioning aid , markings may be provided at the proximal end of the control stem / shaft , outside the patient &# 39 ; s body . as the device is released into the vessel , the markings may be utilized to indicate the distance of device deployment , past the catheter tip . during collection of fluid from the vessel , low pressures may develop at the collection device tip , particularly where a roller / peristaltic pump or the like is used to draw fluid from the target organ out of the vessel . this may be indicated by pressures in a lumen feeding into the pump as low as , for example , \u2212 190 mmhg , although clearly these pressures are variable depending on the vessel type , health and age of the subject , characteristics of the perfusion circuit and the like . in the absence of the inventive support device , these pressures can cause the vessel to collapse . not only would vessel collapse affect the perfusion procedure , vessel collapse can also cause venous pooling in the organ and irreversible tissue damage . the advantages and benefits of the present invention will be expanded upon in the following detailed description presenting some of the preferred embodiments of the invention , and the specific examples which follow . it is to be understood that the embodiments and examples provided herein are intended to indicate how the present invention may be performed and are not intended to be limiting on the scope of protection sought as is defined in the claims appended hereto . fig1 a shows an example of an expandable member , in its expanded condition , suitable for supporting a vessel . expandable member 104 is provided in the form of an expandable framework and is adapted to be percutaneously deliverable to the blood vessel in a collapsed condition . fig1 b shows the expandable member in a collapsed condition within a catheter 110 , in which ends 105 , 107 have been drawn apart to radially reduce the member . when collapsed within catheter 110 , atraumatic tip 101 may protrude from the catheter to assist in guiding the support device into the vessel prior to deployment . when the expandable member has been guided into the target blood vessel , the catheter 110 is retracted ( or the expandable member is pushed out of the catheter ), deploying the device into the vessel where it expands . fig1 c shows the support device fully deployed from the catheter , with the expandable member in its fully expanded condition . a guidewire or stem 106 extends within the catheter 110 and is used to deliver the device from a point of entry through the peripheral vasculature to the target vessel . atraumatic tip 101 coupled to the expandable member 104 , is adapted to make atraumatic contact with vessel walls during placement of the device by deforming or deflecting off the vessel wall on contact . this can be achieved by incorporating flexibility into the tip so that it deforms upon contact with the vessel wall . alternatively or additionally , the tip may be shaped or curved to avoid trauma . the atraumatic tip may take any one of a number of forms . in the examples illustrated in fig1 to 3 , the atraumatic tip 101 , 201 is j - shaped . however , other shapes are considered to be suitable , including but not limited to those illustrated in fig4 . for example , the atraumatic tip may have a cross section which is enlarged relative to the guidewire radius , and have a smooth surface so as to avoid causing perforation when the tip comes into contact with the vessel wall . one such example is shown in fig4 a where the atraumatic tip 401 is tear - shaped . alternatively , the atraumatic tip may include a portion having a pigtail shaped curve 402 ( fig4 b ), or an angled tip ( not shown ). preferably , the expandable member is formed from a biocompatible superelastic material , or alternatively from a shape memory material or a material which exhibits both of these properties , being capable of recovery after deformation for delivery in a collapsed or compressed state within a catheter . devices manufactured using these materials can be collapsed for percutaneous delivery to a deployment site and then resume a known shape on deployment . a range of biocompatible materials may be suitable such as alloys of nickel and titanium ( e . g . nitinol ). other suitable biocompatible materials include but are not limited to polymers and plastics such as hydrophilic plastics , ceramics and the like . fig3 illustrates the support device of fig1 a to 1 c , with an occluding balloon inflated around catheter 110 . the occluding balloon 114 may be utilized during collection of fluid from an organ or region of the body in isolation , where substantially all of the fluid flowing out of the organ or region is collected by the catheter 110 . the occluding means substantially prevents blood , therapeutic agent and / or other fluids entering the vessel from flowing on to other organs or regions , and permits collection of substantially all of the fluid entering the vessel . collected fluid may then be analyzed and / or re - oxygenated and / or perfused through the organ , discarded or handled otherwise . the occlusion means may include an occluding balloon , flange , disc or other means . catheter 110 is delivered to the vessel with the balloon 114 in a deflated condition . the expandable member is delivered , through the catheter , and deployed inside the vessel . the balloon is then inflated around the catheter and substantially all the fluid in the vessel flows through the catheter and into a perfusion set or reservoir to which it is connected . a pump , syringe or other means may be incorporated into the perfusion set to draw fluid out of the vessel , through the catheter , at a rate which substantially maintains the required flow through the organ or region , or through a re - perfusion circuit . as fluid is drawn out of the vessel through the catheter , the expanded support structure supports the vessel walls , preventing collapse or cavitation which might otherwise result from the low pressures or high flow rates generated at the catheter tip , maintaining patency and ensuring flow in the circuit . the expandable member may also anchor the device in position within the vessel , substantially precluding movement of the device and ensuring that the catheter is retained in an optimal location for collection of fluid . the expandable member may take a range of different shapes when in an expanded ( or collapsed ) configuration , and may provide any number of supporting filaments or struts . the design of the expandable member may be based on a range of criteria including but not limited to the size and strength of the vessel wall and the flow rates and pressures likely to be generated near the device . some of these embodiments are illustrated in fig8 a to 8 c although these are examples only and are not intended to limit the scope of the invention as broadly described herein . fig8 a to 8 c illustrate expandable members having elongate portions in the supporting struts adapted for contact with the vessel wall . in the example in fig8 b , the supporting struts are slightly rounded to reduce trauma to the vessel walls . fig8 c provides additional struts when compared with fig8 a , as may be necessitated in particularly flaccid vessels requiring more substantial support . embodiments illustrated herein provide expandable members with a substantially elongate structure adapted for coaxial insertion into and placement within the vessel . the elongate structure supports the vessel over a length on the elongate portions of the struts substantially parallel to and in contact with the vessel wall . these elongate portions may be substantially straight , or may be curved ( e . g . fig8 b ). supporting the vessel wall over a length of the support device , compared with the point of supports of the prior art , improves the capacity of the device to maintain patency , even when very low pressures and high flow rates are generated at the catheter tip , and also reduces the likelihood of the device causing damage to the vessel wall . the elongate portions may have a length which is about the same as or greater than the diameter of the vessel being supported , or some multiple of the vessel diameter , or for example from 1 mm up to 30 mm depending on the vessel size and structure . the length of the elongate portion may be selected according to the vessel being supported , the size of the catheter being used and the flow rates and pressures likely to be generated at the catheter tip . preferably , the elongate portions of the expandable member which contact the vessel wall , are just adjacent the distal tip of the catheter when the device is fully deployed . thus , a proximal end of one or more of the elongate portions may commence , for example , within 0 . 1 to 25 mm of the catheter tip , or at least at a distance which is less than the diameter of the catheter opening . this prevents the vessel wall from being drawn into the space between the catheter tip and the start of the elongate portion of the expandable member which contacts the vessel wall . further , the device may be configured so that when it is in an expanded condition , the distance between adjacent elongate portions is sufficiently small to prevent the vessel wall from being drawn into gaps between them . for example , the distance between adjacent elongate portions may be less than the diameter of the catheter . alternatively , the distance between the adjacent elongate portions may be less than , for example , 3 , 2 . 5 , 2 , 1 . 5 , 1 or 0 . 5 mm , depending on the size and type of the target vessel , and the diameter of the collection catheter being used . preferably , the support device possesses sufficient mechanical strength to maintain patency during collection of fluid , withstanding the deformation forces which may occur in response to suction or low pressures produced at the collection catheter tip . in some embodiments however , it may also be desirable for the device to exhibit some flexibility , and conform to the shape of the vessel when deployed . thus , the support device is capable of providing support and maintaining patency along a length of the vessel , even where there is a curve in the vessel wall . an alternative embodiment of a support device 200 is illustrated in fig2 . proximal end 205 of the expandable member 204 is fixedly attached to a stem or shaft 206 , whereas distal end 203 of the expandable member is movable and able to slide over part of the shaft . this enables the member to collapse radially for delivery inside a delivery catheter , and also facilitates recapture of the device . fig5 illustrates another alternative embodiment of a support device shown at 500 in an expanded condition . in this embodiment , both the proximal end 505 and the distal end 503 of the expandable member are movable along a stem or shaft 506 used to deliver the device to the vessel . stops 508 a , 508 b are provided at fixed locations on a distal portion of the shaft , arranged between ends 503 , 505 of the expandable member . these stops may consist of a small ring , crimp or node of increased diameter , relative to the shaft diameter , and prevent the ends of the expandable member from moving across the stop . this facilitates deployment and retrieval of the expandable member from a catheter . fig1 illustrates a support device 151 consisting of an expandable framework 155 having a woven or braided , basket - like configuration when in the expanded condition . in this arrangement , the support device may also include occluding means in the form of a thin flow - proof coating 156 on the inner and / or outer surface of framework 155 to prevent flow of liquid from the vessel . thus , substantially all fluid in the vessel may be collected by catheter 160 . the flow - proof coating may be made from biocompatible silicon , elastomer or flow - proof polymer . preferably , the support device includes a radiopaque or other marker so that it can be positioned within the target vessel using an imaging system such as those generally known in the art . this enables the physician to position and deploy the expandable member into the blood vessel accurately . the marker may be incorporated into the expandable member and / or into an atraumatic guiding tip which may be incorporated into the support device . preferably , the atraumatic tip is manufactured from , includes or is coated with a lubricant and / or a material having a low coefficient of friction . many materials having low coefficient of friction properties may be used including but not limited to biocompatible high density polyethylene ( hdpe ), teflon \u00ae, polypropylene , polyethylene , microglide \u2122, low friction chromium and silicon to name a few . this improves the performance of the atraumatic tip , so that it \u201c slides \u201d along the vessel wall upon making contact , thereby substantially avoiding trauma . use of an atraumatic guiding tip improves the safety and ability to position the expandable member in the target vessel . moreover , since the atraumatic tip may exhibit greater flexibility than the rest of the device , the device is easier to manipulate into position . the atraumatic tip may be provided at a distance from the distal end of the expandable member which enables a physician to guide the expandable member into position within the target vessel . this distance may be anywhere from , for example , 0 . 25 to 5 centimeters from the distal end of the expandable member when in an expanded condition , although it is to be understood that larger or smaller distances may be utilized , depending on the location of the target vessel and the anatomy surrounding it . referring now to fig6 a and 6 b , another example of a support device 600 is shown . a lumen 602 has a control stem 601 extending therein . four loop portions 603 are provided . each loop portion is attached at a first loop end to a distal end 604 of the lumen , and at a second loop end to the control stem at 605 . the loop portions are controllably expandable by advancing the control stem within the lumen in the direction shown by arrow 606 ( fig6 b ). the support device is percutaneously deliverable with the plurality of loop portions housed substantially within the lumen 602 as illustrated in fig6 a and expandable as illustrated in fig6 b . whilst the embodiment illustrated in fig6 a and 6 b provides 4 loop portions , it is to be understood that any number of loop portions may be used . the number of loop portions incorporated into the device may depend on , for example , the anatomy of the vessel being supported , and / or the size of the catheter used to deliver the device . fig7 a and 7 b illustrate another example of a support device 700 which provides 3 loop portions 703 attached to control stem 701 at juncture 705 . the 3 loop portions are contained during delivery substantially within lumen 702 ( fig7 a ), and are controllably expandable to maintain patency within the blood vessel by advancing control stem 701 in the direction of arrow 706 ( fig7 b ). the rounded edges of the loop portions present a reduced risk of damaging the vessel walls , e . g . by perforation or bruising during delivery . the one or more loop portions may be attached to or near the distal end of the delivery lumen in any suitable manner . the point of attachment may be inside or outside the lumen . the loops may be manufactured from any suitable material such as a metal , metal alloy , plastic , polymer , or other filamentous material or composite . the one or more loop portions may be attached at a second loop end to the control stem by soldering , fusing , an adhesive , or any other suitable means . in another embodiment , the loop portions may be attached to a first and a second loop end to the control stem . the support structure of fig6 a , 6 b , 7 a and 7 b may further include an atraumatic guiding tip of the kind described above to aid in positioning the support structure within the blood vessel . alternatively , parts of the loop portions which may protrude from the lumen when the loop portions are in their collapsed state may be used to guide the support structure into the blood vessel . one or more of the loop portions may be provided with a radiopaque or other marker to assist in this regard . retention means may also be provided with the support structure to retain the expandable member in an expanded condition within the vessel . the retention means may be in the form of a clamp , clip , thumb - slide or the like accessible from outside the patient &# 39 ; s body , and may facilitate adjustment of a deployed expandable member during a procedure . retention means may also impart additional rigidity and strength to the expandable member . thus , the retention member may be used to counteract excessively low pressures which may otherwise cause the expanded member to fail . a support structure of the kind illustrated in the figures may be delivered within a multilumen catheter 900 of the kind illustrated in cross section in fig9 . using this catheter , the support device 910 can be delivered through a first internal lumen 901 without interfering with flow in a second lumen 902 . a third lumen 903 may be provided for monitoring flow rates and pressures , for blood analysis or for delivering other percutaneous tools or devices to the vessel or as an inflation lumen for an occlusion balloon . it is to be understood that in the various embodiments of the present invention , the expanded member does not require constant contact with the vessel walls to provide the required support . for example , the diameter of the expanded member may be less than the diameter of the vessel so that the expanded member only contacts the vessel wall when the vessel begins to collapse . patency is considered to be maintained as long as the support device keeps the vessel open to a degree which is sufficient to maintain continuous flow . to avoid causing turbulence or other undesirable blood flow effects within the vessel , and to optimize flow in the vessel it may be desirable to substantially match the diameter of the expanded member to the diameter of the vessel . alternatively the expandable member may be shaped , e . g . as a coil or helix , to have minimal effect on the flow in the vessel . in one embodiment , the expandable member may have a slightly larger expanded diameter than the relaxed vessel to create an anchoring effect . depending on the size of the outflow vessel from which blood and perfusate is collected from the target region , there may be a natural tendency for the collection catheter tip to move about and contact the vessel , thus increasing the risk of vessel collapse or invagination of the catheter tip into the vessel wall . this can cause pooling of fluid in the isolated target region and may cause serious and permanent damage to the organ or region of the patient being treated . use of a support structure in conjunction with the collection catheter to maintain patency of the outflow vessel , in accordance with embodiments of the present invention can minimize the risk of these complications eventuating . thus , a collection catheter associated with the expanding member can be retained in position during fluid collection . this minimizes movement of the catheter tip , ensures that it is substantially centered relative to the vessel walls and improves withdrawal of fluid out of the vessel . at completion of the procedure , it is desirable that the expanded member is collapsed or compressed and recaptured , preferably in the catheter from which it was deployed . this facilitates removal of the support device from the patient . a reinforcing tip may be provided on the catheter end to strengthen it for recapture . alternatively or additionally , the tip may be coated with a lubricant and / or material having a low coefficient of friction to facilitate smooth recapture of the expandable member . the catheter may also have an internal coating of lubricant and / or a material having a low coefficient of friction to assist translation of support device along its interior during delivery and removal of the device from the patient . although the present invention has been described in terms of the presently preferred embodiments , it is to be understood that the disclosure is not to be interpreted as limiting . various alterations and modifications will no doubt become apparent to those skilled in the art after having read the above disclosure and it is intended that the present disclosure be interpreted as covering all alterations and modifications as fall within the true spirit and scope of the invention . effect of support device on flow rates and pressures achievable during recirculation in sheep right hepatic vein , cephalic vein , coronary sinus and renal vein during recirculation a 0 . 014 \u2033 diameter superelastic nitinol wire stem of 1 . 35 m length was used , coupled to an expandable member having 6 pre - shaped elliptical loop portions welded to the stem . a 0 . 024 \u2033 od atraumatic tip of 2 cm length attached to the distal end of the expandable member was used to position the device in the blood vessel . a balloon occlusion catheter was positioned in the vessel and the expandable member deployed at the tip of the catheter . the balloon was inflated to isolate and capture flows in the vessel and the catheter was connected to a standard extracorporeal circuit for blood circulation . negative pressures were observed in perfusion lines draining the coronary sinus , renal vein , right hepatic vein and cephalic vein during recirculation both with and without a support device . these data show that cavitation is prevented at certain pressures in the vessels tested where a support device is used , but is not prevented where the support device is absent in the vessel at those pressures . although cavitation may occur even with the support device , it occurs at higher flows . also , cavitation ceases sooner where the support device was employed allowing flow to return to normal . in the coronary sinus , recovery from cavitation was not possible without the support device , emphasizing the importance of the device in the procedure . the data further demonstrates that vessel collapse can be irreversible in the absence of a support structure . however , where a support structure is present , the vessel collapse may be reversed by increasing pressure in the vessel or by slowing or reversing the flow rate of fluid through the vessel . more specifically , considering the data for the right hepatic vein , flow rates of up to 250 ml per minute may be achieved before cavitation occurs where a support device is present in the vessel . under the same conditions but where there is no support device , flow rates of only up to 180 ml per minute are possible . a more striking example of the advantages of the support device is seen for the cephalic vein where no flow is achievable without the device . when the vessel wall is supported by the device flow rates of up to 200 ml per minute are noted before cavitation occurs . when the vessel wall is supported flow rates of up to 200 ml per minute are noted before cavitation occurs ."}
|
{"category": "Physics", "patent": "while the support device of the present invention may be used in a range of different vessels , including blood vessels , it has particular application in procedures where an organ or anatomical region is undergoing localized perfusion with a therapeutic , diagnostic or other agent . for simplicity , these agents will be hereinafter referred to as therapeutic agents . however , it is to be understood that the term \u201c therapeutic \u201d is not to be construed as limiting , and that it includes , without limitation , therapeutic , diagnostic , prophylactic and other agents not specifically identified herein , but which would be considered by the relevant skilled addressee to be suitable for perfusion to an organ or anatomical region . perfusion may be total perfusion , where the entire organ is totally or substantially isolated from the systemic flow , or partial perfusion where only a portion of the organ is substantially isolated . localized perfusion of this kind presents advantages by improving efficacy and the time exposure of the therapeutic agent to the relevant cells . it also limits exposure and hence toxicity to non - target cells as described in brief above . however , it is to be understood that the present invention may also be used simply to collect or drain fluid from an organ or region . collected fluid may be removed from the subject and re - circulated into the organ , filtered and / or treated , or discarded . in some organs , it may be difficult to achieve total isolation , so partial isolation and perfusion may be performed , for example to the right or left lobe of the liver . despite partial perfusion being capable of delivering therapeutic agent to merely a part of the organ , significant therapeutic benefit may still be achieved . particular benefit may be achieved where perfusate is collected after perfusing the target organ , so as to prevent subsequent circulation of the therapeutic agent to other regions of the body where toxic effects may be observed , or the therapeutic agent wasted . the benefit may be improved further where collected perfusate is re - circulated into the target organ utilizing any therapeutic agent which remains after a first pass through the target organ . this may be achieved using the approach described in published patent application wo2005 / 082440 , the entire contents of which are herein incorporated by reference . as discussed infra , when fluid is collected from vessels draining from a target organ or region , one or more of these vessels may require cannulation with a collection catheter . when fluid is drained through these collection catheters , the vessels in which they are positioned become susceptible to collapse as the pressure inside decreases . while some vessels may be more susceptible to collapse than others , the support device of the present invention can provide advantages by supporting and stabilizing the vessel and even anchoring the collection catheter in position . the support device of the present invention may facilitate or at least improve the performance of perfusion . in some instances , the advantages of the present invention have been found to be essential to maintaining adequate positioning of collection catheters and flow rates within the vessel during perfusion . the right and left lobes of the liver have been identified as possible target regions and in this context , the support device may be deployed in one of the hepatic veins to support and maintain patency of the vein as fluid ( e . g . perfusate ) is collected from the liver . however , it is to be understood that fluid from many other organs or regions may be accessed in this way . deploying the support device may also protect the vessel wall by maintaining the tip of the catheter substantially centrally of the vessel or at least at a distance from the vessel walls to prevent aspiration or cavitation . deployment of the device may refer to partial or complete deployment . in complete deployment , the entire expandable member is released from the catheter and expanded to its full extent . in partial deployment , part of the expandable member is retained within the catheter and the amount of expansion is limited by the diameter of the catheter opening . partial deployment may be useful where , for example , during deployment it is found that the diameter of the expandable member may exceed the vessel diameter by an unsafe amount and complete deployment is likely to damage the vessel wall . limiting expansion of the device by partial deployment may avoid vessel damage . partial deployment may also stabilize the expandable member by limiting its movement relative to the catheter tip . thus by retaining part of the expandable member within the catheter , torsional , axial and lateral movement of the member , relative to the catheter is prevented or at least minimized by the struts of the expandable member being in abutment with the internal surface of the catheter . alternatively , the expandable member may be modified at the proximal end , for example by incorporating a lead , a link or other means to limit the extent of movement possible between the catheter tip and the expandable member once deployed . as a further positioning aid , markings may be provided at the proximal end of the control stem / shaft , outside the patient &# 39 ; s body . as the device is released into the vessel , the markings may be utilized to indicate the distance of device deployment , past the catheter tip . during collection of fluid from the vessel , low pressures may develop at the collection device tip , particularly where a roller / peristaltic pump or the like is used to draw fluid from the target organ out of the vessel . this may be indicated by pressures in a lumen feeding into the pump as low as , for example , \u2212 190 mmhg , although clearly these pressures are variable depending on the vessel type , health and age of the subject , characteristics of the perfusion circuit and the like . in the absence of the inventive support device , these pressures can cause the vessel to collapse . not only would vessel collapse affect the perfusion procedure , vessel collapse can also cause venous pooling in the organ and irreversible tissue damage . the advantages and benefits of the present invention will be expanded upon in the following detailed description presenting some of the preferred embodiments of the invention , and the specific examples which follow . it is to be understood that the embodiments and examples provided herein are intended to indicate how the present invention may be performed and are not intended to be limiting on the scope of protection sought as is defined in the claims appended hereto . fig1 a shows an example of an expandable member , in its expanded condition , suitable for supporting a vessel . expandable member 104 is provided in the form of an expandable framework and is adapted to be percutaneously deliverable to the blood vessel in a collapsed condition . fig1 b shows the expandable member in a collapsed condition within a catheter 110 , in which ends 105 , 107 have been drawn apart to radially reduce the member . when collapsed within catheter 110 , atraumatic tip 101 may protrude from the catheter to assist in guiding the support device into the vessel prior to deployment . when the expandable member has been guided into the target blood vessel , the catheter 110 is retracted ( or the expandable member is pushed out of the catheter ), deploying the device into the vessel where it expands . fig1 c shows the support device fully deployed from the catheter , with the expandable member in its fully expanded condition . a guidewire or stem 106 extends within the catheter 110 and is used to deliver the device from a point of entry through the peripheral vasculature to the target vessel . atraumatic tip 101 coupled to the expandable member 104 , is adapted to make atraumatic contact with vessel walls during placement of the device by deforming or deflecting off the vessel wall on contact . this can be achieved by incorporating flexibility into the tip so that it deforms upon contact with the vessel wall . alternatively or additionally , the tip may be shaped or curved to avoid trauma . the atraumatic tip may take any one of a number of forms . in the examples illustrated in fig1 to 3 , the atraumatic tip 101 , 201 is j - shaped . however , other shapes are considered to be suitable , including but not limited to those illustrated in fig4 . for example , the atraumatic tip may have a cross section which is enlarged relative to the guidewire radius , and have a smooth surface so as to avoid causing perforation when the tip comes into contact with the vessel wall . one such example is shown in fig4 a where the atraumatic tip 401 is tear - shaped . alternatively , the atraumatic tip may include a portion having a pigtail shaped curve 402 ( fig4 b ), or an angled tip ( not shown ). preferably , the expandable member is formed from a biocompatible superelastic material , or alternatively from a shape memory material or a material which exhibits both of these properties , being capable of recovery after deformation for delivery in a collapsed or compressed state within a catheter . devices manufactured using these materials can be collapsed for percutaneous delivery to a deployment site and then resume a known shape on deployment . a range of biocompatible materials may be suitable such as alloys of nickel and titanium ( e . g . nitinol ). other suitable biocompatible materials include but are not limited to polymers and plastics such as hydrophilic plastics , ceramics and the like . fig3 illustrates the support device of fig1 a to 1 c , with an occluding balloon inflated around catheter 110 . the occluding balloon 114 may be utilized during collection of fluid from an organ or region of the body in isolation , where substantially all of the fluid flowing out of the organ or region is collected by the catheter 110 . the occluding means substantially prevents blood , therapeutic agent and / or other fluids entering the vessel from flowing on to other organs or regions , and permits collection of substantially all of the fluid entering the vessel . collected fluid may then be analyzed and / or re - oxygenated and / or perfused through the organ , discarded or handled otherwise . the occlusion means may include an occluding balloon , flange , disc or other means . catheter 110 is delivered to the vessel with the balloon 114 in a deflated condition . the expandable member is delivered , through the catheter , and deployed inside the vessel . the balloon is then inflated around the catheter and substantially all the fluid in the vessel flows through the catheter and into a perfusion set or reservoir to which it is connected . a pump , syringe or other means may be incorporated into the perfusion set to draw fluid out of the vessel , through the catheter , at a rate which substantially maintains the required flow through the organ or region , or through a re - perfusion circuit . as fluid is drawn out of the vessel through the catheter , the expanded support structure supports the vessel walls , preventing collapse or cavitation which might otherwise result from the low pressures or high flow rates generated at the catheter tip , maintaining patency and ensuring flow in the circuit . the expandable member may also anchor the device in position within the vessel , substantially precluding movement of the device and ensuring that the catheter is retained in an optimal location for collection of fluid . the expandable member may take a range of different shapes when in an expanded ( or collapsed ) configuration , and may provide any number of supporting filaments or struts . the design of the expandable member may be based on a range of criteria including but not limited to the size and strength of the vessel wall and the flow rates and pressures likely to be generated near the device . some of these embodiments are illustrated in fig8 a to 8 c although these are examples only and are not intended to limit the scope of the invention as broadly described herein . fig8 a to 8 c illustrate expandable members having elongate portions in the supporting struts adapted for contact with the vessel wall . in the example in fig8 b , the supporting struts are slightly rounded to reduce trauma to the vessel walls . fig8 c provides additional struts when compared with fig8 a , as may be necessitated in particularly flaccid vessels requiring more substantial support . embodiments illustrated herein provide expandable members with a substantially elongate structure adapted for coaxial insertion into and placement within the vessel . the elongate structure supports the vessel over a length on the elongate portions of the struts substantially parallel to and in contact with the vessel wall . these elongate portions may be substantially straight , or may be curved ( e . g . fig8 b ). supporting the vessel wall over a length of the support device , compared with the point of supports of the prior art , improves the capacity of the device to maintain patency , even when very low pressures and high flow rates are generated at the catheter tip , and also reduces the likelihood of the device causing damage to the vessel wall . the elongate portions may have a length which is about the same as or greater than the diameter of the vessel being supported , or some multiple of the vessel diameter , or for example from 1 mm up to 30 mm depending on the vessel size and structure . the length of the elongate portion may be selected according to the vessel being supported , the size of the catheter being used and the flow rates and pressures likely to be generated at the catheter tip . preferably , the elongate portions of the expandable member which contact the vessel wall , are just adjacent the distal tip of the catheter when the device is fully deployed . thus , a proximal end of one or more of the elongate portions may commence , for example , within 0 . 1 to 25 mm of the catheter tip , or at least at a distance which is less than the diameter of the catheter opening . this prevents the vessel wall from being drawn into the space between the catheter tip and the start of the elongate portion of the expandable member which contacts the vessel wall . further , the device may be configured so that when it is in an expanded condition , the distance between adjacent elongate portions is sufficiently small to prevent the vessel wall from being drawn into gaps between them . for example , the distance between adjacent elongate portions may be less than the diameter of the catheter . alternatively , the distance between the adjacent elongate portions may be less than , for example , 3 , 2 . 5 , 2 , 1 . 5 , 1 or 0 . 5 mm , depending on the size and type of the target vessel , and the diameter of the collection catheter being used . preferably , the support device possesses sufficient mechanical strength to maintain patency during collection of fluid , withstanding the deformation forces which may occur in response to suction or low pressures produced at the collection catheter tip . in some embodiments however , it may also be desirable for the device to exhibit some flexibility , and conform to the shape of the vessel when deployed . thus , the support device is capable of providing support and maintaining patency along a length of the vessel , even where there is a curve in the vessel wall . an alternative embodiment of a support device 200 is illustrated in fig2 . proximal end 205 of the expandable member 204 is fixedly attached to a stem or shaft 206 , whereas distal end 203 of the expandable member is movable and able to slide over part of the shaft . this enables the member to collapse radially for delivery inside a delivery catheter , and also facilitates recapture of the device . fig5 illustrates another alternative embodiment of a support device shown at 500 in an expanded condition . in this embodiment , both the proximal end 505 and the distal end 503 of the expandable member are movable along a stem or shaft 506 used to deliver the device to the vessel . stops 508 a , 508 b are provided at fixed locations on a distal portion of the shaft , arranged between ends 503 , 505 of the expandable member . these stops may consist of a small ring , crimp or node of increased diameter , relative to the shaft diameter , and prevent the ends of the expandable member from moving across the stop . this facilitates deployment and retrieval of the expandable member from a catheter . fig1 illustrates a support device 151 consisting of an expandable framework 155 having a woven or braided , basket - like configuration when in the expanded condition . in this arrangement , the support device may also include occluding means in the form of a thin flow - proof coating 156 on the inner and / or outer surface of framework 155 to prevent flow of liquid from the vessel . thus , substantially all fluid in the vessel may be collected by catheter 160 . the flow - proof coating may be made from biocompatible silicon , elastomer or flow - proof polymer . preferably , the support device includes a radiopaque or other marker so that it can be positioned within the target vessel using an imaging system such as those generally known in the art . this enables the physician to position and deploy the expandable member into the blood vessel accurately . the marker may be incorporated into the expandable member and / or into an atraumatic guiding tip which may be incorporated into the support device . preferably , the atraumatic tip is manufactured from , includes or is coated with a lubricant and / or a material having a low coefficient of friction . many materials having low coefficient of friction properties may be used including but not limited to biocompatible high density polyethylene ( hdpe ), teflon \u00ae, polypropylene , polyethylene , microglide \u2122, low friction chromium and silicon to name a few . this improves the performance of the atraumatic tip , so that it \u201c slides \u201d along the vessel wall upon making contact , thereby substantially avoiding trauma . use of an atraumatic guiding tip improves the safety and ability to position the expandable member in the target vessel . moreover , since the atraumatic tip may exhibit greater flexibility than the rest of the device , the device is easier to manipulate into position . the atraumatic tip may be provided at a distance from the distal end of the expandable member which enables a physician to guide the expandable member into position within the target vessel . this distance may be anywhere from , for example , 0 . 25 to 5 centimeters from the distal end of the expandable member when in an expanded condition , although it is to be understood that larger or smaller distances may be utilized , depending on the location of the target vessel and the anatomy surrounding it . referring now to fig6 a and 6 b , another example of a support device 600 is shown . a lumen 602 has a control stem 601 extending therein . four loop portions 603 are provided . each loop portion is attached at a first loop end to a distal end 604 of the lumen , and at a second loop end to the control stem at 605 . the loop portions are controllably expandable by advancing the control stem within the lumen in the direction shown by arrow 606 ( fig6 b ). the support device is percutaneously deliverable with the plurality of loop portions housed substantially within the lumen 602 as illustrated in fig6 a and expandable as illustrated in fig6 b . whilst the embodiment illustrated in fig6 a and 6 b provides 4 loop portions , it is to be understood that any number of loop portions may be used . the number of loop portions incorporated into the device may depend on , for example , the anatomy of the vessel being supported , and / or the size of the catheter used to deliver the device . fig7 a and 7 b illustrate another example of a support device 700 which provides 3 loop portions 703 attached to control stem 701 at juncture 705 . the 3 loop portions are contained during delivery substantially within lumen 702 ( fig7 a ), and are controllably expandable to maintain patency within the blood vessel by advancing control stem 701 in the direction of arrow 706 ( fig7 b ). the rounded edges of the loop portions present a reduced risk of damaging the vessel walls , e . g . by perforation or bruising during delivery . the one or more loop portions may be attached to or near the distal end of the delivery lumen in any suitable manner . the point of attachment may be inside or outside the lumen . the loops may be manufactured from any suitable material such as a metal , metal alloy , plastic , polymer , or other filamentous material or composite . the one or more loop portions may be attached at a second loop end to the control stem by soldering , fusing , an adhesive , or any other suitable means . in another embodiment , the loop portions may be attached to a first and a second loop end to the control stem . the support structure of fig6 a , 6 b , 7 a and 7 b may further include an atraumatic guiding tip of the kind described above to aid in positioning the support structure within the blood vessel . alternatively , parts of the loop portions which may protrude from the lumen when the loop portions are in their collapsed state may be used to guide the support structure into the blood vessel . one or more of the loop portions may be provided with a radiopaque or other marker to assist in this regard . retention means may also be provided with the support structure to retain the expandable member in an expanded condition within the vessel . the retention means may be in the form of a clamp , clip , thumb - slide or the like accessible from outside the patient &# 39 ; s body , and may facilitate adjustment of a deployed expandable member during a procedure . retention means may also impart additional rigidity and strength to the expandable member . thus , the retention member may be used to counteract excessively low pressures which may otherwise cause the expanded member to fail . a support structure of the kind illustrated in the figures may be delivered within a multilumen catheter 900 of the kind illustrated in cross section in fig9 . using this catheter , the support device 910 can be delivered through a first internal lumen 901 without interfering with flow in a second lumen 902 . a third lumen 903 may be provided for monitoring flow rates and pressures , for blood analysis or for delivering other percutaneous tools or devices to the vessel or as an inflation lumen for an occlusion balloon . it is to be understood that in the various embodiments of the present invention , the expanded member does not require constant contact with the vessel walls to provide the required support . for example , the diameter of the expanded member may be less than the diameter of the vessel so that the expanded member only contacts the vessel wall when the vessel begins to collapse . patency is considered to be maintained as long as the support device keeps the vessel open to a degree which is sufficient to maintain continuous flow . to avoid causing turbulence or other undesirable blood flow effects within the vessel , and to optimize flow in the vessel it may be desirable to substantially match the diameter of the expanded member to the diameter of the vessel . alternatively the expandable member may be shaped , e . g . as a coil or helix , to have minimal effect on the flow in the vessel . in one embodiment , the expandable member may have a slightly larger expanded diameter than the relaxed vessel to create an anchoring effect . depending on the size of the outflow vessel from which blood and perfusate is collected from the target region , there may be a natural tendency for the collection catheter tip to move about and contact the vessel , thus increasing the risk of vessel collapse or invagination of the catheter tip into the vessel wall . this can cause pooling of fluid in the isolated target region and may cause serious and permanent damage to the organ or region of the patient being treated . use of a support structure in conjunction with the collection catheter to maintain patency of the outflow vessel , in accordance with embodiments of the present invention can minimize the risk of these complications eventuating . thus , a collection catheter associated with the expanding member can be retained in position during fluid collection . this minimizes movement of the catheter tip , ensures that it is substantially centered relative to the vessel walls and improves withdrawal of fluid out of the vessel . at completion of the procedure , it is desirable that the expanded member is collapsed or compressed and recaptured , preferably in the catheter from which it was deployed . this facilitates removal of the support device from the patient . a reinforcing tip may be provided on the catheter end to strengthen it for recapture . alternatively or additionally , the tip may be coated with a lubricant and / or material having a low coefficient of friction to facilitate smooth recapture of the expandable member . the catheter may also have an internal coating of lubricant and / or a material having a low coefficient of friction to assist translation of support device along its interior during delivery and removal of the device from the patient . although the present invention has been described in terms of the presently preferred embodiments , it is to be understood that the disclosure is not to be interpreted as limiting . various alterations and modifications will no doubt become apparent to those skilled in the art after having read the above disclosure and it is intended that the present disclosure be interpreted as covering all alterations and modifications as fall within the true spirit and scope of the invention . effect of support device on flow rates and pressures achievable during recirculation in sheep right hepatic vein , cephalic vein , coronary sinus and renal vein during recirculation a 0 . 014 \u2033 diameter superelastic nitinol wire stem of 1 . 35 m length was used , coupled to an expandable member having 6 pre - shaped elliptical loop portions welded to the stem . a 0 . 024 \u2033 od atraumatic tip of 2 cm length attached to the distal end of the expandable member was used to position the device in the blood vessel . a balloon occlusion catheter was positioned in the vessel and the expandable member deployed at the tip of the catheter . the balloon was inflated to isolate and capture flows in the vessel and the catheter was connected to a standard extracorporeal circuit for blood circulation . negative pressures were observed in perfusion lines draining the coronary sinus , renal vein , right hepatic vein and cephalic vein during recirculation both with and without a support device . these data show that cavitation is prevented at certain pressures in the vessels tested where a support device is used , but is not prevented where the support device is absent in the vessel at those pressures . although cavitation may occur even with the support device , it occurs at higher flows . also , cavitation ceases sooner where the support device was employed allowing flow to return to normal . in the coronary sinus , recovery from cavitation was not possible without the support device , emphasizing the importance of the device in the procedure . the data further demonstrates that vessel collapse can be irreversible in the absence of a support structure . however , where a support structure is present , the vessel collapse may be reversed by increasing pressure in the vessel or by slowing or reversing the flow rate of fluid through the vessel . more specifically , considering the data for the right hepatic vein , flow rates of up to 250 ml per minute may be achieved before cavitation occurs where a support device is present in the vessel . under the same conditions but where there is no support device , flow rates of only up to 180 ml per minute are possible . a more striking example of the advantages of the support device is seen for the cephalic vein where no flow is achievable without the device . when the vessel wall is supported by the device flow rates of up to 200 ml per minute are noted before cavitation occurs . when the vessel wall is supported flow rates of up to 200 ml per minute are noted before cavitation occurs ."}
|
Is the categorization of this patent accurate?
| 0.25 |
1d5133cde40c92eaabe8131cbdf960efd3029fddffc551de233b4ea69b05bd19
| 0.066406 | 0.031128 | 0.040283 | 0.029297 | 0.067383 | 0.056641 |
null |
{"category": "Human Necessities", "patent": "while the support device of the present invention may be used in a range of different vessels , including blood vessels , it has particular application in procedures where an organ or anatomical region is undergoing localized perfusion with a therapeutic , diagnostic or other agent . for simplicity , these agents will be hereinafter referred to as therapeutic agents . however , it is to be understood that the term \u201c therapeutic \u201d is not to be construed as limiting , and that it includes , without limitation , therapeutic , diagnostic , prophylactic and other agents not specifically identified herein , but which would be considered by the relevant skilled addressee to be suitable for perfusion to an organ or anatomical region . perfusion may be total perfusion , where the entire organ is totally or substantially isolated from the systemic flow , or partial perfusion where only a portion of the organ is substantially isolated . localized perfusion of this kind presents advantages by improving efficacy and the time exposure of the therapeutic agent to the relevant cells . it also limits exposure and hence toxicity to non - target cells as described in brief above . however , it is to be understood that the present invention may also be used simply to collect or drain fluid from an organ or region . collected fluid may be removed from the subject and re - circulated into the organ , filtered and / or treated , or discarded . in some organs , it may be difficult to achieve total isolation , so partial isolation and perfusion may be performed , for example to the right or left lobe of the liver . despite partial perfusion being capable of delivering therapeutic agent to merely a part of the organ , significant therapeutic benefit may still be achieved . particular benefit may be achieved where perfusate is collected after perfusing the target organ , so as to prevent subsequent circulation of the therapeutic agent to other regions of the body where toxic effects may be observed , or the therapeutic agent wasted . the benefit may be improved further where collected perfusate is re - circulated into the target organ utilizing any therapeutic agent which remains after a first pass through the target organ . this may be achieved using the approach described in published patent application wo2005 / 082440 , the entire contents of which are herein incorporated by reference . as discussed infra , when fluid is collected from vessels draining from a target organ or region , one or more of these vessels may require cannulation with a collection catheter . when fluid is drained through these collection catheters , the vessels in which they are positioned become susceptible to collapse as the pressure inside decreases . while some vessels may be more susceptible to collapse than others , the support device of the present invention can provide advantages by supporting and stabilizing the vessel and even anchoring the collection catheter in position . the support device of the present invention may facilitate or at least improve the performance of perfusion . in some instances , the advantages of the present invention have been found to be essential to maintaining adequate positioning of collection catheters and flow rates within the vessel during perfusion . the right and left lobes of the liver have been identified as possible target regions and in this context , the support device may be deployed in one of the hepatic veins to support and maintain patency of the vein as fluid ( e . g . perfusate ) is collected from the liver . however , it is to be understood that fluid from many other organs or regions may be accessed in this way . deploying the support device may also protect the vessel wall by maintaining the tip of the catheter substantially centrally of the vessel or at least at a distance from the vessel walls to prevent aspiration or cavitation . deployment of the device may refer to partial or complete deployment . in complete deployment , the entire expandable member is released from the catheter and expanded to its full extent . in partial deployment , part of the expandable member is retained within the catheter and the amount of expansion is limited by the diameter of the catheter opening . partial deployment may be useful where , for example , during deployment it is found that the diameter of the expandable member may exceed the vessel diameter by an unsafe amount and complete deployment is likely to damage the vessel wall . limiting expansion of the device by partial deployment may avoid vessel damage . partial deployment may also stabilize the expandable member by limiting its movement relative to the catheter tip . thus by retaining part of the expandable member within the catheter , torsional , axial and lateral movement of the member , relative to the catheter is prevented or at least minimized by the struts of the expandable member being in abutment with the internal surface of the catheter . alternatively , the expandable member may be modified at the proximal end , for example by incorporating a lead , a link or other means to limit the extent of movement possible between the catheter tip and the expandable member once deployed . as a further positioning aid , markings may be provided at the proximal end of the control stem / shaft , outside the patient &# 39 ; s body . as the device is released into the vessel , the markings may be utilized to indicate the distance of device deployment , past the catheter tip . during collection of fluid from the vessel , low pressures may develop at the collection device tip , particularly where a roller / peristaltic pump or the like is used to draw fluid from the target organ out of the vessel . this may be indicated by pressures in a lumen feeding into the pump as low as , for example , \u2212 190 mmhg , although clearly these pressures are variable depending on the vessel type , health and age of the subject , characteristics of the perfusion circuit and the like . in the absence of the inventive support device , these pressures can cause the vessel to collapse . not only would vessel collapse affect the perfusion procedure , vessel collapse can also cause venous pooling in the organ and irreversible tissue damage . the advantages and benefits of the present invention will be expanded upon in the following detailed description presenting some of the preferred embodiments of the invention , and the specific examples which follow . it is to be understood that the embodiments and examples provided herein are intended to indicate how the present invention may be performed and are not intended to be limiting on the scope of protection sought as is defined in the claims appended hereto . fig1 a shows an example of an expandable member , in its expanded condition , suitable for supporting a vessel . expandable member 104 is provided in the form of an expandable framework and is adapted to be percutaneously deliverable to the blood vessel in a collapsed condition . fig1 b shows the expandable member in a collapsed condition within a catheter 110 , in which ends 105 , 107 have been drawn apart to radially reduce the member . when collapsed within catheter 110 , atraumatic tip 101 may protrude from the catheter to assist in guiding the support device into the vessel prior to deployment . when the expandable member has been guided into the target blood vessel , the catheter 110 is retracted ( or the expandable member is pushed out of the catheter ), deploying the device into the vessel where it expands . fig1 c shows the support device fully deployed from the catheter , with the expandable member in its fully expanded condition . a guidewire or stem 106 extends within the catheter 110 and is used to deliver the device from a point of entry through the peripheral vasculature to the target vessel . atraumatic tip 101 coupled to the expandable member 104 , is adapted to make atraumatic contact with vessel walls during placement of the device by deforming or deflecting off the vessel wall on contact . this can be achieved by incorporating flexibility into the tip so that it deforms upon contact with the vessel wall . alternatively or additionally , the tip may be shaped or curved to avoid trauma . the atraumatic tip may take any one of a number of forms . in the examples illustrated in fig1 to 3 , the atraumatic tip 101 , 201 is j - shaped . however , other shapes are considered to be suitable , including but not limited to those illustrated in fig4 . for example , the atraumatic tip may have a cross section which is enlarged relative to the guidewire radius , and have a smooth surface so as to avoid causing perforation when the tip comes into contact with the vessel wall . one such example is shown in fig4 a where the atraumatic tip 401 is tear - shaped . alternatively , the atraumatic tip may include a portion having a pigtail shaped curve 402 ( fig4 b ), or an angled tip ( not shown ). preferably , the expandable member is formed from a biocompatible superelastic material , or alternatively from a shape memory material or a material which exhibits both of these properties , being capable of recovery after deformation for delivery in a collapsed or compressed state within a catheter . devices manufactured using these materials can be collapsed for percutaneous delivery to a deployment site and then resume a known shape on deployment . a range of biocompatible materials may be suitable such as alloys of nickel and titanium ( e . g . nitinol ). other suitable biocompatible materials include but are not limited to polymers and plastics such as hydrophilic plastics , ceramics and the like . fig3 illustrates the support device of fig1 a to 1 c , with an occluding balloon inflated around catheter 110 . the occluding balloon 114 may be utilized during collection of fluid from an organ or region of the body in isolation , where substantially all of the fluid flowing out of the organ or region is collected by the catheter 110 . the occluding means substantially prevents blood , therapeutic agent and / or other fluids entering the vessel from flowing on to other organs or regions , and permits collection of substantially all of the fluid entering the vessel . collected fluid may then be analyzed and / or re - oxygenated and / or perfused through the organ , discarded or handled otherwise . the occlusion means may include an occluding balloon , flange , disc or other means . catheter 110 is delivered to the vessel with the balloon 114 in a deflated condition . the expandable member is delivered , through the catheter , and deployed inside the vessel . the balloon is then inflated around the catheter and substantially all the fluid in the vessel flows through the catheter and into a perfusion set or reservoir to which it is connected . a pump , syringe or other means may be incorporated into the perfusion set to draw fluid out of the vessel , through the catheter , at a rate which substantially maintains the required flow through the organ or region , or through a re - perfusion circuit . as fluid is drawn out of the vessel through the catheter , the expanded support structure supports the vessel walls , preventing collapse or cavitation which might otherwise result from the low pressures or high flow rates generated at the catheter tip , maintaining patency and ensuring flow in the circuit . the expandable member may also anchor the device in position within the vessel , substantially precluding movement of the device and ensuring that the catheter is retained in an optimal location for collection of fluid . the expandable member may take a range of different shapes when in an expanded ( or collapsed ) configuration , and may provide any number of supporting filaments or struts . the design of the expandable member may be based on a range of criteria including but not limited to the size and strength of the vessel wall and the flow rates and pressures likely to be generated near the device . some of these embodiments are illustrated in fig8 a to 8 c although these are examples only and are not intended to limit the scope of the invention as broadly described herein . fig8 a to 8 c illustrate expandable members having elongate portions in the supporting struts adapted for contact with the vessel wall . in the example in fig8 b , the supporting struts are slightly rounded to reduce trauma to the vessel walls . fig8 c provides additional struts when compared with fig8 a , as may be necessitated in particularly flaccid vessels requiring more substantial support . embodiments illustrated herein provide expandable members with a substantially elongate structure adapted for coaxial insertion into and placement within the vessel . the elongate structure supports the vessel over a length on the elongate portions of the struts substantially parallel to and in contact with the vessel wall . these elongate portions may be substantially straight , or may be curved ( e . g . fig8 b ). supporting the vessel wall over a length of the support device , compared with the point of supports of the prior art , improves the capacity of the device to maintain patency , even when very low pressures and high flow rates are generated at the catheter tip , and also reduces the likelihood of the device causing damage to the vessel wall . the elongate portions may have a length which is about the same as or greater than the diameter of the vessel being supported , or some multiple of the vessel diameter , or for example from 1 mm up to 30 mm depending on the vessel size and structure . the length of the elongate portion may be selected according to the vessel being supported , the size of the catheter being used and the flow rates and pressures likely to be generated at the catheter tip . preferably , the elongate portions of the expandable member which contact the vessel wall , are just adjacent the distal tip of the catheter when the device is fully deployed . thus , a proximal end of one or more of the elongate portions may commence , for example , within 0 . 1 to 25 mm of the catheter tip , or at least at a distance which is less than the diameter of the catheter opening . this prevents the vessel wall from being drawn into the space between the catheter tip and the start of the elongate portion of the expandable member which contacts the vessel wall . further , the device may be configured so that when it is in an expanded condition , the distance between adjacent elongate portions is sufficiently small to prevent the vessel wall from being drawn into gaps between them . for example , the distance between adjacent elongate portions may be less than the diameter of the catheter . alternatively , the distance between the adjacent elongate portions may be less than , for example , 3 , 2 . 5 , 2 , 1 . 5 , 1 or 0 . 5 mm , depending on the size and type of the target vessel , and the diameter of the collection catheter being used . preferably , the support device possesses sufficient mechanical strength to maintain patency during collection of fluid , withstanding the deformation forces which may occur in response to suction or low pressures produced at the collection catheter tip . in some embodiments however , it may also be desirable for the device to exhibit some flexibility , and conform to the shape of the vessel when deployed . thus , the support device is capable of providing support and maintaining patency along a length of the vessel , even where there is a curve in the vessel wall . an alternative embodiment of a support device 200 is illustrated in fig2 . proximal end 205 of the expandable member 204 is fixedly attached to a stem or shaft 206 , whereas distal end 203 of the expandable member is movable and able to slide over part of the shaft . this enables the member to collapse radially for delivery inside a delivery catheter , and also facilitates recapture of the device . fig5 illustrates another alternative embodiment of a support device shown at 500 in an expanded condition . in this embodiment , both the proximal end 505 and the distal end 503 of the expandable member are movable along a stem or shaft 506 used to deliver the device to the vessel . stops 508 a , 508 b are provided at fixed locations on a distal portion of the shaft , arranged between ends 503 , 505 of the expandable member . these stops may consist of a small ring , crimp or node of increased diameter , relative to the shaft diameter , and prevent the ends of the expandable member from moving across the stop . this facilitates deployment and retrieval of the expandable member from a catheter . fig1 illustrates a support device 151 consisting of an expandable framework 155 having a woven or braided , basket - like configuration when in the expanded condition . in this arrangement , the support device may also include occluding means in the form of a thin flow - proof coating 156 on the inner and / or outer surface of framework 155 to prevent flow of liquid from the vessel . thus , substantially all fluid in the vessel may be collected by catheter 160 . the flow - proof coating may be made from biocompatible silicon , elastomer or flow - proof polymer . preferably , the support device includes a radiopaque or other marker so that it can be positioned within the target vessel using an imaging system such as those generally known in the art . this enables the physician to position and deploy the expandable member into the blood vessel accurately . the marker may be incorporated into the expandable member and / or into an atraumatic guiding tip which may be incorporated into the support device . preferably , the atraumatic tip is manufactured from , includes or is coated with a lubricant and / or a material having a low coefficient of friction . many materials having low coefficient of friction properties may be used including but not limited to biocompatible high density polyethylene ( hdpe ), teflon \u00ae, polypropylene , polyethylene , microglide \u2122, low friction chromium and silicon to name a few . this improves the performance of the atraumatic tip , so that it \u201c slides \u201d along the vessel wall upon making contact , thereby substantially avoiding trauma . use of an atraumatic guiding tip improves the safety and ability to position the expandable member in the target vessel . moreover , since the atraumatic tip may exhibit greater flexibility than the rest of the device , the device is easier to manipulate into position . the atraumatic tip may be provided at a distance from the distal end of the expandable member which enables a physician to guide the expandable member into position within the target vessel . this distance may be anywhere from , for example , 0 . 25 to 5 centimeters from the distal end of the expandable member when in an expanded condition , although it is to be understood that larger or smaller distances may be utilized , depending on the location of the target vessel and the anatomy surrounding it . referring now to fig6 a and 6 b , another example of a support device 600 is shown . a lumen 602 has a control stem 601 extending therein . four loop portions 603 are provided . each loop portion is attached at a first loop end to a distal end 604 of the lumen , and at a second loop end to the control stem at 605 . the loop portions are controllably expandable by advancing the control stem within the lumen in the direction shown by arrow 606 ( fig6 b ). the support device is percutaneously deliverable with the plurality of loop portions housed substantially within the lumen 602 as illustrated in fig6 a and expandable as illustrated in fig6 b . whilst the embodiment illustrated in fig6 a and 6 b provides 4 loop portions , it is to be understood that any number of loop portions may be used . the number of loop portions incorporated into the device may depend on , for example , the anatomy of the vessel being supported , and / or the size of the catheter used to deliver the device . fig7 a and 7 b illustrate another example of a support device 700 which provides 3 loop portions 703 attached to control stem 701 at juncture 705 . the 3 loop portions are contained during delivery substantially within lumen 702 ( fig7 a ), and are controllably expandable to maintain patency within the blood vessel by advancing control stem 701 in the direction of arrow 706 ( fig7 b ). the rounded edges of the loop portions present a reduced risk of damaging the vessel walls , e . g . by perforation or bruising during delivery . the one or more loop portions may be attached to or near the distal end of the delivery lumen in any suitable manner . the point of attachment may be inside or outside the lumen . the loops may be manufactured from any suitable material such as a metal , metal alloy , plastic , polymer , or other filamentous material or composite . the one or more loop portions may be attached at a second loop end to the control stem by soldering , fusing , an adhesive , or any other suitable means . in another embodiment , the loop portions may be attached to a first and a second loop end to the control stem . the support structure of fig6 a , 6 b , 7 a and 7 b may further include an atraumatic guiding tip of the kind described above to aid in positioning the support structure within the blood vessel . alternatively , parts of the loop portions which may protrude from the lumen when the loop portions are in their collapsed state may be used to guide the support structure into the blood vessel . one or more of the loop portions may be provided with a radiopaque or other marker to assist in this regard . retention means may also be provided with the support structure to retain the expandable member in an expanded condition within the vessel . the retention means may be in the form of a clamp , clip , thumb - slide or the like accessible from outside the patient &# 39 ; s body , and may facilitate adjustment of a deployed expandable member during a procedure . retention means may also impart additional rigidity and strength to the expandable member . thus , the retention member may be used to counteract excessively low pressures which may otherwise cause the expanded member to fail . a support structure of the kind illustrated in the figures may be delivered within a multilumen catheter 900 of the kind illustrated in cross section in fig9 . using this catheter , the support device 910 can be delivered through a first internal lumen 901 without interfering with flow in a second lumen 902 . a third lumen 903 may be provided for monitoring flow rates and pressures , for blood analysis or for delivering other percutaneous tools or devices to the vessel or as an inflation lumen for an occlusion balloon . it is to be understood that in the various embodiments of the present invention , the expanded member does not require constant contact with the vessel walls to provide the required support . for example , the diameter of the expanded member may be less than the diameter of the vessel so that the expanded member only contacts the vessel wall when the vessel begins to collapse . patency is considered to be maintained as long as the support device keeps the vessel open to a degree which is sufficient to maintain continuous flow . to avoid causing turbulence or other undesirable blood flow effects within the vessel , and to optimize flow in the vessel it may be desirable to substantially match the diameter of the expanded member to the diameter of the vessel . alternatively the expandable member may be shaped , e . g . as a coil or helix , to have minimal effect on the flow in the vessel . in one embodiment , the expandable member may have a slightly larger expanded diameter than the relaxed vessel to create an anchoring effect . depending on the size of the outflow vessel from which blood and perfusate is collected from the target region , there may be a natural tendency for the collection catheter tip to move about and contact the vessel , thus increasing the risk of vessel collapse or invagination of the catheter tip into the vessel wall . this can cause pooling of fluid in the isolated target region and may cause serious and permanent damage to the organ or region of the patient being treated . use of a support structure in conjunction with the collection catheter to maintain patency of the outflow vessel , in accordance with embodiments of the present invention can minimize the risk of these complications eventuating . thus , a collection catheter associated with the expanding member can be retained in position during fluid collection . this minimizes movement of the catheter tip , ensures that it is substantially centered relative to the vessel walls and improves withdrawal of fluid out of the vessel . at completion of the procedure , it is desirable that the expanded member is collapsed or compressed and recaptured , preferably in the catheter from which it was deployed . this facilitates removal of the support device from the patient . a reinforcing tip may be provided on the catheter end to strengthen it for recapture . alternatively or additionally , the tip may be coated with a lubricant and / or material having a low coefficient of friction to facilitate smooth recapture of the expandable member . the catheter may also have an internal coating of lubricant and / or a material having a low coefficient of friction to assist translation of support device along its interior during delivery and removal of the device from the patient . although the present invention has been described in terms of the presently preferred embodiments , it is to be understood that the disclosure is not to be interpreted as limiting . various alterations and modifications will no doubt become apparent to those skilled in the art after having read the above disclosure and it is intended that the present disclosure be interpreted as covering all alterations and modifications as fall within the true spirit and scope of the invention . effect of support device on flow rates and pressures achievable during recirculation in sheep right hepatic vein , cephalic vein , coronary sinus and renal vein during recirculation a 0 . 014 \u2033 diameter superelastic nitinol wire stem of 1 . 35 m length was used , coupled to an expandable member having 6 pre - shaped elliptical loop portions welded to the stem . a 0 . 024 \u2033 od atraumatic tip of 2 cm length attached to the distal end of the expandable member was used to position the device in the blood vessel . a balloon occlusion catheter was positioned in the vessel and the expandable member deployed at the tip of the catheter . the balloon was inflated to isolate and capture flows in the vessel and the catheter was connected to a standard extracorporeal circuit for blood circulation . negative pressures were observed in perfusion lines draining the coronary sinus , renal vein , right hepatic vein and cephalic vein during recirculation both with and without a support device . these data show that cavitation is prevented at certain pressures in the vessels tested where a support device is used , but is not prevented where the support device is absent in the vessel at those pressures . although cavitation may occur even with the support device , it occurs at higher flows . also , cavitation ceases sooner where the support device was employed allowing flow to return to normal . in the coronary sinus , recovery from cavitation was not possible without the support device , emphasizing the importance of the device in the procedure . the data further demonstrates that vessel collapse can be irreversible in the absence of a support structure . however , where a support structure is present , the vessel collapse may be reversed by increasing pressure in the vessel or by slowing or reversing the flow rate of fluid through the vessel . more specifically , considering the data for the right hepatic vein , flow rates of up to 250 ml per minute may be achieved before cavitation occurs where a support device is present in the vessel . under the same conditions but where there is no support device , flow rates of only up to 180 ml per minute are possible . a more striking example of the advantages of the support device is seen for the cephalic vein where no flow is achievable without the device . when the vessel wall is supported by the device flow rates of up to 200 ml per minute are noted before cavitation occurs . when the vessel wall is supported flow rates of up to 200 ml per minute are noted before cavitation occurs ."}
|
{"patent": "while the support device of the present invention may be used in a range of different vessels , including blood vessels , it has particular application in procedures where an organ or anatomical region is undergoing localized perfusion with a therapeutic , diagnostic or other agent . for simplicity , these agents will be hereinafter referred to as therapeutic agents . however , it is to be understood that the term \u201c therapeutic \u201d is not to be construed as limiting , and that it includes , without limitation , therapeutic , diagnostic , prophylactic and other agents not specifically identified herein , but which would be considered by the relevant skilled addressee to be suitable for perfusion to an organ or anatomical region . perfusion may be total perfusion , where the entire organ is totally or substantially isolated from the systemic flow , or partial perfusion where only a portion of the organ is substantially isolated . localized perfusion of this kind presents advantages by improving efficacy and the time exposure of the therapeutic agent to the relevant cells . it also limits exposure and hence toxicity to non - target cells as described in brief above . however , it is to be understood that the present invention may also be used simply to collect or drain fluid from an organ or region . collected fluid may be removed from the subject and re - circulated into the organ , filtered and / or treated , or discarded . in some organs , it may be difficult to achieve total isolation , so partial isolation and perfusion may be performed , for example to the right or left lobe of the liver . despite partial perfusion being capable of delivering therapeutic agent to merely a part of the organ , significant therapeutic benefit may still be achieved . particular benefit may be achieved where perfusate is collected after perfusing the target organ , so as to prevent subsequent circulation of the therapeutic agent to other regions of the body where toxic effects may be observed , or the therapeutic agent wasted . the benefit may be improved further where collected perfusate is re - circulated into the target organ utilizing any therapeutic agent which remains after a first pass through the target organ . this may be achieved using the approach described in published patent application wo2005 / 082440 , the entire contents of which are herein incorporated by reference . as discussed infra , when fluid is collected from vessels draining from a target organ or region , one or more of these vessels may require cannulation with a collection catheter . when fluid is drained through these collection catheters , the vessels in which they are positioned become susceptible to collapse as the pressure inside decreases . while some vessels may be more susceptible to collapse than others , the support device of the present invention can provide advantages by supporting and stabilizing the vessel and even anchoring the collection catheter in position . the support device of the present invention may facilitate or at least improve the performance of perfusion . in some instances , the advantages of the present invention have been found to be essential to maintaining adequate positioning of collection catheters and flow rates within the vessel during perfusion . the right and left lobes of the liver have been identified as possible target regions and in this context , the support device may be deployed in one of the hepatic veins to support and maintain patency of the vein as fluid ( e . g . perfusate ) is collected from the liver . however , it is to be understood that fluid from many other organs or regions may be accessed in this way . deploying the support device may also protect the vessel wall by maintaining the tip of the catheter substantially centrally of the vessel or at least at a distance from the vessel walls to prevent aspiration or cavitation . deployment of the device may refer to partial or complete deployment . in complete deployment , the entire expandable member is released from the catheter and expanded to its full extent . in partial deployment , part of the expandable member is retained within the catheter and the amount of expansion is limited by the diameter of the catheter opening . partial deployment may be useful where , for example , during deployment it is found that the diameter of the expandable member may exceed the vessel diameter by an unsafe amount and complete deployment is likely to damage the vessel wall . limiting expansion of the device by partial deployment may avoid vessel damage . partial deployment may also stabilize the expandable member by limiting its movement relative to the catheter tip . thus by retaining part of the expandable member within the catheter , torsional , axial and lateral movement of the member , relative to the catheter is prevented or at least minimized by the struts of the expandable member being in abutment with the internal surface of the catheter . alternatively , the expandable member may be modified at the proximal end , for example by incorporating a lead , a link or other means to limit the extent of movement possible between the catheter tip and the expandable member once deployed . as a further positioning aid , markings may be provided at the proximal end of the control stem / shaft , outside the patient &# 39 ; s body . as the device is released into the vessel , the markings may be utilized to indicate the distance of device deployment , past the catheter tip . during collection of fluid from the vessel , low pressures may develop at the collection device tip , particularly where a roller / peristaltic pump or the like is used to draw fluid from the target organ out of the vessel . this may be indicated by pressures in a lumen feeding into the pump as low as , for example , \u2212 190 mmhg , although clearly these pressures are variable depending on the vessel type , health and age of the subject , characteristics of the perfusion circuit and the like . in the absence of the inventive support device , these pressures can cause the vessel to collapse . not only would vessel collapse affect the perfusion procedure , vessel collapse can also cause venous pooling in the organ and irreversible tissue damage . the advantages and benefits of the present invention will be expanded upon in the following detailed description presenting some of the preferred embodiments of the invention , and the specific examples which follow . it is to be understood that the embodiments and examples provided herein are intended to indicate how the present invention may be performed and are not intended to be limiting on the scope of protection sought as is defined in the claims appended hereto . fig1 a shows an example of an expandable member , in its expanded condition , suitable for supporting a vessel . expandable member 104 is provided in the form of an expandable framework and is adapted to be percutaneously deliverable to the blood vessel in a collapsed condition . fig1 b shows the expandable member in a collapsed condition within a catheter 110 , in which ends 105 , 107 have been drawn apart to radially reduce the member . when collapsed within catheter 110 , atraumatic tip 101 may protrude from the catheter to assist in guiding the support device into the vessel prior to deployment . when the expandable member has been guided into the target blood vessel , the catheter 110 is retracted ( or the expandable member is pushed out of the catheter ), deploying the device into the vessel where it expands . fig1 c shows the support device fully deployed from the catheter , with the expandable member in its fully expanded condition . a guidewire or stem 106 extends within the catheter 110 and is used to deliver the device from a point of entry through the peripheral vasculature to the target vessel . atraumatic tip 101 coupled to the expandable member 104 , is adapted to make atraumatic contact with vessel walls during placement of the device by deforming or deflecting off the vessel wall on contact . this can be achieved by incorporating flexibility into the tip so that it deforms upon contact with the vessel wall . alternatively or additionally , the tip may be shaped or curved to avoid trauma . the atraumatic tip may take any one of a number of forms . in the examples illustrated in fig1 to 3 , the atraumatic tip 101 , 201 is j - shaped . however , other shapes are considered to be suitable , including but not limited to those illustrated in fig4 . for example , the atraumatic tip may have a cross section which is enlarged relative to the guidewire radius , and have a smooth surface so as to avoid causing perforation when the tip comes into contact with the vessel wall . one such example is shown in fig4 a where the atraumatic tip 401 is tear - shaped . alternatively , the atraumatic tip may include a portion having a pigtail shaped curve 402 ( fig4 b ), or an angled tip ( not shown ). preferably , the expandable member is formed from a biocompatible superelastic material , or alternatively from a shape memory material or a material which exhibits both of these properties , being capable of recovery after deformation for delivery in a collapsed or compressed state within a catheter . devices manufactured using these materials can be collapsed for percutaneous delivery to a deployment site and then resume a known shape on deployment . a range of biocompatible materials may be suitable such as alloys of nickel and titanium ( e . g . nitinol ). other suitable biocompatible materials include but are not limited to polymers and plastics such as hydrophilic plastics , ceramics and the like . fig3 illustrates the support device of fig1 a to 1 c , with an occluding balloon inflated around catheter 110 . the occluding balloon 114 may be utilized during collection of fluid from an organ or region of the body in isolation , where substantially all of the fluid flowing out of the organ or region is collected by the catheter 110 . the occluding means substantially prevents blood , therapeutic agent and / or other fluids entering the vessel from flowing on to other organs or regions , and permits collection of substantially all of the fluid entering the vessel . collected fluid may then be analyzed and / or re - oxygenated and / or perfused through the organ , discarded or handled otherwise . the occlusion means may include an occluding balloon , flange , disc or other means . catheter 110 is delivered to the vessel with the balloon 114 in a deflated condition . the expandable member is delivered , through the catheter , and deployed inside the vessel . the balloon is then inflated around the catheter and substantially all the fluid in the vessel flows through the catheter and into a perfusion set or reservoir to which it is connected . a pump , syringe or other means may be incorporated into the perfusion set to draw fluid out of the vessel , through the catheter , at a rate which substantially maintains the required flow through the organ or region , or through a re - perfusion circuit . as fluid is drawn out of the vessel through the catheter , the expanded support structure supports the vessel walls , preventing collapse or cavitation which might otherwise result from the low pressures or high flow rates generated at the catheter tip , maintaining patency and ensuring flow in the circuit . the expandable member may also anchor the device in position within the vessel , substantially precluding movement of the device and ensuring that the catheter is retained in an optimal location for collection of fluid . the expandable member may take a range of different shapes when in an expanded ( or collapsed ) configuration , and may provide any number of supporting filaments or struts . the design of the expandable member may be based on a range of criteria including but not limited to the size and strength of the vessel wall and the flow rates and pressures likely to be generated near the device . some of these embodiments are illustrated in fig8 a to 8 c although these are examples only and are not intended to limit the scope of the invention as broadly described herein . fig8 a to 8 c illustrate expandable members having elongate portions in the supporting struts adapted for contact with the vessel wall . in the example in fig8 b , the supporting struts are slightly rounded to reduce trauma to the vessel walls . fig8 c provides additional struts when compared with fig8 a , as may be necessitated in particularly flaccid vessels requiring more substantial support . embodiments illustrated herein provide expandable members with a substantially elongate structure adapted for coaxial insertion into and placement within the vessel . the elongate structure supports the vessel over a length on the elongate portions of the struts substantially parallel to and in contact with the vessel wall . these elongate portions may be substantially straight , or may be curved ( e . g . fig8 b ). supporting the vessel wall over a length of the support device , compared with the point of supports of the prior art , improves the capacity of the device to maintain patency , even when very low pressures and high flow rates are generated at the catheter tip , and also reduces the likelihood of the device causing damage to the vessel wall . the elongate portions may have a length which is about the same as or greater than the diameter of the vessel being supported , or some multiple of the vessel diameter , or for example from 1 mm up to 30 mm depending on the vessel size and structure . the length of the elongate portion may be selected according to the vessel being supported , the size of the catheter being used and the flow rates and pressures likely to be generated at the catheter tip . preferably , the elongate portions of the expandable member which contact the vessel wall , are just adjacent the distal tip of the catheter when the device is fully deployed . thus , a proximal end of one or more of the elongate portions may commence , for example , within 0 . 1 to 25 mm of the catheter tip , or at least at a distance which is less than the diameter of the catheter opening . this prevents the vessel wall from being drawn into the space between the catheter tip and the start of the elongate portion of the expandable member which contacts the vessel wall . further , the device may be configured so that when it is in an expanded condition , the distance between adjacent elongate portions is sufficiently small to prevent the vessel wall from being drawn into gaps between them . for example , the distance between adjacent elongate portions may be less than the diameter of the catheter . alternatively , the distance between the adjacent elongate portions may be less than , for example , 3 , 2 . 5 , 2 , 1 . 5 , 1 or 0 . 5 mm , depending on the size and type of the target vessel , and the diameter of the collection catheter being used . preferably , the support device possesses sufficient mechanical strength to maintain patency during collection of fluid , withstanding the deformation forces which may occur in response to suction or low pressures produced at the collection catheter tip . in some embodiments however , it may also be desirable for the device to exhibit some flexibility , and conform to the shape of the vessel when deployed . thus , the support device is capable of providing support and maintaining patency along a length of the vessel , even where there is a curve in the vessel wall . an alternative embodiment of a support device 200 is illustrated in fig2 . proximal end 205 of the expandable member 204 is fixedly attached to a stem or shaft 206 , whereas distal end 203 of the expandable member is movable and able to slide over part of the shaft . this enables the member to collapse radially for delivery inside a delivery catheter , and also facilitates recapture of the device . fig5 illustrates another alternative embodiment of a support device shown at 500 in an expanded condition . in this embodiment , both the proximal end 505 and the distal end 503 of the expandable member are movable along a stem or shaft 506 used to deliver the device to the vessel . stops 508 a , 508 b are provided at fixed locations on a distal portion of the shaft , arranged between ends 503 , 505 of the expandable member . these stops may consist of a small ring , crimp or node of increased diameter , relative to the shaft diameter , and prevent the ends of the expandable member from moving across the stop . this facilitates deployment and retrieval of the expandable member from a catheter . fig1 illustrates a support device 151 consisting of an expandable framework 155 having a woven or braided , basket - like configuration when in the expanded condition . in this arrangement , the support device may also include occluding means in the form of a thin flow - proof coating 156 on the inner and / or outer surface of framework 155 to prevent flow of liquid from the vessel . thus , substantially all fluid in the vessel may be collected by catheter 160 . the flow - proof coating may be made from biocompatible silicon , elastomer or flow - proof polymer . preferably , the support device includes a radiopaque or other marker so that it can be positioned within the target vessel using an imaging system such as those generally known in the art . this enables the physician to position and deploy the expandable member into the blood vessel accurately . the marker may be incorporated into the expandable member and / or into an atraumatic guiding tip which may be incorporated into the support device . preferably , the atraumatic tip is manufactured from , includes or is coated with a lubricant and / or a material having a low coefficient of friction . many materials having low coefficient of friction properties may be used including but not limited to biocompatible high density polyethylene ( hdpe ), teflon \u00ae, polypropylene , polyethylene , microglide \u2122, low friction chromium and silicon to name a few . this improves the performance of the atraumatic tip , so that it \u201c slides \u201d along the vessel wall upon making contact , thereby substantially avoiding trauma . use of an atraumatic guiding tip improves the safety and ability to position the expandable member in the target vessel . moreover , since the atraumatic tip may exhibit greater flexibility than the rest of the device , the device is easier to manipulate into position . the atraumatic tip may be provided at a distance from the distal end of the expandable member which enables a physician to guide the expandable member into position within the target vessel . this distance may be anywhere from , for example , 0 . 25 to 5 centimeters from the distal end of the expandable member when in an expanded condition , although it is to be understood that larger or smaller distances may be utilized , depending on the location of the target vessel and the anatomy surrounding it . referring now to fig6 a and 6 b , another example of a support device 600 is shown . a lumen 602 has a control stem 601 extending therein . four loop portions 603 are provided . each loop portion is attached at a first loop end to a distal end 604 of the lumen , and at a second loop end to the control stem at 605 . the loop portions are controllably expandable by advancing the control stem within the lumen in the direction shown by arrow 606 ( fig6 b ). the support device is percutaneously deliverable with the plurality of loop portions housed substantially within the lumen 602 as illustrated in fig6 a and expandable as illustrated in fig6 b . whilst the embodiment illustrated in fig6 a and 6 b provides 4 loop portions , it is to be understood that any number of loop portions may be used . the number of loop portions incorporated into the device may depend on , for example , the anatomy of the vessel being supported , and / or the size of the catheter used to deliver the device . fig7 a and 7 b illustrate another example of a support device 700 which provides 3 loop portions 703 attached to control stem 701 at juncture 705 . the 3 loop portions are contained during delivery substantially within lumen 702 ( fig7 a ), and are controllably expandable to maintain patency within the blood vessel by advancing control stem 701 in the direction of arrow 706 ( fig7 b ). the rounded edges of the loop portions present a reduced risk of damaging the vessel walls , e . g . by perforation or bruising during delivery . the one or more loop portions may be attached to or near the distal end of the delivery lumen in any suitable manner . the point of attachment may be inside or outside the lumen . the loops may be manufactured from any suitable material such as a metal , metal alloy , plastic , polymer , or other filamentous material or composite . the one or more loop portions may be attached at a second loop end to the control stem by soldering , fusing , an adhesive , or any other suitable means . in another embodiment , the loop portions may be attached to a first and a second loop end to the control stem . the support structure of fig6 a , 6 b , 7 a and 7 b may further include an atraumatic guiding tip of the kind described above to aid in positioning the support structure within the blood vessel . alternatively , parts of the loop portions which may protrude from the lumen when the loop portions are in their collapsed state may be used to guide the support structure into the blood vessel . one or more of the loop portions may be provided with a radiopaque or other marker to assist in this regard . retention means may also be provided with the support structure to retain the expandable member in an expanded condition within the vessel . the retention means may be in the form of a clamp , clip , thumb - slide or the like accessible from outside the patient &# 39 ; s body , and may facilitate adjustment of a deployed expandable member during a procedure . retention means may also impart additional rigidity and strength to the expandable member . thus , the retention member may be used to counteract excessively low pressures which may otherwise cause the expanded member to fail . a support structure of the kind illustrated in the figures may be delivered within a multilumen catheter 900 of the kind illustrated in cross section in fig9 . using this catheter , the support device 910 can be delivered through a first internal lumen 901 without interfering with flow in a second lumen 902 . a third lumen 903 may be provided for monitoring flow rates and pressures , for blood analysis or for delivering other percutaneous tools or devices to the vessel or as an inflation lumen for an occlusion balloon . it is to be understood that in the various embodiments of the present invention , the expanded member does not require constant contact with the vessel walls to provide the required support . for example , the diameter of the expanded member may be less than the diameter of the vessel so that the expanded member only contacts the vessel wall when the vessel begins to collapse . patency is considered to be maintained as long as the support device keeps the vessel open to a degree which is sufficient to maintain continuous flow . to avoid causing turbulence or other undesirable blood flow effects within the vessel , and to optimize flow in the vessel it may be desirable to substantially match the diameter of the expanded member to the diameter of the vessel . alternatively the expandable member may be shaped , e . g . as a coil or helix , to have minimal effect on the flow in the vessel . in one embodiment , the expandable member may have a slightly larger expanded diameter than the relaxed vessel to create an anchoring effect . depending on the size of the outflow vessel from which blood and perfusate is collected from the target region , there may be a natural tendency for the collection catheter tip to move about and contact the vessel , thus increasing the risk of vessel collapse or invagination of the catheter tip into the vessel wall . this can cause pooling of fluid in the isolated target region and may cause serious and permanent damage to the organ or region of the patient being treated . use of a support structure in conjunction with the collection catheter to maintain patency of the outflow vessel , in accordance with embodiments of the present invention can minimize the risk of these complications eventuating . thus , a collection catheter associated with the expanding member can be retained in position during fluid collection . this minimizes movement of the catheter tip , ensures that it is substantially centered relative to the vessel walls and improves withdrawal of fluid out of the vessel . at completion of the procedure , it is desirable that the expanded member is collapsed or compressed and recaptured , preferably in the catheter from which it was deployed . this facilitates removal of the support device from the patient . a reinforcing tip may be provided on the catheter end to strengthen it for recapture . alternatively or additionally , the tip may be coated with a lubricant and / or material having a low coefficient of friction to facilitate smooth recapture of the expandable member . the catheter may also have an internal coating of lubricant and / or a material having a low coefficient of friction to assist translation of support device along its interior during delivery and removal of the device from the patient . although the present invention has been described in terms of the presently preferred embodiments , it is to be understood that the disclosure is not to be interpreted as limiting . various alterations and modifications will no doubt become apparent to those skilled in the art after having read the above disclosure and it is intended that the present disclosure be interpreted as covering all alterations and modifications as fall within the true spirit and scope of the invention . effect of support device on flow rates and pressures achievable during recirculation in sheep right hepatic vein , cephalic vein , coronary sinus and renal vein during recirculation a 0 . 014 \u2033 diameter superelastic nitinol wire stem of 1 . 35 m length was used , coupled to an expandable member having 6 pre - shaped elliptical loop portions welded to the stem . a 0 . 024 \u2033 od atraumatic tip of 2 cm length attached to the distal end of the expandable member was used to position the device in the blood vessel . a balloon occlusion catheter was positioned in the vessel and the expandable member deployed at the tip of the catheter . the balloon was inflated to isolate and capture flows in the vessel and the catheter was connected to a standard extracorporeal circuit for blood circulation . negative pressures were observed in perfusion lines draining the coronary sinus , renal vein , right hepatic vein and cephalic vein during recirculation both with and without a support device . these data show that cavitation is prevented at certain pressures in the vessels tested where a support device is used , but is not prevented where the support device is absent in the vessel at those pressures . although cavitation may occur even with the support device , it occurs at higher flows . also , cavitation ceases sooner where the support device was employed allowing flow to return to normal . in the coronary sinus , recovery from cavitation was not possible without the support device , emphasizing the importance of the device in the procedure . the data further demonstrates that vessel collapse can be irreversible in the absence of a support structure . however , where a support structure is present , the vessel collapse may be reversed by increasing pressure in the vessel or by slowing or reversing the flow rate of fluid through the vessel . more specifically , considering the data for the right hepatic vein , flow rates of up to 250 ml per minute may be achieved before cavitation occurs where a support device is present in the vessel . under the same conditions but where there is no support device , flow rates of only up to 180 ml per minute are possible . a more striking example of the advantages of the support device is seen for the cephalic vein where no flow is achievable without the device . when the vessel wall is supported by the device flow rates of up to 200 ml per minute are noted before cavitation occurs . when the vessel wall is supported flow rates of up to 200 ml per minute are noted before cavitation occurs .", "category": "Electricity"}
|
Is the categorization of this patent accurate?
| 0.25 |
1d5133cde40c92eaabe8131cbdf960efd3029fddffc551de233b4ea69b05bd19
| 0.066406 | 0.000216 | 0.040283 | 0.002884 | 0.067383 | 0.000368 |
null |
{"category": "Human Necessities", "patent": "while the support device of the present invention may be used in a range of different vessels , including blood vessels , it has particular application in procedures where an organ or anatomical region is undergoing localized perfusion with a therapeutic , diagnostic or other agent . for simplicity , these agents will be hereinafter referred to as therapeutic agents . however , it is to be understood that the term \u201c therapeutic \u201d is not to be construed as limiting , and that it includes , without limitation , therapeutic , diagnostic , prophylactic and other agents not specifically identified herein , but which would be considered by the relevant skilled addressee to be suitable for perfusion to an organ or anatomical region . perfusion may be total perfusion , where the entire organ is totally or substantially isolated from the systemic flow , or partial perfusion where only a portion of the organ is substantially isolated . localized perfusion of this kind presents advantages by improving efficacy and the time exposure of the therapeutic agent to the relevant cells . it also limits exposure and hence toxicity to non - target cells as described in brief above . however , it is to be understood that the present invention may also be used simply to collect or drain fluid from an organ or region . collected fluid may be removed from the subject and re - circulated into the organ , filtered and / or treated , or discarded . in some organs , it may be difficult to achieve total isolation , so partial isolation and perfusion may be performed , for example to the right or left lobe of the liver . despite partial perfusion being capable of delivering therapeutic agent to merely a part of the organ , significant therapeutic benefit may still be achieved . particular benefit may be achieved where perfusate is collected after perfusing the target organ , so as to prevent subsequent circulation of the therapeutic agent to other regions of the body where toxic effects may be observed , or the therapeutic agent wasted . the benefit may be improved further where collected perfusate is re - circulated into the target organ utilizing any therapeutic agent which remains after a first pass through the target organ . this may be achieved using the approach described in published patent application wo2005 / 082440 , the entire contents of which are herein incorporated by reference . as discussed infra , when fluid is collected from vessels draining from a target organ or region , one or more of these vessels may require cannulation with a collection catheter . when fluid is drained through these collection catheters , the vessels in which they are positioned become susceptible to collapse as the pressure inside decreases . while some vessels may be more susceptible to collapse than others , the support device of the present invention can provide advantages by supporting and stabilizing the vessel and even anchoring the collection catheter in position . the support device of the present invention may facilitate or at least improve the performance of perfusion . in some instances , the advantages of the present invention have been found to be essential to maintaining adequate positioning of collection catheters and flow rates within the vessel during perfusion . the right and left lobes of the liver have been identified as possible target regions and in this context , the support device may be deployed in one of the hepatic veins to support and maintain patency of the vein as fluid ( e . g . perfusate ) is collected from the liver . however , it is to be understood that fluid from many other organs or regions may be accessed in this way . deploying the support device may also protect the vessel wall by maintaining the tip of the catheter substantially centrally of the vessel or at least at a distance from the vessel walls to prevent aspiration or cavitation . deployment of the device may refer to partial or complete deployment . in complete deployment , the entire expandable member is released from the catheter and expanded to its full extent . in partial deployment , part of the expandable member is retained within the catheter and the amount of expansion is limited by the diameter of the catheter opening . partial deployment may be useful where , for example , during deployment it is found that the diameter of the expandable member may exceed the vessel diameter by an unsafe amount and complete deployment is likely to damage the vessel wall . limiting expansion of the device by partial deployment may avoid vessel damage . partial deployment may also stabilize the expandable member by limiting its movement relative to the catheter tip . thus by retaining part of the expandable member within the catheter , torsional , axial and lateral movement of the member , relative to the catheter is prevented or at least minimized by the struts of the expandable member being in abutment with the internal surface of the catheter . alternatively , the expandable member may be modified at the proximal end , for example by incorporating a lead , a link or other means to limit the extent of movement possible between the catheter tip and the expandable member once deployed . as a further positioning aid , markings may be provided at the proximal end of the control stem / shaft , outside the patient &# 39 ; s body . as the device is released into the vessel , the markings may be utilized to indicate the distance of device deployment , past the catheter tip . during collection of fluid from the vessel , low pressures may develop at the collection device tip , particularly where a roller / peristaltic pump or the like is used to draw fluid from the target organ out of the vessel . this may be indicated by pressures in a lumen feeding into the pump as low as , for example , \u2212 190 mmhg , although clearly these pressures are variable depending on the vessel type , health and age of the subject , characteristics of the perfusion circuit and the like . in the absence of the inventive support device , these pressures can cause the vessel to collapse . not only would vessel collapse affect the perfusion procedure , vessel collapse can also cause venous pooling in the organ and irreversible tissue damage . the advantages and benefits of the present invention will be expanded upon in the following detailed description presenting some of the preferred embodiments of the invention , and the specific examples which follow . it is to be understood that the embodiments and examples provided herein are intended to indicate how the present invention may be performed and are not intended to be limiting on the scope of protection sought as is defined in the claims appended hereto . fig1 a shows an example of an expandable member , in its expanded condition , suitable for supporting a vessel . expandable member 104 is provided in the form of an expandable framework and is adapted to be percutaneously deliverable to the blood vessel in a collapsed condition . fig1 b shows the expandable member in a collapsed condition within a catheter 110 , in which ends 105 , 107 have been drawn apart to radially reduce the member . when collapsed within catheter 110 , atraumatic tip 101 may protrude from the catheter to assist in guiding the support device into the vessel prior to deployment . when the expandable member has been guided into the target blood vessel , the catheter 110 is retracted ( or the expandable member is pushed out of the catheter ), deploying the device into the vessel where it expands . fig1 c shows the support device fully deployed from the catheter , with the expandable member in its fully expanded condition . a guidewire or stem 106 extends within the catheter 110 and is used to deliver the device from a point of entry through the peripheral vasculature to the target vessel . atraumatic tip 101 coupled to the expandable member 104 , is adapted to make atraumatic contact with vessel walls during placement of the device by deforming or deflecting off the vessel wall on contact . this can be achieved by incorporating flexibility into the tip so that it deforms upon contact with the vessel wall . alternatively or additionally , the tip may be shaped or curved to avoid trauma . the atraumatic tip may take any one of a number of forms . in the examples illustrated in fig1 to 3 , the atraumatic tip 101 , 201 is j - shaped . however , other shapes are considered to be suitable , including but not limited to those illustrated in fig4 . for example , the atraumatic tip may have a cross section which is enlarged relative to the guidewire radius , and have a smooth surface so as to avoid causing perforation when the tip comes into contact with the vessel wall . one such example is shown in fig4 a where the atraumatic tip 401 is tear - shaped . alternatively , the atraumatic tip may include a portion having a pigtail shaped curve 402 ( fig4 b ), or an angled tip ( not shown ). preferably , the expandable member is formed from a biocompatible superelastic material , or alternatively from a shape memory material or a material which exhibits both of these properties , being capable of recovery after deformation for delivery in a collapsed or compressed state within a catheter . devices manufactured using these materials can be collapsed for percutaneous delivery to a deployment site and then resume a known shape on deployment . a range of biocompatible materials may be suitable such as alloys of nickel and titanium ( e . g . nitinol ). other suitable biocompatible materials include but are not limited to polymers and plastics such as hydrophilic plastics , ceramics and the like . fig3 illustrates the support device of fig1 a to 1 c , with an occluding balloon inflated around catheter 110 . the occluding balloon 114 may be utilized during collection of fluid from an organ or region of the body in isolation , where substantially all of the fluid flowing out of the organ or region is collected by the catheter 110 . the occluding means substantially prevents blood , therapeutic agent and / or other fluids entering the vessel from flowing on to other organs or regions , and permits collection of substantially all of the fluid entering the vessel . collected fluid may then be analyzed and / or re - oxygenated and / or perfused through the organ , discarded or handled otherwise . the occlusion means may include an occluding balloon , flange , disc or other means . catheter 110 is delivered to the vessel with the balloon 114 in a deflated condition . the expandable member is delivered , through the catheter , and deployed inside the vessel . the balloon is then inflated around the catheter and substantially all the fluid in the vessel flows through the catheter and into a perfusion set or reservoir to which it is connected . a pump , syringe or other means may be incorporated into the perfusion set to draw fluid out of the vessel , through the catheter , at a rate which substantially maintains the required flow through the organ or region , or through a re - perfusion circuit . as fluid is drawn out of the vessel through the catheter , the expanded support structure supports the vessel walls , preventing collapse or cavitation which might otherwise result from the low pressures or high flow rates generated at the catheter tip , maintaining patency and ensuring flow in the circuit . the expandable member may also anchor the device in position within the vessel , substantially precluding movement of the device and ensuring that the catheter is retained in an optimal location for collection of fluid . the expandable member may take a range of different shapes when in an expanded ( or collapsed ) configuration , and may provide any number of supporting filaments or struts . the design of the expandable member may be based on a range of criteria including but not limited to the size and strength of the vessel wall and the flow rates and pressures likely to be generated near the device . some of these embodiments are illustrated in fig8 a to 8 c although these are examples only and are not intended to limit the scope of the invention as broadly described herein . fig8 a to 8 c illustrate expandable members having elongate portions in the supporting struts adapted for contact with the vessel wall . in the example in fig8 b , the supporting struts are slightly rounded to reduce trauma to the vessel walls . fig8 c provides additional struts when compared with fig8 a , as may be necessitated in particularly flaccid vessels requiring more substantial support . embodiments illustrated herein provide expandable members with a substantially elongate structure adapted for coaxial insertion into and placement within the vessel . the elongate structure supports the vessel over a length on the elongate portions of the struts substantially parallel to and in contact with the vessel wall . these elongate portions may be substantially straight , or may be curved ( e . g . fig8 b ). supporting the vessel wall over a length of the support device , compared with the point of supports of the prior art , improves the capacity of the device to maintain patency , even when very low pressures and high flow rates are generated at the catheter tip , and also reduces the likelihood of the device causing damage to the vessel wall . the elongate portions may have a length which is about the same as or greater than the diameter of the vessel being supported , or some multiple of the vessel diameter , or for example from 1 mm up to 30 mm depending on the vessel size and structure . the length of the elongate portion may be selected according to the vessel being supported , the size of the catheter being used and the flow rates and pressures likely to be generated at the catheter tip . preferably , the elongate portions of the expandable member which contact the vessel wall , are just adjacent the distal tip of the catheter when the device is fully deployed . thus , a proximal end of one or more of the elongate portions may commence , for example , within 0 . 1 to 25 mm of the catheter tip , or at least at a distance which is less than the diameter of the catheter opening . this prevents the vessel wall from being drawn into the space between the catheter tip and the start of the elongate portion of the expandable member which contacts the vessel wall . further , the device may be configured so that when it is in an expanded condition , the distance between adjacent elongate portions is sufficiently small to prevent the vessel wall from being drawn into gaps between them . for example , the distance between adjacent elongate portions may be less than the diameter of the catheter . alternatively , the distance between the adjacent elongate portions may be less than , for example , 3 , 2 . 5 , 2 , 1 . 5 , 1 or 0 . 5 mm , depending on the size and type of the target vessel , and the diameter of the collection catheter being used . preferably , the support device possesses sufficient mechanical strength to maintain patency during collection of fluid , withstanding the deformation forces which may occur in response to suction or low pressures produced at the collection catheter tip . in some embodiments however , it may also be desirable for the device to exhibit some flexibility , and conform to the shape of the vessel when deployed . thus , the support device is capable of providing support and maintaining patency along a length of the vessel , even where there is a curve in the vessel wall . an alternative embodiment of a support device 200 is illustrated in fig2 . proximal end 205 of the expandable member 204 is fixedly attached to a stem or shaft 206 , whereas distal end 203 of the expandable member is movable and able to slide over part of the shaft . this enables the member to collapse radially for delivery inside a delivery catheter , and also facilitates recapture of the device . fig5 illustrates another alternative embodiment of a support device shown at 500 in an expanded condition . in this embodiment , both the proximal end 505 and the distal end 503 of the expandable member are movable along a stem or shaft 506 used to deliver the device to the vessel . stops 508 a , 508 b are provided at fixed locations on a distal portion of the shaft , arranged between ends 503 , 505 of the expandable member . these stops may consist of a small ring , crimp or node of increased diameter , relative to the shaft diameter , and prevent the ends of the expandable member from moving across the stop . this facilitates deployment and retrieval of the expandable member from a catheter . fig1 illustrates a support device 151 consisting of an expandable framework 155 having a woven or braided , basket - like configuration when in the expanded condition . in this arrangement , the support device may also include occluding means in the form of a thin flow - proof coating 156 on the inner and / or outer surface of framework 155 to prevent flow of liquid from the vessel . thus , substantially all fluid in the vessel may be collected by catheter 160 . the flow - proof coating may be made from biocompatible silicon , elastomer or flow - proof polymer . preferably , the support device includes a radiopaque or other marker so that it can be positioned within the target vessel using an imaging system such as those generally known in the art . this enables the physician to position and deploy the expandable member into the blood vessel accurately . the marker may be incorporated into the expandable member and / or into an atraumatic guiding tip which may be incorporated into the support device . preferably , the atraumatic tip is manufactured from , includes or is coated with a lubricant and / or a material having a low coefficient of friction . many materials having low coefficient of friction properties may be used including but not limited to biocompatible high density polyethylene ( hdpe ), teflon \u00ae, polypropylene , polyethylene , microglide \u2122, low friction chromium and silicon to name a few . this improves the performance of the atraumatic tip , so that it \u201c slides \u201d along the vessel wall upon making contact , thereby substantially avoiding trauma . use of an atraumatic guiding tip improves the safety and ability to position the expandable member in the target vessel . moreover , since the atraumatic tip may exhibit greater flexibility than the rest of the device , the device is easier to manipulate into position . the atraumatic tip may be provided at a distance from the distal end of the expandable member which enables a physician to guide the expandable member into position within the target vessel . this distance may be anywhere from , for example , 0 . 25 to 5 centimeters from the distal end of the expandable member when in an expanded condition , although it is to be understood that larger or smaller distances may be utilized , depending on the location of the target vessel and the anatomy surrounding it . referring now to fig6 a and 6 b , another example of a support device 600 is shown . a lumen 602 has a control stem 601 extending therein . four loop portions 603 are provided . each loop portion is attached at a first loop end to a distal end 604 of the lumen , and at a second loop end to the control stem at 605 . the loop portions are controllably expandable by advancing the control stem within the lumen in the direction shown by arrow 606 ( fig6 b ). the support device is percutaneously deliverable with the plurality of loop portions housed substantially within the lumen 602 as illustrated in fig6 a and expandable as illustrated in fig6 b . whilst the embodiment illustrated in fig6 a and 6 b provides 4 loop portions , it is to be understood that any number of loop portions may be used . the number of loop portions incorporated into the device may depend on , for example , the anatomy of the vessel being supported , and / or the size of the catheter used to deliver the device . fig7 a and 7 b illustrate another example of a support device 700 which provides 3 loop portions 703 attached to control stem 701 at juncture 705 . the 3 loop portions are contained during delivery substantially within lumen 702 ( fig7 a ), and are controllably expandable to maintain patency within the blood vessel by advancing control stem 701 in the direction of arrow 706 ( fig7 b ). the rounded edges of the loop portions present a reduced risk of damaging the vessel walls , e . g . by perforation or bruising during delivery . the one or more loop portions may be attached to or near the distal end of the delivery lumen in any suitable manner . the point of attachment may be inside or outside the lumen . the loops may be manufactured from any suitable material such as a metal , metal alloy , plastic , polymer , or other filamentous material or composite . the one or more loop portions may be attached at a second loop end to the control stem by soldering , fusing , an adhesive , or any other suitable means . in another embodiment , the loop portions may be attached to a first and a second loop end to the control stem . the support structure of fig6 a , 6 b , 7 a and 7 b may further include an atraumatic guiding tip of the kind described above to aid in positioning the support structure within the blood vessel . alternatively , parts of the loop portions which may protrude from the lumen when the loop portions are in their collapsed state may be used to guide the support structure into the blood vessel . one or more of the loop portions may be provided with a radiopaque or other marker to assist in this regard . retention means may also be provided with the support structure to retain the expandable member in an expanded condition within the vessel . the retention means may be in the form of a clamp , clip , thumb - slide or the like accessible from outside the patient &# 39 ; s body , and may facilitate adjustment of a deployed expandable member during a procedure . retention means may also impart additional rigidity and strength to the expandable member . thus , the retention member may be used to counteract excessively low pressures which may otherwise cause the expanded member to fail . a support structure of the kind illustrated in the figures may be delivered within a multilumen catheter 900 of the kind illustrated in cross section in fig9 . using this catheter , the support device 910 can be delivered through a first internal lumen 901 without interfering with flow in a second lumen 902 . a third lumen 903 may be provided for monitoring flow rates and pressures , for blood analysis or for delivering other percutaneous tools or devices to the vessel or as an inflation lumen for an occlusion balloon . it is to be understood that in the various embodiments of the present invention , the expanded member does not require constant contact with the vessel walls to provide the required support . for example , the diameter of the expanded member may be less than the diameter of the vessel so that the expanded member only contacts the vessel wall when the vessel begins to collapse . patency is considered to be maintained as long as the support device keeps the vessel open to a degree which is sufficient to maintain continuous flow . to avoid causing turbulence or other undesirable blood flow effects within the vessel , and to optimize flow in the vessel it may be desirable to substantially match the diameter of the expanded member to the diameter of the vessel . alternatively the expandable member may be shaped , e . g . as a coil or helix , to have minimal effect on the flow in the vessel . in one embodiment , the expandable member may have a slightly larger expanded diameter than the relaxed vessel to create an anchoring effect . depending on the size of the outflow vessel from which blood and perfusate is collected from the target region , there may be a natural tendency for the collection catheter tip to move about and contact the vessel , thus increasing the risk of vessel collapse or invagination of the catheter tip into the vessel wall . this can cause pooling of fluid in the isolated target region and may cause serious and permanent damage to the organ or region of the patient being treated . use of a support structure in conjunction with the collection catheter to maintain patency of the outflow vessel , in accordance with embodiments of the present invention can minimize the risk of these complications eventuating . thus , a collection catheter associated with the expanding member can be retained in position during fluid collection . this minimizes movement of the catheter tip , ensures that it is substantially centered relative to the vessel walls and improves withdrawal of fluid out of the vessel . at completion of the procedure , it is desirable that the expanded member is collapsed or compressed and recaptured , preferably in the catheter from which it was deployed . this facilitates removal of the support device from the patient . a reinforcing tip may be provided on the catheter end to strengthen it for recapture . alternatively or additionally , the tip may be coated with a lubricant and / or material having a low coefficient of friction to facilitate smooth recapture of the expandable member . the catheter may also have an internal coating of lubricant and / or a material having a low coefficient of friction to assist translation of support device along its interior during delivery and removal of the device from the patient . although the present invention has been described in terms of the presently preferred embodiments , it is to be understood that the disclosure is not to be interpreted as limiting . various alterations and modifications will no doubt become apparent to those skilled in the art after having read the above disclosure and it is intended that the present disclosure be interpreted as covering all alterations and modifications as fall within the true spirit and scope of the invention . effect of support device on flow rates and pressures achievable during recirculation in sheep right hepatic vein , cephalic vein , coronary sinus and renal vein during recirculation a 0 . 014 \u2033 diameter superelastic nitinol wire stem of 1 . 35 m length was used , coupled to an expandable member having 6 pre - shaped elliptical loop portions welded to the stem . a 0 . 024 \u2033 od atraumatic tip of 2 cm length attached to the distal end of the expandable member was used to position the device in the blood vessel . a balloon occlusion catheter was positioned in the vessel and the expandable member deployed at the tip of the catheter . the balloon was inflated to isolate and capture flows in the vessel and the catheter was connected to a standard extracorporeal circuit for blood circulation . negative pressures were observed in perfusion lines draining the coronary sinus , renal vein , right hepatic vein and cephalic vein during recirculation both with and without a support device . these data show that cavitation is prevented at certain pressures in the vessels tested where a support device is used , but is not prevented where the support device is absent in the vessel at those pressures . although cavitation may occur even with the support device , it occurs at higher flows . also , cavitation ceases sooner where the support device was employed allowing flow to return to normal . in the coronary sinus , recovery from cavitation was not possible without the support device , emphasizing the importance of the device in the procedure . the data further demonstrates that vessel collapse can be irreversible in the absence of a support structure . however , where a support structure is present , the vessel collapse may be reversed by increasing pressure in the vessel or by slowing or reversing the flow rate of fluid through the vessel . more specifically , considering the data for the right hepatic vein , flow rates of up to 250 ml per minute may be achieved before cavitation occurs where a support device is present in the vessel . under the same conditions but where there is no support device , flow rates of only up to 180 ml per minute are possible . a more striking example of the advantages of the support device is seen for the cephalic vein where no flow is achievable without the device . when the vessel wall is supported by the device flow rates of up to 200 ml per minute are noted before cavitation occurs . when the vessel wall is supported flow rates of up to 200 ml per minute are noted before cavitation occurs ."}
|
{"patent": "while the support device of the present invention may be used in a range of different vessels , including blood vessels , it has particular application in procedures where an organ or anatomical region is undergoing localized perfusion with a therapeutic , diagnostic or other agent . for simplicity , these agents will be hereinafter referred to as therapeutic agents . however , it is to be understood that the term \u201c therapeutic \u201d is not to be construed as limiting , and that it includes , without limitation , therapeutic , diagnostic , prophylactic and other agents not specifically identified herein , but which would be considered by the relevant skilled addressee to be suitable for perfusion to an organ or anatomical region . perfusion may be total perfusion , where the entire organ is totally or substantially isolated from the systemic flow , or partial perfusion where only a portion of the organ is substantially isolated . localized perfusion of this kind presents advantages by improving efficacy and the time exposure of the therapeutic agent to the relevant cells . it also limits exposure and hence toxicity to non - target cells as described in brief above . however , it is to be understood that the present invention may also be used simply to collect or drain fluid from an organ or region . collected fluid may be removed from the subject and re - circulated into the organ , filtered and / or treated , or discarded . in some organs , it may be difficult to achieve total isolation , so partial isolation and perfusion may be performed , for example to the right or left lobe of the liver . despite partial perfusion being capable of delivering therapeutic agent to merely a part of the organ , significant therapeutic benefit may still be achieved . particular benefit may be achieved where perfusate is collected after perfusing the target organ , so as to prevent subsequent circulation of the therapeutic agent to other regions of the body where toxic effects may be observed , or the therapeutic agent wasted . the benefit may be improved further where collected perfusate is re - circulated into the target organ utilizing any therapeutic agent which remains after a first pass through the target organ . this may be achieved using the approach described in published patent application wo2005 / 082440 , the entire contents of which are herein incorporated by reference . as discussed infra , when fluid is collected from vessels draining from a target organ or region , one or more of these vessels may require cannulation with a collection catheter . when fluid is drained through these collection catheters , the vessels in which they are positioned become susceptible to collapse as the pressure inside decreases . while some vessels may be more susceptible to collapse than others , the support device of the present invention can provide advantages by supporting and stabilizing the vessel and even anchoring the collection catheter in position . the support device of the present invention may facilitate or at least improve the performance of perfusion . in some instances , the advantages of the present invention have been found to be essential to maintaining adequate positioning of collection catheters and flow rates within the vessel during perfusion . the right and left lobes of the liver have been identified as possible target regions and in this context , the support device may be deployed in one of the hepatic veins to support and maintain patency of the vein as fluid ( e . g . perfusate ) is collected from the liver . however , it is to be understood that fluid from many other organs or regions may be accessed in this way . deploying the support device may also protect the vessel wall by maintaining the tip of the catheter substantially centrally of the vessel or at least at a distance from the vessel walls to prevent aspiration or cavitation . deployment of the device may refer to partial or complete deployment . in complete deployment , the entire expandable member is released from the catheter and expanded to its full extent . in partial deployment , part of the expandable member is retained within the catheter and the amount of expansion is limited by the diameter of the catheter opening . partial deployment may be useful where , for example , during deployment it is found that the diameter of the expandable member may exceed the vessel diameter by an unsafe amount and complete deployment is likely to damage the vessel wall . limiting expansion of the device by partial deployment may avoid vessel damage . partial deployment may also stabilize the expandable member by limiting its movement relative to the catheter tip . thus by retaining part of the expandable member within the catheter , torsional , axial and lateral movement of the member , relative to the catheter is prevented or at least minimized by the struts of the expandable member being in abutment with the internal surface of the catheter . alternatively , the expandable member may be modified at the proximal end , for example by incorporating a lead , a link or other means to limit the extent of movement possible between the catheter tip and the expandable member once deployed . as a further positioning aid , markings may be provided at the proximal end of the control stem / shaft , outside the patient &# 39 ; s body . as the device is released into the vessel , the markings may be utilized to indicate the distance of device deployment , past the catheter tip . during collection of fluid from the vessel , low pressures may develop at the collection device tip , particularly where a roller / peristaltic pump or the like is used to draw fluid from the target organ out of the vessel . this may be indicated by pressures in a lumen feeding into the pump as low as , for example , \u2212 190 mmhg , although clearly these pressures are variable depending on the vessel type , health and age of the subject , characteristics of the perfusion circuit and the like . in the absence of the inventive support device , these pressures can cause the vessel to collapse . not only would vessel collapse affect the perfusion procedure , vessel collapse can also cause venous pooling in the organ and irreversible tissue damage . the advantages and benefits of the present invention will be expanded upon in the following detailed description presenting some of the preferred embodiments of the invention , and the specific examples which follow . it is to be understood that the embodiments and examples provided herein are intended to indicate how the present invention may be performed and are not intended to be limiting on the scope of protection sought as is defined in the claims appended hereto . fig1 a shows an example of an expandable member , in its expanded condition , suitable for supporting a vessel . expandable member 104 is provided in the form of an expandable framework and is adapted to be percutaneously deliverable to the blood vessel in a collapsed condition . fig1 b shows the expandable member in a collapsed condition within a catheter 110 , in which ends 105 , 107 have been drawn apart to radially reduce the member . when collapsed within catheter 110 , atraumatic tip 101 may protrude from the catheter to assist in guiding the support device into the vessel prior to deployment . when the expandable member has been guided into the target blood vessel , the catheter 110 is retracted ( or the expandable member is pushed out of the catheter ), deploying the device into the vessel where it expands . fig1 c shows the support device fully deployed from the catheter , with the expandable member in its fully expanded condition . a guidewire or stem 106 extends within the catheter 110 and is used to deliver the device from a point of entry through the peripheral vasculature to the target vessel . atraumatic tip 101 coupled to the expandable member 104 , is adapted to make atraumatic contact with vessel walls during placement of the device by deforming or deflecting off the vessel wall on contact . this can be achieved by incorporating flexibility into the tip so that it deforms upon contact with the vessel wall . alternatively or additionally , the tip may be shaped or curved to avoid trauma . the atraumatic tip may take any one of a number of forms . in the examples illustrated in fig1 to 3 , the atraumatic tip 101 , 201 is j - shaped . however , other shapes are considered to be suitable , including but not limited to those illustrated in fig4 . for example , the atraumatic tip may have a cross section which is enlarged relative to the guidewire radius , and have a smooth surface so as to avoid causing perforation when the tip comes into contact with the vessel wall . one such example is shown in fig4 a where the atraumatic tip 401 is tear - shaped . alternatively , the atraumatic tip may include a portion having a pigtail shaped curve 402 ( fig4 b ), or an angled tip ( not shown ). preferably , the expandable member is formed from a biocompatible superelastic material , or alternatively from a shape memory material or a material which exhibits both of these properties , being capable of recovery after deformation for delivery in a collapsed or compressed state within a catheter . devices manufactured using these materials can be collapsed for percutaneous delivery to a deployment site and then resume a known shape on deployment . a range of biocompatible materials may be suitable such as alloys of nickel and titanium ( e . g . nitinol ). other suitable biocompatible materials include but are not limited to polymers and plastics such as hydrophilic plastics , ceramics and the like . fig3 illustrates the support device of fig1 a to 1 c , with an occluding balloon inflated around catheter 110 . the occluding balloon 114 may be utilized during collection of fluid from an organ or region of the body in isolation , where substantially all of the fluid flowing out of the organ or region is collected by the catheter 110 . the occluding means substantially prevents blood , therapeutic agent and / or other fluids entering the vessel from flowing on to other organs or regions , and permits collection of substantially all of the fluid entering the vessel . collected fluid may then be analyzed and / or re - oxygenated and / or perfused through the organ , discarded or handled otherwise . the occlusion means may include an occluding balloon , flange , disc or other means . catheter 110 is delivered to the vessel with the balloon 114 in a deflated condition . the expandable member is delivered , through the catheter , and deployed inside the vessel . the balloon is then inflated around the catheter and substantially all the fluid in the vessel flows through the catheter and into a perfusion set or reservoir to which it is connected . a pump , syringe or other means may be incorporated into the perfusion set to draw fluid out of the vessel , through the catheter , at a rate which substantially maintains the required flow through the organ or region , or through a re - perfusion circuit . as fluid is drawn out of the vessel through the catheter , the expanded support structure supports the vessel walls , preventing collapse or cavitation which might otherwise result from the low pressures or high flow rates generated at the catheter tip , maintaining patency and ensuring flow in the circuit . the expandable member may also anchor the device in position within the vessel , substantially precluding movement of the device and ensuring that the catheter is retained in an optimal location for collection of fluid . the expandable member may take a range of different shapes when in an expanded ( or collapsed ) configuration , and may provide any number of supporting filaments or struts . the design of the expandable member may be based on a range of criteria including but not limited to the size and strength of the vessel wall and the flow rates and pressures likely to be generated near the device . some of these embodiments are illustrated in fig8 a to 8 c although these are examples only and are not intended to limit the scope of the invention as broadly described herein . fig8 a to 8 c illustrate expandable members having elongate portions in the supporting struts adapted for contact with the vessel wall . in the example in fig8 b , the supporting struts are slightly rounded to reduce trauma to the vessel walls . fig8 c provides additional struts when compared with fig8 a , as may be necessitated in particularly flaccid vessels requiring more substantial support . embodiments illustrated herein provide expandable members with a substantially elongate structure adapted for coaxial insertion into and placement within the vessel . the elongate structure supports the vessel over a length on the elongate portions of the struts substantially parallel to and in contact with the vessel wall . these elongate portions may be substantially straight , or may be curved ( e . g . fig8 b ). supporting the vessel wall over a length of the support device , compared with the point of supports of the prior art , improves the capacity of the device to maintain patency , even when very low pressures and high flow rates are generated at the catheter tip , and also reduces the likelihood of the device causing damage to the vessel wall . the elongate portions may have a length which is about the same as or greater than the diameter of the vessel being supported , or some multiple of the vessel diameter , or for example from 1 mm up to 30 mm depending on the vessel size and structure . the length of the elongate portion may be selected according to the vessel being supported , the size of the catheter being used and the flow rates and pressures likely to be generated at the catheter tip . preferably , the elongate portions of the expandable member which contact the vessel wall , are just adjacent the distal tip of the catheter when the device is fully deployed . thus , a proximal end of one or more of the elongate portions may commence , for example , within 0 . 1 to 25 mm of the catheter tip , or at least at a distance which is less than the diameter of the catheter opening . this prevents the vessel wall from being drawn into the space between the catheter tip and the start of the elongate portion of the expandable member which contacts the vessel wall . further , the device may be configured so that when it is in an expanded condition , the distance between adjacent elongate portions is sufficiently small to prevent the vessel wall from being drawn into gaps between them . for example , the distance between adjacent elongate portions may be less than the diameter of the catheter . alternatively , the distance between the adjacent elongate portions may be less than , for example , 3 , 2 . 5 , 2 , 1 . 5 , 1 or 0 . 5 mm , depending on the size and type of the target vessel , and the diameter of the collection catheter being used . preferably , the support device possesses sufficient mechanical strength to maintain patency during collection of fluid , withstanding the deformation forces which may occur in response to suction or low pressures produced at the collection catheter tip . in some embodiments however , it may also be desirable for the device to exhibit some flexibility , and conform to the shape of the vessel when deployed . thus , the support device is capable of providing support and maintaining patency along a length of the vessel , even where there is a curve in the vessel wall . an alternative embodiment of a support device 200 is illustrated in fig2 . proximal end 205 of the expandable member 204 is fixedly attached to a stem or shaft 206 , whereas distal end 203 of the expandable member is movable and able to slide over part of the shaft . this enables the member to collapse radially for delivery inside a delivery catheter , and also facilitates recapture of the device . fig5 illustrates another alternative embodiment of a support device shown at 500 in an expanded condition . in this embodiment , both the proximal end 505 and the distal end 503 of the expandable member are movable along a stem or shaft 506 used to deliver the device to the vessel . stops 508 a , 508 b are provided at fixed locations on a distal portion of the shaft , arranged between ends 503 , 505 of the expandable member . these stops may consist of a small ring , crimp or node of increased diameter , relative to the shaft diameter , and prevent the ends of the expandable member from moving across the stop . this facilitates deployment and retrieval of the expandable member from a catheter . fig1 illustrates a support device 151 consisting of an expandable framework 155 having a woven or braided , basket - like configuration when in the expanded condition . in this arrangement , the support device may also include occluding means in the form of a thin flow - proof coating 156 on the inner and / or outer surface of framework 155 to prevent flow of liquid from the vessel . thus , substantially all fluid in the vessel may be collected by catheter 160 . the flow - proof coating may be made from biocompatible silicon , elastomer or flow - proof polymer . preferably , the support device includes a radiopaque or other marker so that it can be positioned within the target vessel using an imaging system such as those generally known in the art . this enables the physician to position and deploy the expandable member into the blood vessel accurately . the marker may be incorporated into the expandable member and / or into an atraumatic guiding tip which may be incorporated into the support device . preferably , the atraumatic tip is manufactured from , includes or is coated with a lubricant and / or a material having a low coefficient of friction . many materials having low coefficient of friction properties may be used including but not limited to biocompatible high density polyethylene ( hdpe ), teflon \u00ae, polypropylene , polyethylene , microglide \u2122, low friction chromium and silicon to name a few . this improves the performance of the atraumatic tip , so that it \u201c slides \u201d along the vessel wall upon making contact , thereby substantially avoiding trauma . use of an atraumatic guiding tip improves the safety and ability to position the expandable member in the target vessel . moreover , since the atraumatic tip may exhibit greater flexibility than the rest of the device , the device is easier to manipulate into position . the atraumatic tip may be provided at a distance from the distal end of the expandable member which enables a physician to guide the expandable member into position within the target vessel . this distance may be anywhere from , for example , 0 . 25 to 5 centimeters from the distal end of the expandable member when in an expanded condition , although it is to be understood that larger or smaller distances may be utilized , depending on the location of the target vessel and the anatomy surrounding it . referring now to fig6 a and 6 b , another example of a support device 600 is shown . a lumen 602 has a control stem 601 extending therein . four loop portions 603 are provided . each loop portion is attached at a first loop end to a distal end 604 of the lumen , and at a second loop end to the control stem at 605 . the loop portions are controllably expandable by advancing the control stem within the lumen in the direction shown by arrow 606 ( fig6 b ). the support device is percutaneously deliverable with the plurality of loop portions housed substantially within the lumen 602 as illustrated in fig6 a and expandable as illustrated in fig6 b . whilst the embodiment illustrated in fig6 a and 6 b provides 4 loop portions , it is to be understood that any number of loop portions may be used . the number of loop portions incorporated into the device may depend on , for example , the anatomy of the vessel being supported , and / or the size of the catheter used to deliver the device . fig7 a and 7 b illustrate another example of a support device 700 which provides 3 loop portions 703 attached to control stem 701 at juncture 705 . the 3 loop portions are contained during delivery substantially within lumen 702 ( fig7 a ), and are controllably expandable to maintain patency within the blood vessel by advancing control stem 701 in the direction of arrow 706 ( fig7 b ). the rounded edges of the loop portions present a reduced risk of damaging the vessel walls , e . g . by perforation or bruising during delivery . the one or more loop portions may be attached to or near the distal end of the delivery lumen in any suitable manner . the point of attachment may be inside or outside the lumen . the loops may be manufactured from any suitable material such as a metal , metal alloy , plastic , polymer , or other filamentous material or composite . the one or more loop portions may be attached at a second loop end to the control stem by soldering , fusing , an adhesive , or any other suitable means . in another embodiment , the loop portions may be attached to a first and a second loop end to the control stem . the support structure of fig6 a , 6 b , 7 a and 7 b may further include an atraumatic guiding tip of the kind described above to aid in positioning the support structure within the blood vessel . alternatively , parts of the loop portions which may protrude from the lumen when the loop portions are in their collapsed state may be used to guide the support structure into the blood vessel . one or more of the loop portions may be provided with a radiopaque or other marker to assist in this regard . retention means may also be provided with the support structure to retain the expandable member in an expanded condition within the vessel . the retention means may be in the form of a clamp , clip , thumb - slide or the like accessible from outside the patient &# 39 ; s body , and may facilitate adjustment of a deployed expandable member during a procedure . retention means may also impart additional rigidity and strength to the expandable member . thus , the retention member may be used to counteract excessively low pressures which may otherwise cause the expanded member to fail . a support structure of the kind illustrated in the figures may be delivered within a multilumen catheter 900 of the kind illustrated in cross section in fig9 . using this catheter , the support device 910 can be delivered through a first internal lumen 901 without interfering with flow in a second lumen 902 . a third lumen 903 may be provided for monitoring flow rates and pressures , for blood analysis or for delivering other percutaneous tools or devices to the vessel or as an inflation lumen for an occlusion balloon . it is to be understood that in the various embodiments of the present invention , the expanded member does not require constant contact with the vessel walls to provide the required support . for example , the diameter of the expanded member may be less than the diameter of the vessel so that the expanded member only contacts the vessel wall when the vessel begins to collapse . patency is considered to be maintained as long as the support device keeps the vessel open to a degree which is sufficient to maintain continuous flow . to avoid causing turbulence or other undesirable blood flow effects within the vessel , and to optimize flow in the vessel it may be desirable to substantially match the diameter of the expanded member to the diameter of the vessel . alternatively the expandable member may be shaped , e . g . as a coil or helix , to have minimal effect on the flow in the vessel . in one embodiment , the expandable member may have a slightly larger expanded diameter than the relaxed vessel to create an anchoring effect . depending on the size of the outflow vessel from which blood and perfusate is collected from the target region , there may be a natural tendency for the collection catheter tip to move about and contact the vessel , thus increasing the risk of vessel collapse or invagination of the catheter tip into the vessel wall . this can cause pooling of fluid in the isolated target region and may cause serious and permanent damage to the organ or region of the patient being treated . use of a support structure in conjunction with the collection catheter to maintain patency of the outflow vessel , in accordance with embodiments of the present invention can minimize the risk of these complications eventuating . thus , a collection catheter associated with the expanding member can be retained in position during fluid collection . this minimizes movement of the catheter tip , ensures that it is substantially centered relative to the vessel walls and improves withdrawal of fluid out of the vessel . at completion of the procedure , it is desirable that the expanded member is collapsed or compressed and recaptured , preferably in the catheter from which it was deployed . this facilitates removal of the support device from the patient . a reinforcing tip may be provided on the catheter end to strengthen it for recapture . alternatively or additionally , the tip may be coated with a lubricant and / or material having a low coefficient of friction to facilitate smooth recapture of the expandable member . the catheter may also have an internal coating of lubricant and / or a material having a low coefficient of friction to assist translation of support device along its interior during delivery and removal of the device from the patient . although the present invention has been described in terms of the presently preferred embodiments , it is to be understood that the disclosure is not to be interpreted as limiting . various alterations and modifications will no doubt become apparent to those skilled in the art after having read the above disclosure and it is intended that the present disclosure be interpreted as covering all alterations and modifications as fall within the true spirit and scope of the invention . effect of support device on flow rates and pressures achievable during recirculation in sheep right hepatic vein , cephalic vein , coronary sinus and renal vein during recirculation a 0 . 014 \u2033 diameter superelastic nitinol wire stem of 1 . 35 m length was used , coupled to an expandable member having 6 pre - shaped elliptical loop portions welded to the stem . a 0 . 024 \u2033 od atraumatic tip of 2 cm length attached to the distal end of the expandable member was used to position the device in the blood vessel . a balloon occlusion catheter was positioned in the vessel and the expandable member deployed at the tip of the catheter . the balloon was inflated to isolate and capture flows in the vessel and the catheter was connected to a standard extracorporeal circuit for blood circulation . negative pressures were observed in perfusion lines draining the coronary sinus , renal vein , right hepatic vein and cephalic vein during recirculation both with and without a support device . these data show that cavitation is prevented at certain pressures in the vessels tested where a support device is used , but is not prevented where the support device is absent in the vessel at those pressures . although cavitation may occur even with the support device , it occurs at higher flows . also , cavitation ceases sooner where the support device was employed allowing flow to return to normal . in the coronary sinus , recovery from cavitation was not possible without the support device , emphasizing the importance of the device in the procedure . the data further demonstrates that vessel collapse can be irreversible in the absence of a support structure . however , where a support structure is present , the vessel collapse may be reversed by increasing pressure in the vessel or by slowing or reversing the flow rate of fluid through the vessel . more specifically , considering the data for the right hepatic vein , flow rates of up to 250 ml per minute may be achieved before cavitation occurs where a support device is present in the vessel . under the same conditions but where there is no support device , flow rates of only up to 180 ml per minute are possible . a more striking example of the advantages of the support device is seen for the cephalic vein where no flow is achievable without the device . when the vessel wall is supported by the device flow rates of up to 200 ml per minute are noted before cavitation occurs . when the vessel wall is supported flow rates of up to 200 ml per minute are noted before cavitation occurs .", "category": "General tagging of new or cross-sectional technology"}
|
Does the category match the content of the patent?
| 0.25 |
1d5133cde40c92eaabe8131cbdf960efd3029fddffc551de233b4ea69b05bd19
| 0.087402 | 0.251953 | 0.047363 | 0.269531 | 0.138672 | 0.119141 |
null |
{"patent": "referring now to fig1 an apparatus for converting a supplied dc voltage to a plurality of regulated dc voltages v1 - v5 will be described . a dc to ac converter 11 supplies a variable frequency ac voltage to a transformer 20 . converter 11 may have any configuration known in the art , a half - bridge converter being shown in the drawing . converter 11 comprises a pair of semiconductor switching devices 16 and 17 , connected in series . switching devices 16 and 17 are shown as field - effect transistors ( fets ), although other devices such as insulated gate transistors ( igts ) or gate turn - off thyristors ( gtos ) may be used . converter 11 is connected to a supplied dc voltage . in this case , the dc voltage is obtained by rectifying an ac source 10 with a voltage doubler 9 comprised of diodes 12 and 13 and capacitors 14 and 15 . a voltage dobler is used to allow the half - bridge converter to provide an ac voltage with a magnitude equal to that of source 10 . on every other half cycle of source 10 , capacitor 14 is charged through diode 12 . capacitor 15 is charged through diode 13 on the other half cycles . a primary winding 21 of transformer 20 is connected between the junction of fets 16 and 17 and the junction of capacitors 14 and 15 . by alternately switching on fets 16 and 17 , an ac voltage having a variable frequency and an rms magnitude equal to the rms magnitude of supply 10 is supplied to primary winding 21 . the gates of fets 16 and 17 are connected to a control circuit which will be described below with reference to fig3 . converter 11 also includes a current sensor 18 for supplying a current signal to the control circuit . transformer 20 has a single secondary winding comprised of an even plurality of secondary winding segments 30 - 39 connected in series . the transformer secondary includes a center tap 40 which serves as ground for the dc outputs . preferably , secondary winding segments 30 - 39 are symmetrical about center tap 40 as to number of turns and wire gauge . as an alternate , electrically equivalent form of transformer 20 secondary , a plurality of separate , series - connected windings may be employed , with taps between adjacent separate windings 30 - 39 as shown in fig1 a . as used herein , the term &# 34 ; multiple secondary windings &# 34 ; is intended to encompass transformer secondaries of the type shown in both fig1 and fig1 a . a preferred construction for transformer 20 is shown in fig2 . primary winding 21 is would on one leg of core 25 multiple . secondary windings 30 - 39 are wound on another leg . to maintain tight coupling between the secondary windings multiple . secondary windings 30 - 39 are layered on top of each other separated by thin sheets of insulator 26 . returning to fig1 a pair of capacitors 41 and 42 are connected across the transformer secondary . the junction of cpacitors 41 and 42 is connected to center tap 40 . the leakage inductance resulting from the relatively loose coupling between primary winding 21 and multiple secondary windings 30 - 39 ( as a result of winding primary and secondaries on different legs of the core ) serves as a resonant inductor which resonates with capacitors 41 and 42 . an optional capacitor 43 may be included in the resonant circuit , directly across the transformer secondary to assist in tuning the resonant circuit . rectifying diodes 50 - 59 rectify the currents from secondary windings 30 - 39 , respectively , providing five dc voltages . the cathodes of diodes 50 - 59 are connected in pairs , the diode anodes in each pair being connected to opposite sides of center tap 40 . preferably , diodes 50 - 59 are connected in a symmetric configuration as in fig1 . the cathodes of each pair of diodes are connected to a separate filter , respectively . thus , a dc output voltage v1 is filtered by an inductor 60 and a capacitor 70 . each of inductors 60 - 64 provides a continuous current with minimum ripple to the load respectively connected thereto , thereby eliminating transient currents in transformer 20 and output capcitors 70 - 74 . inductors 60 - 64 are all wound on a single core and are tightly coupled magnetically . the number of turns of each inductor is proportional to each respective output voltage v1 - v5 . this winding arrangement improves the cross - regulation ( i . e . indirect regulation ) of the output voltages v1 - v5 which are not coupled to the control circuit , as described below . since multiple secondary windings 30 - 39 are layered , they are less closely coupled than bifilar windings . this loss in coupling tends to introduce transient voltage spikes across the rectifying diodes as current commutates from one half of the transformer secondary ( e . g . multiple secondary windings 30 - 34 ) to the other half ( e . g . multiple secondary windings 35 - 39 ). however , the placement of capacitors 41 and 42 from line to center tap ( common ) smoothes the current commutations . thus , resonant capacitors 41 and 42 act as lossless snubbers , reducing the voltage stresses and switching losses of diodes 50 - 59 . turning now to fig3 a control circuit is shown which regulates dc output voltages v1 - v5 ( fig1 ) by varying the output frequency of dc to ac converter 11 ( fig1 ). one of the dc output voltages , v2 for example , is provided to a voltage divider / attenuator 82 to provide a measured voltage which may be directly regulated by the control . the measured voltage is compared to a voltage reference 80 in a summer 83 , producing an error signal . the error signal is processed in a proportional - integral ( p - i ) controller 84 which in turn controls a voltage controlled oscillator ( vco ) 85 . preferably , the output of vco 85 is limited to frequencies above resonance of the resonant circuit . a driver circuit 86 is connected to vco 85 and to the gates of fets 16 and 17 . such driver circuits are known in the art . by way of example , driver circuit 86 may include a flip - flop for toggling by the output signal of vco 85 to provide two complementary signals . in addition , a lock - out circuit ( not shown ) in driver 86 may provide a minimum dead time between successive switchings to prevent shoot through of converter 11 . the control circuit also provides overcurrent protection . a current signal from current sensor 18 ( fig1 ) is provided to voltage divider / attenuator 87 to provide a measured current signal . the measured current signal is provided to the noninverting input of a comparator 88 . a current reference signal 81 is provided to the inverting input of comparator 88 . the output of comparator 88 is coupled to vco 85 . thus , if the measured current signal exceeds the current reference signal , vco 85 is set to its maximum . this increases the frequency of converter 11 and reduces the total current therein . by providing a substantial amount of hysteresis in comparator 88 , the output frequency of vco 85 will be lowered only gradually . the foregoing discloses a resonant power supply with multiple dc outputs but only a single power converter and a single transformer . by winding the output filter inductors on a sngle core , the multiple output voltages more nearly track one another . by layering the secondary windings , savings in size , weight and cost are achieved over bifilar windings . by placing a pair of resonant capacitors from each line to center tap , voltage transients across the rectifier diodes are reduced . a control circuit provides direct regulation of one of the output voltages and indirect regulation of the others by virtue of the close coupling of the transformer secondary windings and the close coupling of the output filter inductors . when desired , local series regulators may be used to improve the regulation of the indirectly regulated output voltages . while preferred embodiments of the present invention have been shown and described herein , it will be obvious that such embodiments are provided by way of example only . numerous variations , changes and substitutions will occur to those skilled in the art without departing from the invention herein . accordingly , it is intended that the invention be limited only by the spirit and scope of the appended claims .", "category": "Electricity"}
|
{"patent": "referring now to fig1 an apparatus for converting a supplied dc voltage to a plurality of regulated dc voltages v1 - v5 will be described . a dc to ac converter 11 supplies a variable frequency ac voltage to a transformer 20 . converter 11 may have any configuration known in the art , a half - bridge converter being shown in the drawing . converter 11 comprises a pair of semiconductor switching devices 16 and 17 , connected in series . switching devices 16 and 17 are shown as field - effect transistors ( fets ), although other devices such as insulated gate transistors ( igts ) or gate turn - off thyristors ( gtos ) may be used . converter 11 is connected to a supplied dc voltage . in this case , the dc voltage is obtained by rectifying an ac source 10 with a voltage doubler 9 comprised of diodes 12 and 13 and capacitors 14 and 15 . a voltage dobler is used to allow the half - bridge converter to provide an ac voltage with a magnitude equal to that of source 10 . on every other half cycle of source 10 , capacitor 14 is charged through diode 12 . capacitor 15 is charged through diode 13 on the other half cycles . a primary winding 21 of transformer 20 is connected between the junction of fets 16 and 17 and the junction of capacitors 14 and 15 . by alternately switching on fets 16 and 17 , an ac voltage having a variable frequency and an rms magnitude equal to the rms magnitude of supply 10 is supplied to primary winding 21 . the gates of fets 16 and 17 are connected to a control circuit which will be described below with reference to fig3 . converter 11 also includes a current sensor 18 for supplying a current signal to the control circuit . transformer 20 has a single secondary winding comprised of an even plurality of secondary winding segments 30 - 39 connected in series . the transformer secondary includes a center tap 40 which serves as ground for the dc outputs . preferably , secondary winding segments 30 - 39 are symmetrical about center tap 40 as to number of turns and wire gauge . as an alternate , electrically equivalent form of transformer 20 secondary , a plurality of separate , series - connected windings may be employed , with taps between adjacent separate windings 30 - 39 as shown in fig1 a . as used herein , the term &# 34 ; multiple secondary windings &# 34 ; is intended to encompass transformer secondaries of the type shown in both fig1 and fig1 a . a preferred construction for transformer 20 is shown in fig2 . primary winding 21 is would on one leg of core 25 multiple . secondary windings 30 - 39 are wound on another leg . to maintain tight coupling between the secondary windings multiple . secondary windings 30 - 39 are layered on top of each other separated by thin sheets of insulator 26 . returning to fig1 a pair of capacitors 41 and 42 are connected across the transformer secondary . the junction of cpacitors 41 and 42 is connected to center tap 40 . the leakage inductance resulting from the relatively loose coupling between primary winding 21 and multiple secondary windings 30 - 39 ( as a result of winding primary and secondaries on different legs of the core ) serves as a resonant inductor which resonates with capacitors 41 and 42 . an optional capacitor 43 may be included in the resonant circuit , directly across the transformer secondary to assist in tuning the resonant circuit . rectifying diodes 50 - 59 rectify the currents from secondary windings 30 - 39 , respectively , providing five dc voltages . the cathodes of diodes 50 - 59 are connected in pairs , the diode anodes in each pair being connected to opposite sides of center tap 40 . preferably , diodes 50 - 59 are connected in a symmetric configuration as in fig1 . the cathodes of each pair of diodes are connected to a separate filter , respectively . thus , a dc output voltage v1 is filtered by an inductor 60 and a capacitor 70 . each of inductors 60 - 64 provides a continuous current with minimum ripple to the load respectively connected thereto , thereby eliminating transient currents in transformer 20 and output capcitors 70 - 74 . inductors 60 - 64 are all wound on a single core and are tightly coupled magnetically . the number of turns of each inductor is proportional to each respective output voltage v1 - v5 . this winding arrangement improves the cross - regulation ( i . e . indirect regulation ) of the output voltages v1 - v5 which are not coupled to the control circuit , as described below . since multiple secondary windings 30 - 39 are layered , they are less closely coupled than bifilar windings . this loss in coupling tends to introduce transient voltage spikes across the rectifying diodes as current commutates from one half of the transformer secondary ( e . g . multiple secondary windings 30 - 34 ) to the other half ( e . g . multiple secondary windings 35 - 39 ). however , the placement of capacitors 41 and 42 from line to center tap ( common ) smoothes the current commutations . thus , resonant capacitors 41 and 42 act as lossless snubbers , reducing the voltage stresses and switching losses of diodes 50 - 59 . turning now to fig3 a control circuit is shown which regulates dc output voltages v1 - v5 ( fig1 ) by varying the output frequency of dc to ac converter 11 ( fig1 ). one of the dc output voltages , v2 for example , is provided to a voltage divider / attenuator 82 to provide a measured voltage which may be directly regulated by the control . the measured voltage is compared to a voltage reference 80 in a summer 83 , producing an error signal . the error signal is processed in a proportional - integral ( p - i ) controller 84 which in turn controls a voltage controlled oscillator ( vco ) 85 . preferably , the output of vco 85 is limited to frequencies above resonance of the resonant circuit . a driver circuit 86 is connected to vco 85 and to the gates of fets 16 and 17 . such driver circuits are known in the art . by way of example , driver circuit 86 may include a flip - flop for toggling by the output signal of vco 85 to provide two complementary signals . in addition , a lock - out circuit ( not shown ) in driver 86 may provide a minimum dead time between successive switchings to prevent shoot through of converter 11 . the control circuit also provides overcurrent protection . a current signal from current sensor 18 ( fig1 ) is provided to voltage divider / attenuator 87 to provide a measured current signal . the measured current signal is provided to the noninverting input of a comparator 88 . a current reference signal 81 is provided to the inverting input of comparator 88 . the output of comparator 88 is coupled to vco 85 . thus , if the measured current signal exceeds the current reference signal , vco 85 is set to its maximum . this increases the frequency of converter 11 and reduces the total current therein . by providing a substantial amount of hysteresis in comparator 88 , the output frequency of vco 85 will be lowered only gradually . the foregoing discloses a resonant power supply with multiple dc outputs but only a single power converter and a single transformer . by winding the output filter inductors on a sngle core , the multiple output voltages more nearly track one another . by layering the secondary windings , savings in size , weight and cost are achieved over bifilar windings . by placing a pair of resonant capacitors from each line to center tap , voltage transients across the rectifier diodes are reduced . a control circuit provides direct regulation of one of the output voltages and indirect regulation of the others by virtue of the close coupling of the transformer secondary windings and the close coupling of the output filter inductors . when desired , local series regulators may be used to improve the regulation of the indirectly regulated output voltages . while preferred embodiments of the present invention have been shown and described herein , it will be obvious that such embodiments are provided by way of example only . numerous variations , changes and substitutions will occur to those skilled in the art without departing from the invention herein . accordingly , it is intended that the invention be limited only by the spirit and scope of the appended claims .", "category": "Human Necessities"}
|
Does the category match the content of the patent?
| 0.25 |
140baa44de85b479375b6ca4965a25a7d7d16b2bf86f08319b15a6f93dadfaf6
| 0.753906 | 0.000123 | 0.675781 | 0.002182 | 0.582031 | 0.008606 |
null |
{"category": "Electricity", "patent": "referring now to fig1 an apparatus for converting a supplied dc voltage to a plurality of regulated dc voltages v1 - v5 will be described . a dc to ac converter 11 supplies a variable frequency ac voltage to a transformer 20 . converter 11 may have any configuration known in the art , a half - bridge converter being shown in the drawing . converter 11 comprises a pair of semiconductor switching devices 16 and 17 , connected in series . switching devices 16 and 17 are shown as field - effect transistors ( fets ), although other devices such as insulated gate transistors ( igts ) or gate turn - off thyristors ( gtos ) may be used . converter 11 is connected to a supplied dc voltage . in this case , the dc voltage is obtained by rectifying an ac source 10 with a voltage doubler 9 comprised of diodes 12 and 13 and capacitors 14 and 15 . a voltage dobler is used to allow the half - bridge converter to provide an ac voltage with a magnitude equal to that of source 10 . on every other half cycle of source 10 , capacitor 14 is charged through diode 12 . capacitor 15 is charged through diode 13 on the other half cycles . a primary winding 21 of transformer 20 is connected between the junction of fets 16 and 17 and the junction of capacitors 14 and 15 . by alternately switching on fets 16 and 17 , an ac voltage having a variable frequency and an rms magnitude equal to the rms magnitude of supply 10 is supplied to primary winding 21 . the gates of fets 16 and 17 are connected to a control circuit which will be described below with reference to fig3 . converter 11 also includes a current sensor 18 for supplying a current signal to the control circuit . transformer 20 has a single secondary winding comprised of an even plurality of secondary winding segments 30 - 39 connected in series . the transformer secondary includes a center tap 40 which serves as ground for the dc outputs . preferably , secondary winding segments 30 - 39 are symmetrical about center tap 40 as to number of turns and wire gauge . as an alternate , electrically equivalent form of transformer 20 secondary , a plurality of separate , series - connected windings may be employed , with taps between adjacent separate windings 30 - 39 as shown in fig1 a . as used herein , the term &# 34 ; multiple secondary windings &# 34 ; is intended to encompass transformer secondaries of the type shown in both fig1 and fig1 a . a preferred construction for transformer 20 is shown in fig2 . primary winding 21 is would on one leg of core 25 multiple . secondary windings 30 - 39 are wound on another leg . to maintain tight coupling between the secondary windings multiple . secondary windings 30 - 39 are layered on top of each other separated by thin sheets of insulator 26 . returning to fig1 a pair of capacitors 41 and 42 are connected across the transformer secondary . the junction of cpacitors 41 and 42 is connected to center tap 40 . the leakage inductance resulting from the relatively loose coupling between primary winding 21 and multiple secondary windings 30 - 39 ( as a result of winding primary and secondaries on different legs of the core ) serves as a resonant inductor which resonates with capacitors 41 and 42 . an optional capacitor 43 may be included in the resonant circuit , directly across the transformer secondary to assist in tuning the resonant circuit . rectifying diodes 50 - 59 rectify the currents from secondary windings 30 - 39 , respectively , providing five dc voltages . the cathodes of diodes 50 - 59 are connected in pairs , the diode anodes in each pair being connected to opposite sides of center tap 40 . preferably , diodes 50 - 59 are connected in a symmetric configuration as in fig1 . the cathodes of each pair of diodes are connected to a separate filter , respectively . thus , a dc output voltage v1 is filtered by an inductor 60 and a capacitor 70 . each of inductors 60 - 64 provides a continuous current with minimum ripple to the load respectively connected thereto , thereby eliminating transient currents in transformer 20 and output capcitors 70 - 74 . inductors 60 - 64 are all wound on a single core and are tightly coupled magnetically . the number of turns of each inductor is proportional to each respective output voltage v1 - v5 . this winding arrangement improves the cross - regulation ( i . e . indirect regulation ) of the output voltages v1 - v5 which are not coupled to the control circuit , as described below . since multiple secondary windings 30 - 39 are layered , they are less closely coupled than bifilar windings . this loss in coupling tends to introduce transient voltage spikes across the rectifying diodes as current commutates from one half of the transformer secondary ( e . g . multiple secondary windings 30 - 34 ) to the other half ( e . g . multiple secondary windings 35 - 39 ). however , the placement of capacitors 41 and 42 from line to center tap ( common ) smoothes the current commutations . thus , resonant capacitors 41 and 42 act as lossless snubbers , reducing the voltage stresses and switching losses of diodes 50 - 59 . turning now to fig3 a control circuit is shown which regulates dc output voltages v1 - v5 ( fig1 ) by varying the output frequency of dc to ac converter 11 ( fig1 ). one of the dc output voltages , v2 for example , is provided to a voltage divider / attenuator 82 to provide a measured voltage which may be directly regulated by the control . the measured voltage is compared to a voltage reference 80 in a summer 83 , producing an error signal . the error signal is processed in a proportional - integral ( p - i ) controller 84 which in turn controls a voltage controlled oscillator ( vco ) 85 . preferably , the output of vco 85 is limited to frequencies above resonance of the resonant circuit . a driver circuit 86 is connected to vco 85 and to the gates of fets 16 and 17 . such driver circuits are known in the art . by way of example , driver circuit 86 may include a flip - flop for toggling by the output signal of vco 85 to provide two complementary signals . in addition , a lock - out circuit ( not shown ) in driver 86 may provide a minimum dead time between successive switchings to prevent shoot through of converter 11 . the control circuit also provides overcurrent protection . a current signal from current sensor 18 ( fig1 ) is provided to voltage divider / attenuator 87 to provide a measured current signal . the measured current signal is provided to the noninverting input of a comparator 88 . a current reference signal 81 is provided to the inverting input of comparator 88 . the output of comparator 88 is coupled to vco 85 . thus , if the measured current signal exceeds the current reference signal , vco 85 is set to its maximum . this increases the frequency of converter 11 and reduces the total current therein . by providing a substantial amount of hysteresis in comparator 88 , the output frequency of vco 85 will be lowered only gradually . the foregoing discloses a resonant power supply with multiple dc outputs but only a single power converter and a single transformer . by winding the output filter inductors on a sngle core , the multiple output voltages more nearly track one another . by layering the secondary windings , savings in size , weight and cost are achieved over bifilar windings . by placing a pair of resonant capacitors from each line to center tap , voltage transients across the rectifier diodes are reduced . a control circuit provides direct regulation of one of the output voltages and indirect regulation of the others by virtue of the close coupling of the transformer secondary windings and the close coupling of the output filter inductors . when desired , local series regulators may be used to improve the regulation of the indirectly regulated output voltages . while preferred embodiments of the present invention have been shown and described herein , it will be obvious that such embodiments are provided by way of example only . numerous variations , changes and substitutions will occur to those skilled in the art without departing from the invention herein . accordingly , it is intended that the invention be limited only by the spirit and scope of the appended claims ."}
|
{"patent": "referring now to fig1 an apparatus for converting a supplied dc voltage to a plurality of regulated dc voltages v1 - v5 will be described . a dc to ac converter 11 supplies a variable frequency ac voltage to a transformer 20 . converter 11 may have any configuration known in the art , a half - bridge converter being shown in the drawing . converter 11 comprises a pair of semiconductor switching devices 16 and 17 , connected in series . switching devices 16 and 17 are shown as field - effect transistors ( fets ), although other devices such as insulated gate transistors ( igts ) or gate turn - off thyristors ( gtos ) may be used . converter 11 is connected to a supplied dc voltage . in this case , the dc voltage is obtained by rectifying an ac source 10 with a voltage doubler 9 comprised of diodes 12 and 13 and capacitors 14 and 15 . a voltage dobler is used to allow the half - bridge converter to provide an ac voltage with a magnitude equal to that of source 10 . on every other half cycle of source 10 , capacitor 14 is charged through diode 12 . capacitor 15 is charged through diode 13 on the other half cycles . a primary winding 21 of transformer 20 is connected between the junction of fets 16 and 17 and the junction of capacitors 14 and 15 . by alternately switching on fets 16 and 17 , an ac voltage having a variable frequency and an rms magnitude equal to the rms magnitude of supply 10 is supplied to primary winding 21 . the gates of fets 16 and 17 are connected to a control circuit which will be described below with reference to fig3 . converter 11 also includes a current sensor 18 for supplying a current signal to the control circuit . transformer 20 has a single secondary winding comprised of an even plurality of secondary winding segments 30 - 39 connected in series . the transformer secondary includes a center tap 40 which serves as ground for the dc outputs . preferably , secondary winding segments 30 - 39 are symmetrical about center tap 40 as to number of turns and wire gauge . as an alternate , electrically equivalent form of transformer 20 secondary , a plurality of separate , series - connected windings may be employed , with taps between adjacent separate windings 30 - 39 as shown in fig1 a . as used herein , the term &# 34 ; multiple secondary windings &# 34 ; is intended to encompass transformer secondaries of the type shown in both fig1 and fig1 a . a preferred construction for transformer 20 is shown in fig2 . primary winding 21 is would on one leg of core 25 multiple . secondary windings 30 - 39 are wound on another leg . to maintain tight coupling between the secondary windings multiple . secondary windings 30 - 39 are layered on top of each other separated by thin sheets of insulator 26 . returning to fig1 a pair of capacitors 41 and 42 are connected across the transformer secondary . the junction of cpacitors 41 and 42 is connected to center tap 40 . the leakage inductance resulting from the relatively loose coupling between primary winding 21 and multiple secondary windings 30 - 39 ( as a result of winding primary and secondaries on different legs of the core ) serves as a resonant inductor which resonates with capacitors 41 and 42 . an optional capacitor 43 may be included in the resonant circuit , directly across the transformer secondary to assist in tuning the resonant circuit . rectifying diodes 50 - 59 rectify the currents from secondary windings 30 - 39 , respectively , providing five dc voltages . the cathodes of diodes 50 - 59 are connected in pairs , the diode anodes in each pair being connected to opposite sides of center tap 40 . preferably , diodes 50 - 59 are connected in a symmetric configuration as in fig1 . the cathodes of each pair of diodes are connected to a separate filter , respectively . thus , a dc output voltage v1 is filtered by an inductor 60 and a capacitor 70 . each of inductors 60 - 64 provides a continuous current with minimum ripple to the load respectively connected thereto , thereby eliminating transient currents in transformer 20 and output capcitors 70 - 74 . inductors 60 - 64 are all wound on a single core and are tightly coupled magnetically . the number of turns of each inductor is proportional to each respective output voltage v1 - v5 . this winding arrangement improves the cross - regulation ( i . e . indirect regulation ) of the output voltages v1 - v5 which are not coupled to the control circuit , as described below . since multiple secondary windings 30 - 39 are layered , they are less closely coupled than bifilar windings . this loss in coupling tends to introduce transient voltage spikes across the rectifying diodes as current commutates from one half of the transformer secondary ( e . g . multiple secondary windings 30 - 34 ) to the other half ( e . g . multiple secondary windings 35 - 39 ). however , the placement of capacitors 41 and 42 from line to center tap ( common ) smoothes the current commutations . thus , resonant capacitors 41 and 42 act as lossless snubbers , reducing the voltage stresses and switching losses of diodes 50 - 59 . turning now to fig3 a control circuit is shown which regulates dc output voltages v1 - v5 ( fig1 ) by varying the output frequency of dc to ac converter 11 ( fig1 ). one of the dc output voltages , v2 for example , is provided to a voltage divider / attenuator 82 to provide a measured voltage which may be directly regulated by the control . the measured voltage is compared to a voltage reference 80 in a summer 83 , producing an error signal . the error signal is processed in a proportional - integral ( p - i ) controller 84 which in turn controls a voltage controlled oscillator ( vco ) 85 . preferably , the output of vco 85 is limited to frequencies above resonance of the resonant circuit . a driver circuit 86 is connected to vco 85 and to the gates of fets 16 and 17 . such driver circuits are known in the art . by way of example , driver circuit 86 may include a flip - flop for toggling by the output signal of vco 85 to provide two complementary signals . in addition , a lock - out circuit ( not shown ) in driver 86 may provide a minimum dead time between successive switchings to prevent shoot through of converter 11 . the control circuit also provides overcurrent protection . a current signal from current sensor 18 ( fig1 ) is provided to voltage divider / attenuator 87 to provide a measured current signal . the measured current signal is provided to the noninverting input of a comparator 88 . a current reference signal 81 is provided to the inverting input of comparator 88 . the output of comparator 88 is coupled to vco 85 . thus , if the measured current signal exceeds the current reference signal , vco 85 is set to its maximum . this increases the frequency of converter 11 and reduces the total current therein . by providing a substantial amount of hysteresis in comparator 88 , the output frequency of vco 85 will be lowered only gradually . the foregoing discloses a resonant power supply with multiple dc outputs but only a single power converter and a single transformer . by winding the output filter inductors on a sngle core , the multiple output voltages more nearly track one another . by layering the secondary windings , savings in size , weight and cost are achieved over bifilar windings . by placing a pair of resonant capacitors from each line to center tap , voltage transients across the rectifier diodes are reduced . a control circuit provides direct regulation of one of the output voltages and indirect regulation of the others by virtue of the close coupling of the transformer secondary windings and the close coupling of the output filter inductors . when desired , local series regulators may be used to improve the regulation of the indirectly regulated output voltages . while preferred embodiments of the present invention have been shown and described herein , it will be obvious that such embodiments are provided by way of example only . numerous variations , changes and substitutions will occur to those skilled in the art without departing from the invention herein . accordingly , it is intended that the invention be limited only by the spirit and scope of the appended claims .", "category": "Performing Operations; Transporting"}
|
Is the patent correctly categorized?
| 0.25 |
140baa44de85b479375b6ca4965a25a7d7d16b2bf86f08319b15a6f93dadfaf6
| 0.917969 | 0.005066 | 0.972656 | 0.029785 | 0.988281 | 0.341797 |
null |
{"patent": "referring now to fig1 an apparatus for converting a supplied dc voltage to a plurality of regulated dc voltages v1 - v5 will be described . a dc to ac converter 11 supplies a variable frequency ac voltage to a transformer 20 . converter 11 may have any configuration known in the art , a half - bridge converter being shown in the drawing . converter 11 comprises a pair of semiconductor switching devices 16 and 17 , connected in series . switching devices 16 and 17 are shown as field - effect transistors ( fets ), although other devices such as insulated gate transistors ( igts ) or gate turn - off thyristors ( gtos ) may be used . converter 11 is connected to a supplied dc voltage . in this case , the dc voltage is obtained by rectifying an ac source 10 with a voltage doubler 9 comprised of diodes 12 and 13 and capacitors 14 and 15 . a voltage dobler is used to allow the half - bridge converter to provide an ac voltage with a magnitude equal to that of source 10 . on every other half cycle of source 10 , capacitor 14 is charged through diode 12 . capacitor 15 is charged through diode 13 on the other half cycles . a primary winding 21 of transformer 20 is connected between the junction of fets 16 and 17 and the junction of capacitors 14 and 15 . by alternately switching on fets 16 and 17 , an ac voltage having a variable frequency and an rms magnitude equal to the rms magnitude of supply 10 is supplied to primary winding 21 . the gates of fets 16 and 17 are connected to a control circuit which will be described below with reference to fig3 . converter 11 also includes a current sensor 18 for supplying a current signal to the control circuit . transformer 20 has a single secondary winding comprised of an even plurality of secondary winding segments 30 - 39 connected in series . the transformer secondary includes a center tap 40 which serves as ground for the dc outputs . preferably , secondary winding segments 30 - 39 are symmetrical about center tap 40 as to number of turns and wire gauge . as an alternate , electrically equivalent form of transformer 20 secondary , a plurality of separate , series - connected windings may be employed , with taps between adjacent separate windings 30 - 39 as shown in fig1 a . as used herein , the term &# 34 ; multiple secondary windings &# 34 ; is intended to encompass transformer secondaries of the type shown in both fig1 and fig1 a . a preferred construction for transformer 20 is shown in fig2 . primary winding 21 is would on one leg of core 25 multiple . secondary windings 30 - 39 are wound on another leg . to maintain tight coupling between the secondary windings multiple . secondary windings 30 - 39 are layered on top of each other separated by thin sheets of insulator 26 . returning to fig1 a pair of capacitors 41 and 42 are connected across the transformer secondary . the junction of cpacitors 41 and 42 is connected to center tap 40 . the leakage inductance resulting from the relatively loose coupling between primary winding 21 and multiple secondary windings 30 - 39 ( as a result of winding primary and secondaries on different legs of the core ) serves as a resonant inductor which resonates with capacitors 41 and 42 . an optional capacitor 43 may be included in the resonant circuit , directly across the transformer secondary to assist in tuning the resonant circuit . rectifying diodes 50 - 59 rectify the currents from secondary windings 30 - 39 , respectively , providing five dc voltages . the cathodes of diodes 50 - 59 are connected in pairs , the diode anodes in each pair being connected to opposite sides of center tap 40 . preferably , diodes 50 - 59 are connected in a symmetric configuration as in fig1 . the cathodes of each pair of diodes are connected to a separate filter , respectively . thus , a dc output voltage v1 is filtered by an inductor 60 and a capacitor 70 . each of inductors 60 - 64 provides a continuous current with minimum ripple to the load respectively connected thereto , thereby eliminating transient currents in transformer 20 and output capcitors 70 - 74 . inductors 60 - 64 are all wound on a single core and are tightly coupled magnetically . the number of turns of each inductor is proportional to each respective output voltage v1 - v5 . this winding arrangement improves the cross - regulation ( i . e . indirect regulation ) of the output voltages v1 - v5 which are not coupled to the control circuit , as described below . since multiple secondary windings 30 - 39 are layered , they are less closely coupled than bifilar windings . this loss in coupling tends to introduce transient voltage spikes across the rectifying diodes as current commutates from one half of the transformer secondary ( e . g . multiple secondary windings 30 - 34 ) to the other half ( e . g . multiple secondary windings 35 - 39 ). however , the placement of capacitors 41 and 42 from line to center tap ( common ) smoothes the current commutations . thus , resonant capacitors 41 and 42 act as lossless snubbers , reducing the voltage stresses and switching losses of diodes 50 - 59 . turning now to fig3 a control circuit is shown which regulates dc output voltages v1 - v5 ( fig1 ) by varying the output frequency of dc to ac converter 11 ( fig1 ). one of the dc output voltages , v2 for example , is provided to a voltage divider / attenuator 82 to provide a measured voltage which may be directly regulated by the control . the measured voltage is compared to a voltage reference 80 in a summer 83 , producing an error signal . the error signal is processed in a proportional - integral ( p - i ) controller 84 which in turn controls a voltage controlled oscillator ( vco ) 85 . preferably , the output of vco 85 is limited to frequencies above resonance of the resonant circuit . a driver circuit 86 is connected to vco 85 and to the gates of fets 16 and 17 . such driver circuits are known in the art . by way of example , driver circuit 86 may include a flip - flop for toggling by the output signal of vco 85 to provide two complementary signals . in addition , a lock - out circuit ( not shown ) in driver 86 may provide a minimum dead time between successive switchings to prevent shoot through of converter 11 . the control circuit also provides overcurrent protection . a current signal from current sensor 18 ( fig1 ) is provided to voltage divider / attenuator 87 to provide a measured current signal . the measured current signal is provided to the noninverting input of a comparator 88 . a current reference signal 81 is provided to the inverting input of comparator 88 . the output of comparator 88 is coupled to vco 85 . thus , if the measured current signal exceeds the current reference signal , vco 85 is set to its maximum . this increases the frequency of converter 11 and reduces the total current therein . by providing a substantial amount of hysteresis in comparator 88 , the output frequency of vco 85 will be lowered only gradually . the foregoing discloses a resonant power supply with multiple dc outputs but only a single power converter and a single transformer . by winding the output filter inductors on a sngle core , the multiple output voltages more nearly track one another . by layering the secondary windings , savings in size , weight and cost are achieved over bifilar windings . by placing a pair of resonant capacitors from each line to center tap , voltage transients across the rectifier diodes are reduced . a control circuit provides direct regulation of one of the output voltages and indirect regulation of the others by virtue of the close coupling of the transformer secondary windings and the close coupling of the output filter inductors . when desired , local series regulators may be used to improve the regulation of the indirectly regulated output voltages . while preferred embodiments of the present invention have been shown and described herein , it will be obvious that such embodiments are provided by way of example only . numerous variations , changes and substitutions will occur to those skilled in the art without departing from the invention herein . accordingly , it is intended that the invention be limited only by the spirit and scope of the appended claims .", "category": "Electricity"}
|
{"category": "Chemistry; Metallurgy", "patent": "referring now to fig1 an apparatus for converting a supplied dc voltage to a plurality of regulated dc voltages v1 - v5 will be described . a dc to ac converter 11 supplies a variable frequency ac voltage to a transformer 20 . converter 11 may have any configuration known in the art , a half - bridge converter being shown in the drawing . converter 11 comprises a pair of semiconductor switching devices 16 and 17 , connected in series . switching devices 16 and 17 are shown as field - effect transistors ( fets ), although other devices such as insulated gate transistors ( igts ) or gate turn - off thyristors ( gtos ) may be used . converter 11 is connected to a supplied dc voltage . in this case , the dc voltage is obtained by rectifying an ac source 10 with a voltage doubler 9 comprised of diodes 12 and 13 and capacitors 14 and 15 . a voltage dobler is used to allow the half - bridge converter to provide an ac voltage with a magnitude equal to that of source 10 . on every other half cycle of source 10 , capacitor 14 is charged through diode 12 . capacitor 15 is charged through diode 13 on the other half cycles . a primary winding 21 of transformer 20 is connected between the junction of fets 16 and 17 and the junction of capacitors 14 and 15 . by alternately switching on fets 16 and 17 , an ac voltage having a variable frequency and an rms magnitude equal to the rms magnitude of supply 10 is supplied to primary winding 21 . the gates of fets 16 and 17 are connected to a control circuit which will be described below with reference to fig3 . converter 11 also includes a current sensor 18 for supplying a current signal to the control circuit . transformer 20 has a single secondary winding comprised of an even plurality of secondary winding segments 30 - 39 connected in series . the transformer secondary includes a center tap 40 which serves as ground for the dc outputs . preferably , secondary winding segments 30 - 39 are symmetrical about center tap 40 as to number of turns and wire gauge . as an alternate , electrically equivalent form of transformer 20 secondary , a plurality of separate , series - connected windings may be employed , with taps between adjacent separate windings 30 - 39 as shown in fig1 a . as used herein , the term &# 34 ; multiple secondary windings &# 34 ; is intended to encompass transformer secondaries of the type shown in both fig1 and fig1 a . a preferred construction for transformer 20 is shown in fig2 . primary winding 21 is would on one leg of core 25 multiple . secondary windings 30 - 39 are wound on another leg . to maintain tight coupling between the secondary windings multiple . secondary windings 30 - 39 are layered on top of each other separated by thin sheets of insulator 26 . returning to fig1 a pair of capacitors 41 and 42 are connected across the transformer secondary . the junction of cpacitors 41 and 42 is connected to center tap 40 . the leakage inductance resulting from the relatively loose coupling between primary winding 21 and multiple secondary windings 30 - 39 ( as a result of winding primary and secondaries on different legs of the core ) serves as a resonant inductor which resonates with capacitors 41 and 42 . an optional capacitor 43 may be included in the resonant circuit , directly across the transformer secondary to assist in tuning the resonant circuit . rectifying diodes 50 - 59 rectify the currents from secondary windings 30 - 39 , respectively , providing five dc voltages . the cathodes of diodes 50 - 59 are connected in pairs , the diode anodes in each pair being connected to opposite sides of center tap 40 . preferably , diodes 50 - 59 are connected in a symmetric configuration as in fig1 . the cathodes of each pair of diodes are connected to a separate filter , respectively . thus , a dc output voltage v1 is filtered by an inductor 60 and a capacitor 70 . each of inductors 60 - 64 provides a continuous current with minimum ripple to the load respectively connected thereto , thereby eliminating transient currents in transformer 20 and output capcitors 70 - 74 . inductors 60 - 64 are all wound on a single core and are tightly coupled magnetically . the number of turns of each inductor is proportional to each respective output voltage v1 - v5 . this winding arrangement improves the cross - regulation ( i . e . indirect regulation ) of the output voltages v1 - v5 which are not coupled to the control circuit , as described below . since multiple secondary windings 30 - 39 are layered , they are less closely coupled than bifilar windings . this loss in coupling tends to introduce transient voltage spikes across the rectifying diodes as current commutates from one half of the transformer secondary ( e . g . multiple secondary windings 30 - 34 ) to the other half ( e . g . multiple secondary windings 35 - 39 ). however , the placement of capacitors 41 and 42 from line to center tap ( common ) smoothes the current commutations . thus , resonant capacitors 41 and 42 act as lossless snubbers , reducing the voltage stresses and switching losses of diodes 50 - 59 . turning now to fig3 a control circuit is shown which regulates dc output voltages v1 - v5 ( fig1 ) by varying the output frequency of dc to ac converter 11 ( fig1 ). one of the dc output voltages , v2 for example , is provided to a voltage divider / attenuator 82 to provide a measured voltage which may be directly regulated by the control . the measured voltage is compared to a voltage reference 80 in a summer 83 , producing an error signal . the error signal is processed in a proportional - integral ( p - i ) controller 84 which in turn controls a voltage controlled oscillator ( vco ) 85 . preferably , the output of vco 85 is limited to frequencies above resonance of the resonant circuit . a driver circuit 86 is connected to vco 85 and to the gates of fets 16 and 17 . such driver circuits are known in the art . by way of example , driver circuit 86 may include a flip - flop for toggling by the output signal of vco 85 to provide two complementary signals . in addition , a lock - out circuit ( not shown ) in driver 86 may provide a minimum dead time between successive switchings to prevent shoot through of converter 11 . the control circuit also provides overcurrent protection . a current signal from current sensor 18 ( fig1 ) is provided to voltage divider / attenuator 87 to provide a measured current signal . the measured current signal is provided to the noninverting input of a comparator 88 . a current reference signal 81 is provided to the inverting input of comparator 88 . the output of comparator 88 is coupled to vco 85 . thus , if the measured current signal exceeds the current reference signal , vco 85 is set to its maximum . this increases the frequency of converter 11 and reduces the total current therein . by providing a substantial amount of hysteresis in comparator 88 , the output frequency of vco 85 will be lowered only gradually . the foregoing discloses a resonant power supply with multiple dc outputs but only a single power converter and a single transformer . by winding the output filter inductors on a sngle core , the multiple output voltages more nearly track one another . by layering the secondary windings , savings in size , weight and cost are achieved over bifilar windings . by placing a pair of resonant capacitors from each line to center tap , voltage transients across the rectifier diodes are reduced . a control circuit provides direct regulation of one of the output voltages and indirect regulation of the others by virtue of the close coupling of the transformer secondary windings and the close coupling of the output filter inductors . when desired , local series regulators may be used to improve the regulation of the indirectly regulated output voltages . while preferred embodiments of the present invention have been shown and described herein , it will be obvious that such embodiments are provided by way of example only . numerous variations , changes and substitutions will occur to those skilled in the art without departing from the invention herein . accordingly , it is intended that the invention be limited only by the spirit and scope of the appended claims ."}
|
Does the category match the content of the patent?
| 0.25 |
140baa44de85b479375b6ca4965a25a7d7d16b2bf86f08319b15a6f93dadfaf6
| 0.757813 | 0.001244 | 0.675781 | 0.003601 | 0.585938 | 0.004913 |
null |
{"patent": "referring now to fig1 an apparatus for converting a supplied dc voltage to a plurality of regulated dc voltages v1 - v5 will be described . a dc to ac converter 11 supplies a variable frequency ac voltage to a transformer 20 . converter 11 may have any configuration known in the art , a half - bridge converter being shown in the drawing . converter 11 comprises a pair of semiconductor switching devices 16 and 17 , connected in series . switching devices 16 and 17 are shown as field - effect transistors ( fets ), although other devices such as insulated gate transistors ( igts ) or gate turn - off thyristors ( gtos ) may be used . converter 11 is connected to a supplied dc voltage . in this case , the dc voltage is obtained by rectifying an ac source 10 with a voltage doubler 9 comprised of diodes 12 and 13 and capacitors 14 and 15 . a voltage dobler is used to allow the half - bridge converter to provide an ac voltage with a magnitude equal to that of source 10 . on every other half cycle of source 10 , capacitor 14 is charged through diode 12 . capacitor 15 is charged through diode 13 on the other half cycles . a primary winding 21 of transformer 20 is connected between the junction of fets 16 and 17 and the junction of capacitors 14 and 15 . by alternately switching on fets 16 and 17 , an ac voltage having a variable frequency and an rms magnitude equal to the rms magnitude of supply 10 is supplied to primary winding 21 . the gates of fets 16 and 17 are connected to a control circuit which will be described below with reference to fig3 . converter 11 also includes a current sensor 18 for supplying a current signal to the control circuit . transformer 20 has a single secondary winding comprised of an even plurality of secondary winding segments 30 - 39 connected in series . the transformer secondary includes a center tap 40 which serves as ground for the dc outputs . preferably , secondary winding segments 30 - 39 are symmetrical about center tap 40 as to number of turns and wire gauge . as an alternate , electrically equivalent form of transformer 20 secondary , a plurality of separate , series - connected windings may be employed , with taps between adjacent separate windings 30 - 39 as shown in fig1 a . as used herein , the term &# 34 ; multiple secondary windings &# 34 ; is intended to encompass transformer secondaries of the type shown in both fig1 and fig1 a . a preferred construction for transformer 20 is shown in fig2 . primary winding 21 is would on one leg of core 25 multiple . secondary windings 30 - 39 are wound on another leg . to maintain tight coupling between the secondary windings multiple . secondary windings 30 - 39 are layered on top of each other separated by thin sheets of insulator 26 . returning to fig1 a pair of capacitors 41 and 42 are connected across the transformer secondary . the junction of cpacitors 41 and 42 is connected to center tap 40 . the leakage inductance resulting from the relatively loose coupling between primary winding 21 and multiple secondary windings 30 - 39 ( as a result of winding primary and secondaries on different legs of the core ) serves as a resonant inductor which resonates with capacitors 41 and 42 . an optional capacitor 43 may be included in the resonant circuit , directly across the transformer secondary to assist in tuning the resonant circuit . rectifying diodes 50 - 59 rectify the currents from secondary windings 30 - 39 , respectively , providing five dc voltages . the cathodes of diodes 50 - 59 are connected in pairs , the diode anodes in each pair being connected to opposite sides of center tap 40 . preferably , diodes 50 - 59 are connected in a symmetric configuration as in fig1 . the cathodes of each pair of diodes are connected to a separate filter , respectively . thus , a dc output voltage v1 is filtered by an inductor 60 and a capacitor 70 . each of inductors 60 - 64 provides a continuous current with minimum ripple to the load respectively connected thereto , thereby eliminating transient currents in transformer 20 and output capcitors 70 - 74 . inductors 60 - 64 are all wound on a single core and are tightly coupled magnetically . the number of turns of each inductor is proportional to each respective output voltage v1 - v5 . this winding arrangement improves the cross - regulation ( i . e . indirect regulation ) of the output voltages v1 - v5 which are not coupled to the control circuit , as described below . since multiple secondary windings 30 - 39 are layered , they are less closely coupled than bifilar windings . this loss in coupling tends to introduce transient voltage spikes across the rectifying diodes as current commutates from one half of the transformer secondary ( e . g . multiple secondary windings 30 - 34 ) to the other half ( e . g . multiple secondary windings 35 - 39 ). however , the placement of capacitors 41 and 42 from line to center tap ( common ) smoothes the current commutations . thus , resonant capacitors 41 and 42 act as lossless snubbers , reducing the voltage stresses and switching losses of diodes 50 - 59 . turning now to fig3 a control circuit is shown which regulates dc output voltages v1 - v5 ( fig1 ) by varying the output frequency of dc to ac converter 11 ( fig1 ). one of the dc output voltages , v2 for example , is provided to a voltage divider / attenuator 82 to provide a measured voltage which may be directly regulated by the control . the measured voltage is compared to a voltage reference 80 in a summer 83 , producing an error signal . the error signal is processed in a proportional - integral ( p - i ) controller 84 which in turn controls a voltage controlled oscillator ( vco ) 85 . preferably , the output of vco 85 is limited to frequencies above resonance of the resonant circuit . a driver circuit 86 is connected to vco 85 and to the gates of fets 16 and 17 . such driver circuits are known in the art . by way of example , driver circuit 86 may include a flip - flop for toggling by the output signal of vco 85 to provide two complementary signals . in addition , a lock - out circuit ( not shown ) in driver 86 may provide a minimum dead time between successive switchings to prevent shoot through of converter 11 . the control circuit also provides overcurrent protection . a current signal from current sensor 18 ( fig1 ) is provided to voltage divider / attenuator 87 to provide a measured current signal . the measured current signal is provided to the noninverting input of a comparator 88 . a current reference signal 81 is provided to the inverting input of comparator 88 . the output of comparator 88 is coupled to vco 85 . thus , if the measured current signal exceeds the current reference signal , vco 85 is set to its maximum . this increases the frequency of converter 11 and reduces the total current therein . by providing a substantial amount of hysteresis in comparator 88 , the output frequency of vco 85 will be lowered only gradually . the foregoing discloses a resonant power supply with multiple dc outputs but only a single power converter and a single transformer . by winding the output filter inductors on a sngle core , the multiple output voltages more nearly track one another . by layering the secondary windings , savings in size , weight and cost are achieved over bifilar windings . by placing a pair of resonant capacitors from each line to center tap , voltage transients across the rectifier diodes are reduced . a control circuit provides direct regulation of one of the output voltages and indirect regulation of the others by virtue of the close coupling of the transformer secondary windings and the close coupling of the output filter inductors . when desired , local series regulators may be used to improve the regulation of the indirectly regulated output voltages . while preferred embodiments of the present invention have been shown and described herein , it will be obvious that such embodiments are provided by way of example only . numerous variations , changes and substitutions will occur to those skilled in the art without departing from the invention herein . accordingly , it is intended that the invention be limited only by the spirit and scope of the appended claims .", "category": "Electricity"}
|
{"category": "Textiles; Paper", "patent": "referring now to fig1 an apparatus for converting a supplied dc voltage to a plurality of regulated dc voltages v1 - v5 will be described . a dc to ac converter 11 supplies a variable frequency ac voltage to a transformer 20 . converter 11 may have any configuration known in the art , a half - bridge converter being shown in the drawing . converter 11 comprises a pair of semiconductor switching devices 16 and 17 , connected in series . switching devices 16 and 17 are shown as field - effect transistors ( fets ), although other devices such as insulated gate transistors ( igts ) or gate turn - off thyristors ( gtos ) may be used . converter 11 is connected to a supplied dc voltage . in this case , the dc voltage is obtained by rectifying an ac source 10 with a voltage doubler 9 comprised of diodes 12 and 13 and capacitors 14 and 15 . a voltage dobler is used to allow the half - bridge converter to provide an ac voltage with a magnitude equal to that of source 10 . on every other half cycle of source 10 , capacitor 14 is charged through diode 12 . capacitor 15 is charged through diode 13 on the other half cycles . a primary winding 21 of transformer 20 is connected between the junction of fets 16 and 17 and the junction of capacitors 14 and 15 . by alternately switching on fets 16 and 17 , an ac voltage having a variable frequency and an rms magnitude equal to the rms magnitude of supply 10 is supplied to primary winding 21 . the gates of fets 16 and 17 are connected to a control circuit which will be described below with reference to fig3 . converter 11 also includes a current sensor 18 for supplying a current signal to the control circuit . transformer 20 has a single secondary winding comprised of an even plurality of secondary winding segments 30 - 39 connected in series . the transformer secondary includes a center tap 40 which serves as ground for the dc outputs . preferably , secondary winding segments 30 - 39 are symmetrical about center tap 40 as to number of turns and wire gauge . as an alternate , electrically equivalent form of transformer 20 secondary , a plurality of separate , series - connected windings may be employed , with taps between adjacent separate windings 30 - 39 as shown in fig1 a . as used herein , the term &# 34 ; multiple secondary windings &# 34 ; is intended to encompass transformer secondaries of the type shown in both fig1 and fig1 a . a preferred construction for transformer 20 is shown in fig2 . primary winding 21 is would on one leg of core 25 multiple . secondary windings 30 - 39 are wound on another leg . to maintain tight coupling between the secondary windings multiple . secondary windings 30 - 39 are layered on top of each other separated by thin sheets of insulator 26 . returning to fig1 a pair of capacitors 41 and 42 are connected across the transformer secondary . the junction of cpacitors 41 and 42 is connected to center tap 40 . the leakage inductance resulting from the relatively loose coupling between primary winding 21 and multiple secondary windings 30 - 39 ( as a result of winding primary and secondaries on different legs of the core ) serves as a resonant inductor which resonates with capacitors 41 and 42 . an optional capacitor 43 may be included in the resonant circuit , directly across the transformer secondary to assist in tuning the resonant circuit . rectifying diodes 50 - 59 rectify the currents from secondary windings 30 - 39 , respectively , providing five dc voltages . the cathodes of diodes 50 - 59 are connected in pairs , the diode anodes in each pair being connected to opposite sides of center tap 40 . preferably , diodes 50 - 59 are connected in a symmetric configuration as in fig1 . the cathodes of each pair of diodes are connected to a separate filter , respectively . thus , a dc output voltage v1 is filtered by an inductor 60 and a capacitor 70 . each of inductors 60 - 64 provides a continuous current with minimum ripple to the load respectively connected thereto , thereby eliminating transient currents in transformer 20 and output capcitors 70 - 74 . inductors 60 - 64 are all wound on a single core and are tightly coupled magnetically . the number of turns of each inductor is proportional to each respective output voltage v1 - v5 . this winding arrangement improves the cross - regulation ( i . e . indirect regulation ) of the output voltages v1 - v5 which are not coupled to the control circuit , as described below . since multiple secondary windings 30 - 39 are layered , they are less closely coupled than bifilar windings . this loss in coupling tends to introduce transient voltage spikes across the rectifying diodes as current commutates from one half of the transformer secondary ( e . g . multiple secondary windings 30 - 34 ) to the other half ( e . g . multiple secondary windings 35 - 39 ). however , the placement of capacitors 41 and 42 from line to center tap ( common ) smoothes the current commutations . thus , resonant capacitors 41 and 42 act as lossless snubbers , reducing the voltage stresses and switching losses of diodes 50 - 59 . turning now to fig3 a control circuit is shown which regulates dc output voltages v1 - v5 ( fig1 ) by varying the output frequency of dc to ac converter 11 ( fig1 ). one of the dc output voltages , v2 for example , is provided to a voltage divider / attenuator 82 to provide a measured voltage which may be directly regulated by the control . the measured voltage is compared to a voltage reference 80 in a summer 83 , producing an error signal . the error signal is processed in a proportional - integral ( p - i ) controller 84 which in turn controls a voltage controlled oscillator ( vco ) 85 . preferably , the output of vco 85 is limited to frequencies above resonance of the resonant circuit . a driver circuit 86 is connected to vco 85 and to the gates of fets 16 and 17 . such driver circuits are known in the art . by way of example , driver circuit 86 may include a flip - flop for toggling by the output signal of vco 85 to provide two complementary signals . in addition , a lock - out circuit ( not shown ) in driver 86 may provide a minimum dead time between successive switchings to prevent shoot through of converter 11 . the control circuit also provides overcurrent protection . a current signal from current sensor 18 ( fig1 ) is provided to voltage divider / attenuator 87 to provide a measured current signal . the measured current signal is provided to the noninverting input of a comparator 88 . a current reference signal 81 is provided to the inverting input of comparator 88 . the output of comparator 88 is coupled to vco 85 . thus , if the measured current signal exceeds the current reference signal , vco 85 is set to its maximum . this increases the frequency of converter 11 and reduces the total current therein . by providing a substantial amount of hysteresis in comparator 88 , the output frequency of vco 85 will be lowered only gradually . the foregoing discloses a resonant power supply with multiple dc outputs but only a single power converter and a single transformer . by winding the output filter inductors on a sngle core , the multiple output voltages more nearly track one another . by layering the secondary windings , savings in size , weight and cost are achieved over bifilar windings . by placing a pair of resonant capacitors from each line to center tap , voltage transients across the rectifier diodes are reduced . a control circuit provides direct regulation of one of the output voltages and indirect regulation of the others by virtue of the close coupling of the transformer secondary windings and the close coupling of the output filter inductors . when desired , local series regulators may be used to improve the regulation of the indirectly regulated output voltages . while preferred embodiments of the present invention have been shown and described herein , it will be obvious that such embodiments are provided by way of example only . numerous variations , changes and substitutions will occur to those skilled in the art without departing from the invention herein . accordingly , it is intended that the invention be limited only by the spirit and scope of the appended claims ."}
|
Is the categorization of this patent accurate?
| 0.25 |
140baa44de85b479375b6ca4965a25a7d7d16b2bf86f08319b15a6f93dadfaf6
| 0.239258 | 0.057373 | 0.519531 | 0.021973 | 0.535156 | 0.208984 |
null |
{"category": "Electricity", "patent": "referring now to fig1 an apparatus for converting a supplied dc voltage to a plurality of regulated dc voltages v1 - v5 will be described . a dc to ac converter 11 supplies a variable frequency ac voltage to a transformer 20 . converter 11 may have any configuration known in the art , a half - bridge converter being shown in the drawing . converter 11 comprises a pair of semiconductor switching devices 16 and 17 , connected in series . switching devices 16 and 17 are shown as field - effect transistors ( fets ), although other devices such as insulated gate transistors ( igts ) or gate turn - off thyristors ( gtos ) may be used . converter 11 is connected to a supplied dc voltage . in this case , the dc voltage is obtained by rectifying an ac source 10 with a voltage doubler 9 comprised of diodes 12 and 13 and capacitors 14 and 15 . a voltage dobler is used to allow the half - bridge converter to provide an ac voltage with a magnitude equal to that of source 10 . on every other half cycle of source 10 , capacitor 14 is charged through diode 12 . capacitor 15 is charged through diode 13 on the other half cycles . a primary winding 21 of transformer 20 is connected between the junction of fets 16 and 17 and the junction of capacitors 14 and 15 . by alternately switching on fets 16 and 17 , an ac voltage having a variable frequency and an rms magnitude equal to the rms magnitude of supply 10 is supplied to primary winding 21 . the gates of fets 16 and 17 are connected to a control circuit which will be described below with reference to fig3 . converter 11 also includes a current sensor 18 for supplying a current signal to the control circuit . transformer 20 has a single secondary winding comprised of an even plurality of secondary winding segments 30 - 39 connected in series . the transformer secondary includes a center tap 40 which serves as ground for the dc outputs . preferably , secondary winding segments 30 - 39 are symmetrical about center tap 40 as to number of turns and wire gauge . as an alternate , electrically equivalent form of transformer 20 secondary , a plurality of separate , series - connected windings may be employed , with taps between adjacent separate windings 30 - 39 as shown in fig1 a . as used herein , the term &# 34 ; multiple secondary windings &# 34 ; is intended to encompass transformer secondaries of the type shown in both fig1 and fig1 a . a preferred construction for transformer 20 is shown in fig2 . primary winding 21 is would on one leg of core 25 multiple . secondary windings 30 - 39 are wound on another leg . to maintain tight coupling between the secondary windings multiple . secondary windings 30 - 39 are layered on top of each other separated by thin sheets of insulator 26 . returning to fig1 a pair of capacitors 41 and 42 are connected across the transformer secondary . the junction of cpacitors 41 and 42 is connected to center tap 40 . the leakage inductance resulting from the relatively loose coupling between primary winding 21 and multiple secondary windings 30 - 39 ( as a result of winding primary and secondaries on different legs of the core ) serves as a resonant inductor which resonates with capacitors 41 and 42 . an optional capacitor 43 may be included in the resonant circuit , directly across the transformer secondary to assist in tuning the resonant circuit . rectifying diodes 50 - 59 rectify the currents from secondary windings 30 - 39 , respectively , providing five dc voltages . the cathodes of diodes 50 - 59 are connected in pairs , the diode anodes in each pair being connected to opposite sides of center tap 40 . preferably , diodes 50 - 59 are connected in a symmetric configuration as in fig1 . the cathodes of each pair of diodes are connected to a separate filter , respectively . thus , a dc output voltage v1 is filtered by an inductor 60 and a capacitor 70 . each of inductors 60 - 64 provides a continuous current with minimum ripple to the load respectively connected thereto , thereby eliminating transient currents in transformer 20 and output capcitors 70 - 74 . inductors 60 - 64 are all wound on a single core and are tightly coupled magnetically . the number of turns of each inductor is proportional to each respective output voltage v1 - v5 . this winding arrangement improves the cross - regulation ( i . e . indirect regulation ) of the output voltages v1 - v5 which are not coupled to the control circuit , as described below . since multiple secondary windings 30 - 39 are layered , they are less closely coupled than bifilar windings . this loss in coupling tends to introduce transient voltage spikes across the rectifying diodes as current commutates from one half of the transformer secondary ( e . g . multiple secondary windings 30 - 34 ) to the other half ( e . g . multiple secondary windings 35 - 39 ). however , the placement of capacitors 41 and 42 from line to center tap ( common ) smoothes the current commutations . thus , resonant capacitors 41 and 42 act as lossless snubbers , reducing the voltage stresses and switching losses of diodes 50 - 59 . turning now to fig3 a control circuit is shown which regulates dc output voltages v1 - v5 ( fig1 ) by varying the output frequency of dc to ac converter 11 ( fig1 ). one of the dc output voltages , v2 for example , is provided to a voltage divider / attenuator 82 to provide a measured voltage which may be directly regulated by the control . the measured voltage is compared to a voltage reference 80 in a summer 83 , producing an error signal . the error signal is processed in a proportional - integral ( p - i ) controller 84 which in turn controls a voltage controlled oscillator ( vco ) 85 . preferably , the output of vco 85 is limited to frequencies above resonance of the resonant circuit . a driver circuit 86 is connected to vco 85 and to the gates of fets 16 and 17 . such driver circuits are known in the art . by way of example , driver circuit 86 may include a flip - flop for toggling by the output signal of vco 85 to provide two complementary signals . in addition , a lock - out circuit ( not shown ) in driver 86 may provide a minimum dead time between successive switchings to prevent shoot through of converter 11 . the control circuit also provides overcurrent protection . a current signal from current sensor 18 ( fig1 ) is provided to voltage divider / attenuator 87 to provide a measured current signal . the measured current signal is provided to the noninverting input of a comparator 88 . a current reference signal 81 is provided to the inverting input of comparator 88 . the output of comparator 88 is coupled to vco 85 . thus , if the measured current signal exceeds the current reference signal , vco 85 is set to its maximum . this increases the frequency of converter 11 and reduces the total current therein . by providing a substantial amount of hysteresis in comparator 88 , the output frequency of vco 85 will be lowered only gradually . the foregoing discloses a resonant power supply with multiple dc outputs but only a single power converter and a single transformer . by winding the output filter inductors on a sngle core , the multiple output voltages more nearly track one another . by layering the secondary windings , savings in size , weight and cost are achieved over bifilar windings . by placing a pair of resonant capacitors from each line to center tap , voltage transients across the rectifier diodes are reduced . a control circuit provides direct regulation of one of the output voltages and indirect regulation of the others by virtue of the close coupling of the transformer secondary windings and the close coupling of the output filter inductors . when desired , local series regulators may be used to improve the regulation of the indirectly regulated output voltages . while preferred embodiments of the present invention have been shown and described herein , it will be obvious that such embodiments are provided by way of example only . numerous variations , changes and substitutions will occur to those skilled in the art without departing from the invention herein . accordingly , it is intended that the invention be limited only by the spirit and scope of the appended claims ."}
|
{"patent": "referring now to fig1 an apparatus for converting a supplied dc voltage to a plurality of regulated dc voltages v1 - v5 will be described . a dc to ac converter 11 supplies a variable frequency ac voltage to a transformer 20 . converter 11 may have any configuration known in the art , a half - bridge converter being shown in the drawing . converter 11 comprises a pair of semiconductor switching devices 16 and 17 , connected in series . switching devices 16 and 17 are shown as field - effect transistors ( fets ), although other devices such as insulated gate transistors ( igts ) or gate turn - off thyristors ( gtos ) may be used . converter 11 is connected to a supplied dc voltage . in this case , the dc voltage is obtained by rectifying an ac source 10 with a voltage doubler 9 comprised of diodes 12 and 13 and capacitors 14 and 15 . a voltage dobler is used to allow the half - bridge converter to provide an ac voltage with a magnitude equal to that of source 10 . on every other half cycle of source 10 , capacitor 14 is charged through diode 12 . capacitor 15 is charged through diode 13 on the other half cycles . a primary winding 21 of transformer 20 is connected between the junction of fets 16 and 17 and the junction of capacitors 14 and 15 . by alternately switching on fets 16 and 17 , an ac voltage having a variable frequency and an rms magnitude equal to the rms magnitude of supply 10 is supplied to primary winding 21 . the gates of fets 16 and 17 are connected to a control circuit which will be described below with reference to fig3 . converter 11 also includes a current sensor 18 for supplying a current signal to the control circuit . transformer 20 has a single secondary winding comprised of an even plurality of secondary winding segments 30 - 39 connected in series . the transformer secondary includes a center tap 40 which serves as ground for the dc outputs . preferably , secondary winding segments 30 - 39 are symmetrical about center tap 40 as to number of turns and wire gauge . as an alternate , electrically equivalent form of transformer 20 secondary , a plurality of separate , series - connected windings may be employed , with taps between adjacent separate windings 30 - 39 as shown in fig1 a . as used herein , the term &# 34 ; multiple secondary windings &# 34 ; is intended to encompass transformer secondaries of the type shown in both fig1 and fig1 a . a preferred construction for transformer 20 is shown in fig2 . primary winding 21 is would on one leg of core 25 multiple . secondary windings 30 - 39 are wound on another leg . to maintain tight coupling between the secondary windings multiple . secondary windings 30 - 39 are layered on top of each other separated by thin sheets of insulator 26 . returning to fig1 a pair of capacitors 41 and 42 are connected across the transformer secondary . the junction of cpacitors 41 and 42 is connected to center tap 40 . the leakage inductance resulting from the relatively loose coupling between primary winding 21 and multiple secondary windings 30 - 39 ( as a result of winding primary and secondaries on different legs of the core ) serves as a resonant inductor which resonates with capacitors 41 and 42 . an optional capacitor 43 may be included in the resonant circuit , directly across the transformer secondary to assist in tuning the resonant circuit . rectifying diodes 50 - 59 rectify the currents from secondary windings 30 - 39 , respectively , providing five dc voltages . the cathodes of diodes 50 - 59 are connected in pairs , the diode anodes in each pair being connected to opposite sides of center tap 40 . preferably , diodes 50 - 59 are connected in a symmetric configuration as in fig1 . the cathodes of each pair of diodes are connected to a separate filter , respectively . thus , a dc output voltage v1 is filtered by an inductor 60 and a capacitor 70 . each of inductors 60 - 64 provides a continuous current with minimum ripple to the load respectively connected thereto , thereby eliminating transient currents in transformer 20 and output capcitors 70 - 74 . inductors 60 - 64 are all wound on a single core and are tightly coupled magnetically . the number of turns of each inductor is proportional to each respective output voltage v1 - v5 . this winding arrangement improves the cross - regulation ( i . e . indirect regulation ) of the output voltages v1 - v5 which are not coupled to the control circuit , as described below . since multiple secondary windings 30 - 39 are layered , they are less closely coupled than bifilar windings . this loss in coupling tends to introduce transient voltage spikes across the rectifying diodes as current commutates from one half of the transformer secondary ( e . g . multiple secondary windings 30 - 34 ) to the other half ( e . g . multiple secondary windings 35 - 39 ). however , the placement of capacitors 41 and 42 from line to center tap ( common ) smoothes the current commutations . thus , resonant capacitors 41 and 42 act as lossless snubbers , reducing the voltage stresses and switching losses of diodes 50 - 59 . turning now to fig3 a control circuit is shown which regulates dc output voltages v1 - v5 ( fig1 ) by varying the output frequency of dc to ac converter 11 ( fig1 ). one of the dc output voltages , v2 for example , is provided to a voltage divider / attenuator 82 to provide a measured voltage which may be directly regulated by the control . the measured voltage is compared to a voltage reference 80 in a summer 83 , producing an error signal . the error signal is processed in a proportional - integral ( p - i ) controller 84 which in turn controls a voltage controlled oscillator ( vco ) 85 . preferably , the output of vco 85 is limited to frequencies above resonance of the resonant circuit . a driver circuit 86 is connected to vco 85 and to the gates of fets 16 and 17 . such driver circuits are known in the art . by way of example , driver circuit 86 may include a flip - flop for toggling by the output signal of vco 85 to provide two complementary signals . in addition , a lock - out circuit ( not shown ) in driver 86 may provide a minimum dead time between successive switchings to prevent shoot through of converter 11 . the control circuit also provides overcurrent protection . a current signal from current sensor 18 ( fig1 ) is provided to voltage divider / attenuator 87 to provide a measured current signal . the measured current signal is provided to the noninverting input of a comparator 88 . a current reference signal 81 is provided to the inverting input of comparator 88 . the output of comparator 88 is coupled to vco 85 . thus , if the measured current signal exceeds the current reference signal , vco 85 is set to its maximum . this increases the frequency of converter 11 and reduces the total current therein . by providing a substantial amount of hysteresis in comparator 88 , the output frequency of vco 85 will be lowered only gradually . the foregoing discloses a resonant power supply with multiple dc outputs but only a single power converter and a single transformer . by winding the output filter inductors on a sngle core , the multiple output voltages more nearly track one another . by layering the secondary windings , savings in size , weight and cost are achieved over bifilar windings . by placing a pair of resonant capacitors from each line to center tap , voltage transients across the rectifier diodes are reduced . a control circuit provides direct regulation of one of the output voltages and indirect regulation of the others by virtue of the close coupling of the transformer secondary windings and the close coupling of the output filter inductors . when desired , local series regulators may be used to improve the regulation of the indirectly regulated output voltages . while preferred embodiments of the present invention have been shown and described herein , it will be obvious that such embodiments are provided by way of example only . numerous variations , changes and substitutions will occur to those skilled in the art without departing from the invention herein . accordingly , it is intended that the invention be limited only by the spirit and scope of the appended claims .", "category": "Fixed Constructions"}
|
Is the patent correctly categorized?
| 0.25 |
140baa44de85b479375b6ca4965a25a7d7d16b2bf86f08319b15a6f93dadfaf6
| 0.917969 | 0.263672 | 0.972656 | 0.369141 | 0.988281 | 0.617188 |
null |
{"category": "Electricity", "patent": "referring now to fig1 an apparatus for converting a supplied dc voltage to a plurality of regulated dc voltages v1 - v5 will be described . a dc to ac converter 11 supplies a variable frequency ac voltage to a transformer 20 . converter 11 may have any configuration known in the art , a half - bridge converter being shown in the drawing . converter 11 comprises a pair of semiconductor switching devices 16 and 17 , connected in series . switching devices 16 and 17 are shown as field - effect transistors ( fets ), although other devices such as insulated gate transistors ( igts ) or gate turn - off thyristors ( gtos ) may be used . converter 11 is connected to a supplied dc voltage . in this case , the dc voltage is obtained by rectifying an ac source 10 with a voltage doubler 9 comprised of diodes 12 and 13 and capacitors 14 and 15 . a voltage dobler is used to allow the half - bridge converter to provide an ac voltage with a magnitude equal to that of source 10 . on every other half cycle of source 10 , capacitor 14 is charged through diode 12 . capacitor 15 is charged through diode 13 on the other half cycles . a primary winding 21 of transformer 20 is connected between the junction of fets 16 and 17 and the junction of capacitors 14 and 15 . by alternately switching on fets 16 and 17 , an ac voltage having a variable frequency and an rms magnitude equal to the rms magnitude of supply 10 is supplied to primary winding 21 . the gates of fets 16 and 17 are connected to a control circuit which will be described below with reference to fig3 . converter 11 also includes a current sensor 18 for supplying a current signal to the control circuit . transformer 20 has a single secondary winding comprised of an even plurality of secondary winding segments 30 - 39 connected in series . the transformer secondary includes a center tap 40 which serves as ground for the dc outputs . preferably , secondary winding segments 30 - 39 are symmetrical about center tap 40 as to number of turns and wire gauge . as an alternate , electrically equivalent form of transformer 20 secondary , a plurality of separate , series - connected windings may be employed , with taps between adjacent separate windings 30 - 39 as shown in fig1 a . as used herein , the term &# 34 ; multiple secondary windings &# 34 ; is intended to encompass transformer secondaries of the type shown in both fig1 and fig1 a . a preferred construction for transformer 20 is shown in fig2 . primary winding 21 is would on one leg of core 25 multiple . secondary windings 30 - 39 are wound on another leg . to maintain tight coupling between the secondary windings multiple . secondary windings 30 - 39 are layered on top of each other separated by thin sheets of insulator 26 . returning to fig1 a pair of capacitors 41 and 42 are connected across the transformer secondary . the junction of cpacitors 41 and 42 is connected to center tap 40 . the leakage inductance resulting from the relatively loose coupling between primary winding 21 and multiple secondary windings 30 - 39 ( as a result of winding primary and secondaries on different legs of the core ) serves as a resonant inductor which resonates with capacitors 41 and 42 . an optional capacitor 43 may be included in the resonant circuit , directly across the transformer secondary to assist in tuning the resonant circuit . rectifying diodes 50 - 59 rectify the currents from secondary windings 30 - 39 , respectively , providing five dc voltages . the cathodes of diodes 50 - 59 are connected in pairs , the diode anodes in each pair being connected to opposite sides of center tap 40 . preferably , diodes 50 - 59 are connected in a symmetric configuration as in fig1 . the cathodes of each pair of diodes are connected to a separate filter , respectively . thus , a dc output voltage v1 is filtered by an inductor 60 and a capacitor 70 . each of inductors 60 - 64 provides a continuous current with minimum ripple to the load respectively connected thereto , thereby eliminating transient currents in transformer 20 and output capcitors 70 - 74 . inductors 60 - 64 are all wound on a single core and are tightly coupled magnetically . the number of turns of each inductor is proportional to each respective output voltage v1 - v5 . this winding arrangement improves the cross - regulation ( i . e . indirect regulation ) of the output voltages v1 - v5 which are not coupled to the control circuit , as described below . since multiple secondary windings 30 - 39 are layered , they are less closely coupled than bifilar windings . this loss in coupling tends to introduce transient voltage spikes across the rectifying diodes as current commutates from one half of the transformer secondary ( e . g . multiple secondary windings 30 - 34 ) to the other half ( e . g . multiple secondary windings 35 - 39 ). however , the placement of capacitors 41 and 42 from line to center tap ( common ) smoothes the current commutations . thus , resonant capacitors 41 and 42 act as lossless snubbers , reducing the voltage stresses and switching losses of diodes 50 - 59 . turning now to fig3 a control circuit is shown which regulates dc output voltages v1 - v5 ( fig1 ) by varying the output frequency of dc to ac converter 11 ( fig1 ). one of the dc output voltages , v2 for example , is provided to a voltage divider / attenuator 82 to provide a measured voltage which may be directly regulated by the control . the measured voltage is compared to a voltage reference 80 in a summer 83 , producing an error signal . the error signal is processed in a proportional - integral ( p - i ) controller 84 which in turn controls a voltage controlled oscillator ( vco ) 85 . preferably , the output of vco 85 is limited to frequencies above resonance of the resonant circuit . a driver circuit 86 is connected to vco 85 and to the gates of fets 16 and 17 . such driver circuits are known in the art . by way of example , driver circuit 86 may include a flip - flop for toggling by the output signal of vco 85 to provide two complementary signals . in addition , a lock - out circuit ( not shown ) in driver 86 may provide a minimum dead time between successive switchings to prevent shoot through of converter 11 . the control circuit also provides overcurrent protection . a current signal from current sensor 18 ( fig1 ) is provided to voltage divider / attenuator 87 to provide a measured current signal . the measured current signal is provided to the noninverting input of a comparator 88 . a current reference signal 81 is provided to the inverting input of comparator 88 . the output of comparator 88 is coupled to vco 85 . thus , if the measured current signal exceeds the current reference signal , vco 85 is set to its maximum . this increases the frequency of converter 11 and reduces the total current therein . by providing a substantial amount of hysteresis in comparator 88 , the output frequency of vco 85 will be lowered only gradually . the foregoing discloses a resonant power supply with multiple dc outputs but only a single power converter and a single transformer . by winding the output filter inductors on a sngle core , the multiple output voltages more nearly track one another . by layering the secondary windings , savings in size , weight and cost are achieved over bifilar windings . by placing a pair of resonant capacitors from each line to center tap , voltage transients across the rectifier diodes are reduced . a control circuit provides direct regulation of one of the output voltages and indirect regulation of the others by virtue of the close coupling of the transformer secondary windings and the close coupling of the output filter inductors . when desired , local series regulators may be used to improve the regulation of the indirectly regulated output voltages . while preferred embodiments of the present invention have been shown and described herein , it will be obvious that such embodiments are provided by way of example only . numerous variations , changes and substitutions will occur to those skilled in the art without departing from the invention herein . accordingly , it is intended that the invention be limited only by the spirit and scope of the appended claims ."}
|
{"category": "Mechanical Engineering; Lightning; Heating; Weapons; Blasting", "patent": "referring now to fig1 an apparatus for converting a supplied dc voltage to a plurality of regulated dc voltages v1 - v5 will be described . a dc to ac converter 11 supplies a variable frequency ac voltage to a transformer 20 . converter 11 may have any configuration known in the art , a half - bridge converter being shown in the drawing . converter 11 comprises a pair of semiconductor switching devices 16 and 17 , connected in series . switching devices 16 and 17 are shown as field - effect transistors ( fets ), although other devices such as insulated gate transistors ( igts ) or gate turn - off thyristors ( gtos ) may be used . converter 11 is connected to a supplied dc voltage . in this case , the dc voltage is obtained by rectifying an ac source 10 with a voltage doubler 9 comprised of diodes 12 and 13 and capacitors 14 and 15 . a voltage dobler is used to allow the half - bridge converter to provide an ac voltage with a magnitude equal to that of source 10 . on every other half cycle of source 10 , capacitor 14 is charged through diode 12 . capacitor 15 is charged through diode 13 on the other half cycles . a primary winding 21 of transformer 20 is connected between the junction of fets 16 and 17 and the junction of capacitors 14 and 15 . by alternately switching on fets 16 and 17 , an ac voltage having a variable frequency and an rms magnitude equal to the rms magnitude of supply 10 is supplied to primary winding 21 . the gates of fets 16 and 17 are connected to a control circuit which will be described below with reference to fig3 . converter 11 also includes a current sensor 18 for supplying a current signal to the control circuit . transformer 20 has a single secondary winding comprised of an even plurality of secondary winding segments 30 - 39 connected in series . the transformer secondary includes a center tap 40 which serves as ground for the dc outputs . preferably , secondary winding segments 30 - 39 are symmetrical about center tap 40 as to number of turns and wire gauge . as an alternate , electrically equivalent form of transformer 20 secondary , a plurality of separate , series - connected windings may be employed , with taps between adjacent separate windings 30 - 39 as shown in fig1 a . as used herein , the term &# 34 ; multiple secondary windings &# 34 ; is intended to encompass transformer secondaries of the type shown in both fig1 and fig1 a . a preferred construction for transformer 20 is shown in fig2 . primary winding 21 is would on one leg of core 25 multiple . secondary windings 30 - 39 are wound on another leg . to maintain tight coupling between the secondary windings multiple . secondary windings 30 - 39 are layered on top of each other separated by thin sheets of insulator 26 . returning to fig1 a pair of capacitors 41 and 42 are connected across the transformer secondary . the junction of cpacitors 41 and 42 is connected to center tap 40 . the leakage inductance resulting from the relatively loose coupling between primary winding 21 and multiple secondary windings 30 - 39 ( as a result of winding primary and secondaries on different legs of the core ) serves as a resonant inductor which resonates with capacitors 41 and 42 . an optional capacitor 43 may be included in the resonant circuit , directly across the transformer secondary to assist in tuning the resonant circuit . rectifying diodes 50 - 59 rectify the currents from secondary windings 30 - 39 , respectively , providing five dc voltages . the cathodes of diodes 50 - 59 are connected in pairs , the diode anodes in each pair being connected to opposite sides of center tap 40 . preferably , diodes 50 - 59 are connected in a symmetric configuration as in fig1 . the cathodes of each pair of diodes are connected to a separate filter , respectively . thus , a dc output voltage v1 is filtered by an inductor 60 and a capacitor 70 . each of inductors 60 - 64 provides a continuous current with minimum ripple to the load respectively connected thereto , thereby eliminating transient currents in transformer 20 and output capcitors 70 - 74 . inductors 60 - 64 are all wound on a single core and are tightly coupled magnetically . the number of turns of each inductor is proportional to each respective output voltage v1 - v5 . this winding arrangement improves the cross - regulation ( i . e . indirect regulation ) of the output voltages v1 - v5 which are not coupled to the control circuit , as described below . since multiple secondary windings 30 - 39 are layered , they are less closely coupled than bifilar windings . this loss in coupling tends to introduce transient voltage spikes across the rectifying diodes as current commutates from one half of the transformer secondary ( e . g . multiple secondary windings 30 - 34 ) to the other half ( e . g . multiple secondary windings 35 - 39 ). however , the placement of capacitors 41 and 42 from line to center tap ( common ) smoothes the current commutations . thus , resonant capacitors 41 and 42 act as lossless snubbers , reducing the voltage stresses and switching losses of diodes 50 - 59 . turning now to fig3 a control circuit is shown which regulates dc output voltages v1 - v5 ( fig1 ) by varying the output frequency of dc to ac converter 11 ( fig1 ). one of the dc output voltages , v2 for example , is provided to a voltage divider / attenuator 82 to provide a measured voltage which may be directly regulated by the control . the measured voltage is compared to a voltage reference 80 in a summer 83 , producing an error signal . the error signal is processed in a proportional - integral ( p - i ) controller 84 which in turn controls a voltage controlled oscillator ( vco ) 85 . preferably , the output of vco 85 is limited to frequencies above resonance of the resonant circuit . a driver circuit 86 is connected to vco 85 and to the gates of fets 16 and 17 . such driver circuits are known in the art . by way of example , driver circuit 86 may include a flip - flop for toggling by the output signal of vco 85 to provide two complementary signals . in addition , a lock - out circuit ( not shown ) in driver 86 may provide a minimum dead time between successive switchings to prevent shoot through of converter 11 . the control circuit also provides overcurrent protection . a current signal from current sensor 18 ( fig1 ) is provided to voltage divider / attenuator 87 to provide a measured current signal . the measured current signal is provided to the noninverting input of a comparator 88 . a current reference signal 81 is provided to the inverting input of comparator 88 . the output of comparator 88 is coupled to vco 85 . thus , if the measured current signal exceeds the current reference signal , vco 85 is set to its maximum . this increases the frequency of converter 11 and reduces the total current therein . by providing a substantial amount of hysteresis in comparator 88 , the output frequency of vco 85 will be lowered only gradually . the foregoing discloses a resonant power supply with multiple dc outputs but only a single power converter and a single transformer . by winding the output filter inductors on a sngle core , the multiple output voltages more nearly track one another . by layering the secondary windings , savings in size , weight and cost are achieved over bifilar windings . by placing a pair of resonant capacitors from each line to center tap , voltage transients across the rectifier diodes are reduced . a control circuit provides direct regulation of one of the output voltages and indirect regulation of the others by virtue of the close coupling of the transformer secondary windings and the close coupling of the output filter inductors . when desired , local series regulators may be used to improve the regulation of the indirectly regulated output voltages . while preferred embodiments of the present invention have been shown and described herein , it will be obvious that such embodiments are provided by way of example only . numerous variations , changes and substitutions will occur to those skilled in the art without departing from the invention herein . accordingly , it is intended that the invention be limited only by the spirit and scope of the appended claims ."}
|
Is the patent correctly categorized?
| 0.25 |
140baa44de85b479375b6ca4965a25a7d7d16b2bf86f08319b15a6f93dadfaf6
| 0.917969 | 0.004456 | 0.972656 | 0.001328 | 0.988281 | 0.170898 |
null |
{"category": "Electricity", "patent": "referring now to fig1 an apparatus for converting a supplied dc voltage to a plurality of regulated dc voltages v1 - v5 will be described . a dc to ac converter 11 supplies a variable frequency ac voltage to a transformer 20 . converter 11 may have any configuration known in the art , a half - bridge converter being shown in the drawing . converter 11 comprises a pair of semiconductor switching devices 16 and 17 , connected in series . switching devices 16 and 17 are shown as field - effect transistors ( fets ), although other devices such as insulated gate transistors ( igts ) or gate turn - off thyristors ( gtos ) may be used . converter 11 is connected to a supplied dc voltage . in this case , the dc voltage is obtained by rectifying an ac source 10 with a voltage doubler 9 comprised of diodes 12 and 13 and capacitors 14 and 15 . a voltage dobler is used to allow the half - bridge converter to provide an ac voltage with a magnitude equal to that of source 10 . on every other half cycle of source 10 , capacitor 14 is charged through diode 12 . capacitor 15 is charged through diode 13 on the other half cycles . a primary winding 21 of transformer 20 is connected between the junction of fets 16 and 17 and the junction of capacitors 14 and 15 . by alternately switching on fets 16 and 17 , an ac voltage having a variable frequency and an rms magnitude equal to the rms magnitude of supply 10 is supplied to primary winding 21 . the gates of fets 16 and 17 are connected to a control circuit which will be described below with reference to fig3 . converter 11 also includes a current sensor 18 for supplying a current signal to the control circuit . transformer 20 has a single secondary winding comprised of an even plurality of secondary winding segments 30 - 39 connected in series . the transformer secondary includes a center tap 40 which serves as ground for the dc outputs . preferably , secondary winding segments 30 - 39 are symmetrical about center tap 40 as to number of turns and wire gauge . as an alternate , electrically equivalent form of transformer 20 secondary , a plurality of separate , series - connected windings may be employed , with taps between adjacent separate windings 30 - 39 as shown in fig1 a . as used herein , the term &# 34 ; multiple secondary windings &# 34 ; is intended to encompass transformer secondaries of the type shown in both fig1 and fig1 a . a preferred construction for transformer 20 is shown in fig2 . primary winding 21 is would on one leg of core 25 multiple . secondary windings 30 - 39 are wound on another leg . to maintain tight coupling between the secondary windings multiple . secondary windings 30 - 39 are layered on top of each other separated by thin sheets of insulator 26 . returning to fig1 a pair of capacitors 41 and 42 are connected across the transformer secondary . the junction of cpacitors 41 and 42 is connected to center tap 40 . the leakage inductance resulting from the relatively loose coupling between primary winding 21 and multiple secondary windings 30 - 39 ( as a result of winding primary and secondaries on different legs of the core ) serves as a resonant inductor which resonates with capacitors 41 and 42 . an optional capacitor 43 may be included in the resonant circuit , directly across the transformer secondary to assist in tuning the resonant circuit . rectifying diodes 50 - 59 rectify the currents from secondary windings 30 - 39 , respectively , providing five dc voltages . the cathodes of diodes 50 - 59 are connected in pairs , the diode anodes in each pair being connected to opposite sides of center tap 40 . preferably , diodes 50 - 59 are connected in a symmetric configuration as in fig1 . the cathodes of each pair of diodes are connected to a separate filter , respectively . thus , a dc output voltage v1 is filtered by an inductor 60 and a capacitor 70 . each of inductors 60 - 64 provides a continuous current with minimum ripple to the load respectively connected thereto , thereby eliminating transient currents in transformer 20 and output capcitors 70 - 74 . inductors 60 - 64 are all wound on a single core and are tightly coupled magnetically . the number of turns of each inductor is proportional to each respective output voltage v1 - v5 . this winding arrangement improves the cross - regulation ( i . e . indirect regulation ) of the output voltages v1 - v5 which are not coupled to the control circuit , as described below . since multiple secondary windings 30 - 39 are layered , they are less closely coupled than bifilar windings . this loss in coupling tends to introduce transient voltage spikes across the rectifying diodes as current commutates from one half of the transformer secondary ( e . g . multiple secondary windings 30 - 34 ) to the other half ( e . g . multiple secondary windings 35 - 39 ). however , the placement of capacitors 41 and 42 from line to center tap ( common ) smoothes the current commutations . thus , resonant capacitors 41 and 42 act as lossless snubbers , reducing the voltage stresses and switching losses of diodes 50 - 59 . turning now to fig3 a control circuit is shown which regulates dc output voltages v1 - v5 ( fig1 ) by varying the output frequency of dc to ac converter 11 ( fig1 ). one of the dc output voltages , v2 for example , is provided to a voltage divider / attenuator 82 to provide a measured voltage which may be directly regulated by the control . the measured voltage is compared to a voltage reference 80 in a summer 83 , producing an error signal . the error signal is processed in a proportional - integral ( p - i ) controller 84 which in turn controls a voltage controlled oscillator ( vco ) 85 . preferably , the output of vco 85 is limited to frequencies above resonance of the resonant circuit . a driver circuit 86 is connected to vco 85 and to the gates of fets 16 and 17 . such driver circuits are known in the art . by way of example , driver circuit 86 may include a flip - flop for toggling by the output signal of vco 85 to provide two complementary signals . in addition , a lock - out circuit ( not shown ) in driver 86 may provide a minimum dead time between successive switchings to prevent shoot through of converter 11 . the control circuit also provides overcurrent protection . a current signal from current sensor 18 ( fig1 ) is provided to voltage divider / attenuator 87 to provide a measured current signal . the measured current signal is provided to the noninverting input of a comparator 88 . a current reference signal 81 is provided to the inverting input of comparator 88 . the output of comparator 88 is coupled to vco 85 . thus , if the measured current signal exceeds the current reference signal , vco 85 is set to its maximum . this increases the frequency of converter 11 and reduces the total current therein . by providing a substantial amount of hysteresis in comparator 88 , the output frequency of vco 85 will be lowered only gradually . the foregoing discloses a resonant power supply with multiple dc outputs but only a single power converter and a single transformer . by winding the output filter inductors on a sngle core , the multiple output voltages more nearly track one another . by layering the secondary windings , savings in size , weight and cost are achieved over bifilar windings . by placing a pair of resonant capacitors from each line to center tap , voltage transients across the rectifier diodes are reduced . a control circuit provides direct regulation of one of the output voltages and indirect regulation of the others by virtue of the close coupling of the transformer secondary windings and the close coupling of the output filter inductors . when desired , local series regulators may be used to improve the regulation of the indirectly regulated output voltages . while preferred embodiments of the present invention have been shown and described herein , it will be obvious that such embodiments are provided by way of example only . numerous variations , changes and substitutions will occur to those skilled in the art without departing from the invention herein . accordingly , it is intended that the invention be limited only by the spirit and scope of the appended claims ."}
|
{"patent": "referring now to fig1 an apparatus for converting a supplied dc voltage to a plurality of regulated dc voltages v1 - v5 will be described . a dc to ac converter 11 supplies a variable frequency ac voltage to a transformer 20 . converter 11 may have any configuration known in the art , a half - bridge converter being shown in the drawing . converter 11 comprises a pair of semiconductor switching devices 16 and 17 , connected in series . switching devices 16 and 17 are shown as field - effect transistors ( fets ), although other devices such as insulated gate transistors ( igts ) or gate turn - off thyristors ( gtos ) may be used . converter 11 is connected to a supplied dc voltage . in this case , the dc voltage is obtained by rectifying an ac source 10 with a voltage doubler 9 comprised of diodes 12 and 13 and capacitors 14 and 15 . a voltage dobler is used to allow the half - bridge converter to provide an ac voltage with a magnitude equal to that of source 10 . on every other half cycle of source 10 , capacitor 14 is charged through diode 12 . capacitor 15 is charged through diode 13 on the other half cycles . a primary winding 21 of transformer 20 is connected between the junction of fets 16 and 17 and the junction of capacitors 14 and 15 . by alternately switching on fets 16 and 17 , an ac voltage having a variable frequency and an rms magnitude equal to the rms magnitude of supply 10 is supplied to primary winding 21 . the gates of fets 16 and 17 are connected to a control circuit which will be described below with reference to fig3 . converter 11 also includes a current sensor 18 for supplying a current signal to the control circuit . transformer 20 has a single secondary winding comprised of an even plurality of secondary winding segments 30 - 39 connected in series . the transformer secondary includes a center tap 40 which serves as ground for the dc outputs . preferably , secondary winding segments 30 - 39 are symmetrical about center tap 40 as to number of turns and wire gauge . as an alternate , electrically equivalent form of transformer 20 secondary , a plurality of separate , series - connected windings may be employed , with taps between adjacent separate windings 30 - 39 as shown in fig1 a . as used herein , the term &# 34 ; multiple secondary windings &# 34 ; is intended to encompass transformer secondaries of the type shown in both fig1 and fig1 a . a preferred construction for transformer 20 is shown in fig2 . primary winding 21 is would on one leg of core 25 multiple . secondary windings 30 - 39 are wound on another leg . to maintain tight coupling between the secondary windings multiple . secondary windings 30 - 39 are layered on top of each other separated by thin sheets of insulator 26 . returning to fig1 a pair of capacitors 41 and 42 are connected across the transformer secondary . the junction of cpacitors 41 and 42 is connected to center tap 40 . the leakage inductance resulting from the relatively loose coupling between primary winding 21 and multiple secondary windings 30 - 39 ( as a result of winding primary and secondaries on different legs of the core ) serves as a resonant inductor which resonates with capacitors 41 and 42 . an optional capacitor 43 may be included in the resonant circuit , directly across the transformer secondary to assist in tuning the resonant circuit . rectifying diodes 50 - 59 rectify the currents from secondary windings 30 - 39 , respectively , providing five dc voltages . the cathodes of diodes 50 - 59 are connected in pairs , the diode anodes in each pair being connected to opposite sides of center tap 40 . preferably , diodes 50 - 59 are connected in a symmetric configuration as in fig1 . the cathodes of each pair of diodes are connected to a separate filter , respectively . thus , a dc output voltage v1 is filtered by an inductor 60 and a capacitor 70 . each of inductors 60 - 64 provides a continuous current with minimum ripple to the load respectively connected thereto , thereby eliminating transient currents in transformer 20 and output capcitors 70 - 74 . inductors 60 - 64 are all wound on a single core and are tightly coupled magnetically . the number of turns of each inductor is proportional to each respective output voltage v1 - v5 . this winding arrangement improves the cross - regulation ( i . e . indirect regulation ) of the output voltages v1 - v5 which are not coupled to the control circuit , as described below . since multiple secondary windings 30 - 39 are layered , they are less closely coupled than bifilar windings . this loss in coupling tends to introduce transient voltage spikes across the rectifying diodes as current commutates from one half of the transformer secondary ( e . g . multiple secondary windings 30 - 34 ) to the other half ( e . g . multiple secondary windings 35 - 39 ). however , the placement of capacitors 41 and 42 from line to center tap ( common ) smoothes the current commutations . thus , resonant capacitors 41 and 42 act as lossless snubbers , reducing the voltage stresses and switching losses of diodes 50 - 59 . turning now to fig3 a control circuit is shown which regulates dc output voltages v1 - v5 ( fig1 ) by varying the output frequency of dc to ac converter 11 ( fig1 ). one of the dc output voltages , v2 for example , is provided to a voltage divider / attenuator 82 to provide a measured voltage which may be directly regulated by the control . the measured voltage is compared to a voltage reference 80 in a summer 83 , producing an error signal . the error signal is processed in a proportional - integral ( p - i ) controller 84 which in turn controls a voltage controlled oscillator ( vco ) 85 . preferably , the output of vco 85 is limited to frequencies above resonance of the resonant circuit . a driver circuit 86 is connected to vco 85 and to the gates of fets 16 and 17 . such driver circuits are known in the art . by way of example , driver circuit 86 may include a flip - flop for toggling by the output signal of vco 85 to provide two complementary signals . in addition , a lock - out circuit ( not shown ) in driver 86 may provide a minimum dead time between successive switchings to prevent shoot through of converter 11 . the control circuit also provides overcurrent protection . a current signal from current sensor 18 ( fig1 ) is provided to voltage divider / attenuator 87 to provide a measured current signal . the measured current signal is provided to the noninverting input of a comparator 88 . a current reference signal 81 is provided to the inverting input of comparator 88 . the output of comparator 88 is coupled to vco 85 . thus , if the measured current signal exceeds the current reference signal , vco 85 is set to its maximum . this increases the frequency of converter 11 and reduces the total current therein . by providing a substantial amount of hysteresis in comparator 88 , the output frequency of vco 85 will be lowered only gradually . the foregoing discloses a resonant power supply with multiple dc outputs but only a single power converter and a single transformer . by winding the output filter inductors on a sngle core , the multiple output voltages more nearly track one another . by layering the secondary windings , savings in size , weight and cost are achieved over bifilar windings . by placing a pair of resonant capacitors from each line to center tap , voltage transients across the rectifier diodes are reduced . a control circuit provides direct regulation of one of the output voltages and indirect regulation of the others by virtue of the close coupling of the transformer secondary windings and the close coupling of the output filter inductors . when desired , local series regulators may be used to improve the regulation of the indirectly regulated output voltages . while preferred embodiments of the present invention have been shown and described herein , it will be obvious that such embodiments are provided by way of example only . numerous variations , changes and substitutions will occur to those skilled in the art without departing from the invention herein . accordingly , it is intended that the invention be limited only by the spirit and scope of the appended claims .", "category": "Physics"}
|
Is the category the most suitable category for the given patent?
| 0.25 |
140baa44de85b479375b6ca4965a25a7d7d16b2bf86f08319b15a6f93dadfaf6
| 0.960938 | 0.236328 | 0.882813 | 0.172852 | 0.996094 | 0.429688 |
null |
{"category": "Electricity", "patent": "referring now to fig1 an apparatus for converting a supplied dc voltage to a plurality of regulated dc voltages v1 - v5 will be described . a dc to ac converter 11 supplies a variable frequency ac voltage to a transformer 20 . converter 11 may have any configuration known in the art , a half - bridge converter being shown in the drawing . converter 11 comprises a pair of semiconductor switching devices 16 and 17 , connected in series . switching devices 16 and 17 are shown as field - effect transistors ( fets ), although other devices such as insulated gate transistors ( igts ) or gate turn - off thyristors ( gtos ) may be used . converter 11 is connected to a supplied dc voltage . in this case , the dc voltage is obtained by rectifying an ac source 10 with a voltage doubler 9 comprised of diodes 12 and 13 and capacitors 14 and 15 . a voltage dobler is used to allow the half - bridge converter to provide an ac voltage with a magnitude equal to that of source 10 . on every other half cycle of source 10 , capacitor 14 is charged through diode 12 . capacitor 15 is charged through diode 13 on the other half cycles . a primary winding 21 of transformer 20 is connected between the junction of fets 16 and 17 and the junction of capacitors 14 and 15 . by alternately switching on fets 16 and 17 , an ac voltage having a variable frequency and an rms magnitude equal to the rms magnitude of supply 10 is supplied to primary winding 21 . the gates of fets 16 and 17 are connected to a control circuit which will be described below with reference to fig3 . converter 11 also includes a current sensor 18 for supplying a current signal to the control circuit . transformer 20 has a single secondary winding comprised of an even plurality of secondary winding segments 30 - 39 connected in series . the transformer secondary includes a center tap 40 which serves as ground for the dc outputs . preferably , secondary winding segments 30 - 39 are symmetrical about center tap 40 as to number of turns and wire gauge . as an alternate , electrically equivalent form of transformer 20 secondary , a plurality of separate , series - connected windings may be employed , with taps between adjacent separate windings 30 - 39 as shown in fig1 a . as used herein , the term &# 34 ; multiple secondary windings &# 34 ; is intended to encompass transformer secondaries of the type shown in both fig1 and fig1 a . a preferred construction for transformer 20 is shown in fig2 . primary winding 21 is would on one leg of core 25 multiple . secondary windings 30 - 39 are wound on another leg . to maintain tight coupling between the secondary windings multiple . secondary windings 30 - 39 are layered on top of each other separated by thin sheets of insulator 26 . returning to fig1 a pair of capacitors 41 and 42 are connected across the transformer secondary . the junction of cpacitors 41 and 42 is connected to center tap 40 . the leakage inductance resulting from the relatively loose coupling between primary winding 21 and multiple secondary windings 30 - 39 ( as a result of winding primary and secondaries on different legs of the core ) serves as a resonant inductor which resonates with capacitors 41 and 42 . an optional capacitor 43 may be included in the resonant circuit , directly across the transformer secondary to assist in tuning the resonant circuit . rectifying diodes 50 - 59 rectify the currents from secondary windings 30 - 39 , respectively , providing five dc voltages . the cathodes of diodes 50 - 59 are connected in pairs , the diode anodes in each pair being connected to opposite sides of center tap 40 . preferably , diodes 50 - 59 are connected in a symmetric configuration as in fig1 . the cathodes of each pair of diodes are connected to a separate filter , respectively . thus , a dc output voltage v1 is filtered by an inductor 60 and a capacitor 70 . each of inductors 60 - 64 provides a continuous current with minimum ripple to the load respectively connected thereto , thereby eliminating transient currents in transformer 20 and output capcitors 70 - 74 . inductors 60 - 64 are all wound on a single core and are tightly coupled magnetically . the number of turns of each inductor is proportional to each respective output voltage v1 - v5 . this winding arrangement improves the cross - regulation ( i . e . indirect regulation ) of the output voltages v1 - v5 which are not coupled to the control circuit , as described below . since multiple secondary windings 30 - 39 are layered , they are less closely coupled than bifilar windings . this loss in coupling tends to introduce transient voltage spikes across the rectifying diodes as current commutates from one half of the transformer secondary ( e . g . multiple secondary windings 30 - 34 ) to the other half ( e . g . multiple secondary windings 35 - 39 ). however , the placement of capacitors 41 and 42 from line to center tap ( common ) smoothes the current commutations . thus , resonant capacitors 41 and 42 act as lossless snubbers , reducing the voltage stresses and switching losses of diodes 50 - 59 . turning now to fig3 a control circuit is shown which regulates dc output voltages v1 - v5 ( fig1 ) by varying the output frequency of dc to ac converter 11 ( fig1 ). one of the dc output voltages , v2 for example , is provided to a voltage divider / attenuator 82 to provide a measured voltage which may be directly regulated by the control . the measured voltage is compared to a voltage reference 80 in a summer 83 , producing an error signal . the error signal is processed in a proportional - integral ( p - i ) controller 84 which in turn controls a voltage controlled oscillator ( vco ) 85 . preferably , the output of vco 85 is limited to frequencies above resonance of the resonant circuit . a driver circuit 86 is connected to vco 85 and to the gates of fets 16 and 17 . such driver circuits are known in the art . by way of example , driver circuit 86 may include a flip - flop for toggling by the output signal of vco 85 to provide two complementary signals . in addition , a lock - out circuit ( not shown ) in driver 86 may provide a minimum dead time between successive switchings to prevent shoot through of converter 11 . the control circuit also provides overcurrent protection . a current signal from current sensor 18 ( fig1 ) is provided to voltage divider / attenuator 87 to provide a measured current signal . the measured current signal is provided to the noninverting input of a comparator 88 . a current reference signal 81 is provided to the inverting input of comparator 88 . the output of comparator 88 is coupled to vco 85 . thus , if the measured current signal exceeds the current reference signal , vco 85 is set to its maximum . this increases the frequency of converter 11 and reduces the total current therein . by providing a substantial amount of hysteresis in comparator 88 , the output frequency of vco 85 will be lowered only gradually . the foregoing discloses a resonant power supply with multiple dc outputs but only a single power converter and a single transformer . by winding the output filter inductors on a sngle core , the multiple output voltages more nearly track one another . by layering the secondary windings , savings in size , weight and cost are achieved over bifilar windings . by placing a pair of resonant capacitors from each line to center tap , voltage transients across the rectifier diodes are reduced . a control circuit provides direct regulation of one of the output voltages and indirect regulation of the others by virtue of the close coupling of the transformer secondary windings and the close coupling of the output filter inductors . when desired , local series regulators may be used to improve the regulation of the indirectly regulated output voltages . while preferred embodiments of the present invention have been shown and described herein , it will be obvious that such embodiments are provided by way of example only . numerous variations , changes and substitutions will occur to those skilled in the art without departing from the invention herein . accordingly , it is intended that the invention be limited only by the spirit and scope of the appended claims ."}
|
{"category": "General tagging of new or cross-sectional technology", "patent": "referring now to fig1 an apparatus for converting a supplied dc voltage to a plurality of regulated dc voltages v1 - v5 will be described . a dc to ac converter 11 supplies a variable frequency ac voltage to a transformer 20 . converter 11 may have any configuration known in the art , a half - bridge converter being shown in the drawing . converter 11 comprises a pair of semiconductor switching devices 16 and 17 , connected in series . switching devices 16 and 17 are shown as field - effect transistors ( fets ), although other devices such as insulated gate transistors ( igts ) or gate turn - off thyristors ( gtos ) may be used . converter 11 is connected to a supplied dc voltage . in this case , the dc voltage is obtained by rectifying an ac source 10 with a voltage doubler 9 comprised of diodes 12 and 13 and capacitors 14 and 15 . a voltage dobler is used to allow the half - bridge converter to provide an ac voltage with a magnitude equal to that of source 10 . on every other half cycle of source 10 , capacitor 14 is charged through diode 12 . capacitor 15 is charged through diode 13 on the other half cycles . a primary winding 21 of transformer 20 is connected between the junction of fets 16 and 17 and the junction of capacitors 14 and 15 . by alternately switching on fets 16 and 17 , an ac voltage having a variable frequency and an rms magnitude equal to the rms magnitude of supply 10 is supplied to primary winding 21 . the gates of fets 16 and 17 are connected to a control circuit which will be described below with reference to fig3 . converter 11 also includes a current sensor 18 for supplying a current signal to the control circuit . transformer 20 has a single secondary winding comprised of an even plurality of secondary winding segments 30 - 39 connected in series . the transformer secondary includes a center tap 40 which serves as ground for the dc outputs . preferably , secondary winding segments 30 - 39 are symmetrical about center tap 40 as to number of turns and wire gauge . as an alternate , electrically equivalent form of transformer 20 secondary , a plurality of separate , series - connected windings may be employed , with taps between adjacent separate windings 30 - 39 as shown in fig1 a . as used herein , the term &# 34 ; multiple secondary windings &# 34 ; is intended to encompass transformer secondaries of the type shown in both fig1 and fig1 a . a preferred construction for transformer 20 is shown in fig2 . primary winding 21 is would on one leg of core 25 multiple . secondary windings 30 - 39 are wound on another leg . to maintain tight coupling between the secondary windings multiple . secondary windings 30 - 39 are layered on top of each other separated by thin sheets of insulator 26 . returning to fig1 a pair of capacitors 41 and 42 are connected across the transformer secondary . the junction of cpacitors 41 and 42 is connected to center tap 40 . the leakage inductance resulting from the relatively loose coupling between primary winding 21 and multiple secondary windings 30 - 39 ( as a result of winding primary and secondaries on different legs of the core ) serves as a resonant inductor which resonates with capacitors 41 and 42 . an optional capacitor 43 may be included in the resonant circuit , directly across the transformer secondary to assist in tuning the resonant circuit . rectifying diodes 50 - 59 rectify the currents from secondary windings 30 - 39 , respectively , providing five dc voltages . the cathodes of diodes 50 - 59 are connected in pairs , the diode anodes in each pair being connected to opposite sides of center tap 40 . preferably , diodes 50 - 59 are connected in a symmetric configuration as in fig1 . the cathodes of each pair of diodes are connected to a separate filter , respectively . thus , a dc output voltage v1 is filtered by an inductor 60 and a capacitor 70 . each of inductors 60 - 64 provides a continuous current with minimum ripple to the load respectively connected thereto , thereby eliminating transient currents in transformer 20 and output capcitors 70 - 74 . inductors 60 - 64 are all wound on a single core and are tightly coupled magnetically . the number of turns of each inductor is proportional to each respective output voltage v1 - v5 . this winding arrangement improves the cross - regulation ( i . e . indirect regulation ) of the output voltages v1 - v5 which are not coupled to the control circuit , as described below . since multiple secondary windings 30 - 39 are layered , they are less closely coupled than bifilar windings . this loss in coupling tends to introduce transient voltage spikes across the rectifying diodes as current commutates from one half of the transformer secondary ( e . g . multiple secondary windings 30 - 34 ) to the other half ( e . g . multiple secondary windings 35 - 39 ). however , the placement of capacitors 41 and 42 from line to center tap ( common ) smoothes the current commutations . thus , resonant capacitors 41 and 42 act as lossless snubbers , reducing the voltage stresses and switching losses of diodes 50 - 59 . turning now to fig3 a control circuit is shown which regulates dc output voltages v1 - v5 ( fig1 ) by varying the output frequency of dc to ac converter 11 ( fig1 ). one of the dc output voltages , v2 for example , is provided to a voltage divider / attenuator 82 to provide a measured voltage which may be directly regulated by the control . the measured voltage is compared to a voltage reference 80 in a summer 83 , producing an error signal . the error signal is processed in a proportional - integral ( p - i ) controller 84 which in turn controls a voltage controlled oscillator ( vco ) 85 . preferably , the output of vco 85 is limited to frequencies above resonance of the resonant circuit . a driver circuit 86 is connected to vco 85 and to the gates of fets 16 and 17 . such driver circuits are known in the art . by way of example , driver circuit 86 may include a flip - flop for toggling by the output signal of vco 85 to provide two complementary signals . in addition , a lock - out circuit ( not shown ) in driver 86 may provide a minimum dead time between successive switchings to prevent shoot through of converter 11 . the control circuit also provides overcurrent protection . a current signal from current sensor 18 ( fig1 ) is provided to voltage divider / attenuator 87 to provide a measured current signal . the measured current signal is provided to the noninverting input of a comparator 88 . a current reference signal 81 is provided to the inverting input of comparator 88 . the output of comparator 88 is coupled to vco 85 . thus , if the measured current signal exceeds the current reference signal , vco 85 is set to its maximum . this increases the frequency of converter 11 and reduces the total current therein . by providing a substantial amount of hysteresis in comparator 88 , the output frequency of vco 85 will be lowered only gradually . the foregoing discloses a resonant power supply with multiple dc outputs but only a single power converter and a single transformer . by winding the output filter inductors on a sngle core , the multiple output voltages more nearly track one another . by layering the secondary windings , savings in size , weight and cost are achieved over bifilar windings . by placing a pair of resonant capacitors from each line to center tap , voltage transients across the rectifier diodes are reduced . a control circuit provides direct regulation of one of the output voltages and indirect regulation of the others by virtue of the close coupling of the transformer secondary windings and the close coupling of the output filter inductors . when desired , local series regulators may be used to improve the regulation of the indirectly regulated output voltages . while preferred embodiments of the present invention have been shown and described herein , it will be obvious that such embodiments are provided by way of example only . numerous variations , changes and substitutions will occur to those skilled in the art without departing from the invention herein . accordingly , it is intended that the invention be limited only by the spirit and scope of the appended claims ."}
|
Does the patent belong in this category?
| 0.25 |
140baa44de85b479375b6ca4965a25a7d7d16b2bf86f08319b15a6f93dadfaf6
| 0.976563 | 0.482422 | 0.984375 | 0.703125 | 0.996094 | 0.648438 |
null |
{"category": "Physics", "patent": "for the purposes of promoting an understanding of the principles of the invention and presenting its currently understood best mode of operation , reference will now be made to the embodiments illustrated and specific language will be used to describe the same . it will nevertheless be understood that no limitation of the scope of the invention is thereby intended , with such alterations and further modifications in the illustrated device and such further applications of the principles of the invention as illustrated therein being contemplated as would normally occur to one skilled in the art to which the invention relates . the present novel technology was developed to improve urban response to ehe &# 39 ; s . consequently , the following examples and embodiments to reflect and reference a study area consisting of a large , urban area . this study environment was selected such that the large , urban area has experienced an extreme heat event . furthermore , this particular urban area was selected in part because for the time of the extreme heat event there was available associated population data and landsat tm imagery data . as such , ehe &# 39 ; s are naturally occurring and infrequent , and not inherently reproducible in a controlled laboratory environment , this study is frequently referenced herein . however , it should be kept in mind that the present novel technology is broadly applicable beyond the specific details and characteristics of the study embodiment referenced herein . census data are derived at the block group level following the socio - economic characteristics of vulnerability ( such as : hispanic population , black population , asian population , native american population , other race population , age 65 and over , age 65 and living below poverty , age 5 and under , population living below poverty , low education , and the like ) to extreme heat . estimates of population are derived by normalizing the total population by the area of residential land use within each block group . the area of residential land use within a block group may be determined through any number of methods including satellite imagery , aerial survey , and the like . the area of residential land use is typically selected over other possible values , such as total block group area , because it provides a truer indicator of residential density within each block group . this is to say that it provides a more accurate description of the residential density within each block group . however , any convenient relevant value may be chosen heat related fatalities for the area in question is obtained . the data is typically filtered to only include those deaths that occurred during the previously mentioned extreme heat event . the addresses of those qualified decedents are then assigned geographic identifies ( hereafter geocoded ). additionally , the deaths within each block group are totaled to produce a dataset representative of block group level ehe mortality . typically , the landsat thermal mapping ( hereafter tm ) imagery is acquired for the time period in question . the thermal band of the image is then converted to an at - satellite brightness temperature per the following equation , where t is an estimate of land surface temperature in kelvin . k2 is the calibration constant for temperature in kelvin and k1 is the constant for radiance in mwcm 2\u03bcm \u2212 1 . l w is the spectral radiance in mwcm 2 , calculated from the digital number values of the landsat tm thermal band . typically , t is then averaged by block group . this is done to determine the mean estimated land surface temperature per block group . the average t values are then uniformly stratified into the number of different levels desired . the typical spatial analysis method is the standard deviational ellipse ( sde ). it is a well known method and highly suited for point patterns . the end result of this analysis is the assignment of a weight to each of the descriptive variables . for example , in the most simple case , the variables found to be highly descriptive of the data are assigned a weight of 1 ( one ) while those not found so are assigned a weight of 0 ( zero ). the calculation of the sde is reasonably uncomplicated and many current gis applications allow for its use . the sde first requires that the centroid of each block group be calculated and the demographic measures , decedents and t measures respectively be assigned . typically , a weighted mean center of the point set will also be calculated as part of the calculation of the sde . use of the weighted mean center provides a better descriptor of vulnerability than a non - weighted mean center . the weighted mean center is obtained by averaging the coordinates of all the points and providing a weight to each based on an attribute variable of interest . after the mean center is calculated , each point is then transformed into a different metric space referenced from the mean center . the equation for this transformation is x \u2032 j = x j \u2212 x weighted mean center with the y transformation essentially being the same equation . the angle of rotation from the transformed points is calculated by the standard distance on x and y are then calculated as \u03b4 x =\u221a{ square root over (( \u03c3 i \u2212 1 n ( x \u2032 i cos \u03b4 y =\u221a{ square root over (( \u03c3 i \u2212 1 n ( x \u2032 i sin \u03b8 x , \u03b8 y , x weighted mean center , y weighted mean center , and area are used to quantitatively compare the spatial distributions of all the variables . standard t - test and f test are used to determine the levels of spatial similarity . this in turn indicates the importance of the actual variables when describing the real data . additional evaluations of the variables and their importance in describing the data are achieved through evaluation of concentration and eccentricity . a death concentration value is calculated within the spatial distributions of t . eccentricity values are calculated as \u03b8 x / \u03b8 y . the concentration value describes the level of concentration of the spatial phenomena and the eccentricity indicates the polarity of the point distribution within the ellipse . with respect to a variable , the greater the concentration of death and / or smaller eccentricity , the greater coverage of that variable &# 39 ; s descriptive capability with respect to the actual data . typically , a multiple regression modeling technique is used . first , all non - zero weighted variables are interpolated to standard sized cells covering the study area using a kernel density function . after calculation of the kernel density , the mean value per block group residential area is calculated . the non - zero weighted variables are evaluated for multi - collinearity . if needed , any collinearity is removed . a mapping of the kernel density of real death points ( the actual data ) is performed . a multiple regression utilizing the non - zero weighted variables and t as the independent variables is performed , with density of ehe death being the dependent variable . outputs of the regression are generated , forming standardized predictive values of risk . these values are mapped at the census - block group level with decedent locations as the validation layer . the standard r 2 test ( a test that determines what fraction of the total squared error is attributed to the model ) or the like may be used to determine the effectiveness of the model in explaining the variation of the dependent variable . following a significant r 2 value , the models outputs can be viewed as spatially predictive values of future risk . maps depicting spatial variation of risk , typically through a 3 - d map with the y axis representative of relative risk or the like of the city can be created . in the event of an extreme heat event or a predicted extreme heat event , health care professionals concentrate intervention measures into areas denoted as at high risk . while the invention has been illustrated and described in detail in the foregoing description , the same is to be considered as illustrative and not restrictive in character . it is understood that the embodiments have been shown and described in the foregoing specification in satisfaction of the best mode and enablement requirements . it is understood that one of ordinary skill in the art could readily make a nigh - infinite number of insubstantial changes and modifications to the above - described embodiments and that it would be impractical to attempt to describe all such embodiment variations in the present specification . accordingly , it is understood that all changes and modifications that come within the spirit of the invention are desired to be protected ."}
|
{"category": "Human Necessities", "patent": "for the purposes of promoting an understanding of the principles of the invention and presenting its currently understood best mode of operation , reference will now be made to the embodiments illustrated and specific language will be used to describe the same . it will nevertheless be understood that no limitation of the scope of the invention is thereby intended , with such alterations and further modifications in the illustrated device and such further applications of the principles of the invention as illustrated therein being contemplated as would normally occur to one skilled in the art to which the invention relates . the present novel technology was developed to improve urban response to ehe &# 39 ; s . consequently , the following examples and embodiments to reflect and reference a study area consisting of a large , urban area . this study environment was selected such that the large , urban area has experienced an extreme heat event . furthermore , this particular urban area was selected in part because for the time of the extreme heat event there was available associated population data and landsat tm imagery data . as such , ehe &# 39 ; s are naturally occurring and infrequent , and not inherently reproducible in a controlled laboratory environment , this study is frequently referenced herein . however , it should be kept in mind that the present novel technology is broadly applicable beyond the specific details and characteristics of the study embodiment referenced herein . census data are derived at the block group level following the socio - economic characteristics of vulnerability ( such as : hispanic population , black population , asian population , native american population , other race population , age 65 and over , age 65 and living below poverty , age 5 and under , population living below poverty , low education , and the like ) to extreme heat . estimates of population are derived by normalizing the total population by the area of residential land use within each block group . the area of residential land use within a block group may be determined through any number of methods including satellite imagery , aerial survey , and the like . the area of residential land use is typically selected over other possible values , such as total block group area , because it provides a truer indicator of residential density within each block group . this is to say that it provides a more accurate description of the residential density within each block group . however , any convenient relevant value may be chosen heat related fatalities for the area in question is obtained . the data is typically filtered to only include those deaths that occurred during the previously mentioned extreme heat event . the addresses of those qualified decedents are then assigned geographic identifies ( hereafter geocoded ). additionally , the deaths within each block group are totaled to produce a dataset representative of block group level ehe mortality . typically , the landsat thermal mapping ( hereafter tm ) imagery is acquired for the time period in question . the thermal band of the image is then converted to an at - satellite brightness temperature per the following equation , where t is an estimate of land surface temperature in kelvin . k2 is the calibration constant for temperature in kelvin and k1 is the constant for radiance in mwcm 2\u03bcm \u2212 1 . l w is the spectral radiance in mwcm 2 , calculated from the digital number values of the landsat tm thermal band . typically , t is then averaged by block group . this is done to determine the mean estimated land surface temperature per block group . the average t values are then uniformly stratified into the number of different levels desired . the typical spatial analysis method is the standard deviational ellipse ( sde ). it is a well known method and highly suited for point patterns . the end result of this analysis is the assignment of a weight to each of the descriptive variables . for example , in the most simple case , the variables found to be highly descriptive of the data are assigned a weight of 1 ( one ) while those not found so are assigned a weight of 0 ( zero ). the calculation of the sde is reasonably uncomplicated and many current gis applications allow for its use . the sde first requires that the centroid of each block group be calculated and the demographic measures , decedents and t measures respectively be assigned . typically , a weighted mean center of the point set will also be calculated as part of the calculation of the sde . use of the weighted mean center provides a better descriptor of vulnerability than a non - weighted mean center . the weighted mean center is obtained by averaging the coordinates of all the points and providing a weight to each based on an attribute variable of interest . after the mean center is calculated , each point is then transformed into a different metric space referenced from the mean center . the equation for this transformation is x \u2032 j = x j \u2212 x weighted mean center with the y transformation essentially being the same equation . the angle of rotation from the transformed points is calculated by the standard distance on x and y are then calculated as \u03b4 x =\u221a{ square root over (( \u03c3 i \u2212 1 n ( x \u2032 i cos \u03b4 y =\u221a{ square root over (( \u03c3 i \u2212 1 n ( x \u2032 i sin \u03b8 x , \u03b8 y , x weighted mean center , y weighted mean center , and area are used to quantitatively compare the spatial distributions of all the variables . standard t - test and f test are used to determine the levels of spatial similarity . this in turn indicates the importance of the actual variables when describing the real data . additional evaluations of the variables and their importance in describing the data are achieved through evaluation of concentration and eccentricity . a death concentration value is calculated within the spatial distributions of t . eccentricity values are calculated as \u03b8 x / \u03b8 y . the concentration value describes the level of concentration of the spatial phenomena and the eccentricity indicates the polarity of the point distribution within the ellipse . with respect to a variable , the greater the concentration of death and / or smaller eccentricity , the greater coverage of that variable &# 39 ; s descriptive capability with respect to the actual data . typically , a multiple regression modeling technique is used . first , all non - zero weighted variables are interpolated to standard sized cells covering the study area using a kernel density function . after calculation of the kernel density , the mean value per block group residential area is calculated . the non - zero weighted variables are evaluated for multi - collinearity . if needed , any collinearity is removed . a mapping of the kernel density of real death points ( the actual data ) is performed . a multiple regression utilizing the non - zero weighted variables and t as the independent variables is performed , with density of ehe death being the dependent variable . outputs of the regression are generated , forming standardized predictive values of risk . these values are mapped at the census - block group level with decedent locations as the validation layer . the standard r 2 test ( a test that determines what fraction of the total squared error is attributed to the model ) or the like may be used to determine the effectiveness of the model in explaining the variation of the dependent variable . following a significant r 2 value , the models outputs can be viewed as spatially predictive values of future risk . maps depicting spatial variation of risk , typically through a 3 - d map with the y axis representative of relative risk or the like of the city can be created . in the event of an extreme heat event or a predicted extreme heat event , health care professionals concentrate intervention measures into areas denoted as at high risk . while the invention has been illustrated and described in detail in the foregoing description , the same is to be considered as illustrative and not restrictive in character . it is understood that the embodiments have been shown and described in the foregoing specification in satisfaction of the best mode and enablement requirements . it is understood that one of ordinary skill in the art could readily make a nigh - infinite number of insubstantial changes and modifications to the above - described embodiments and that it would be impractical to attempt to describe all such embodiment variations in the present specification . accordingly , it is understood that all changes and modifications that come within the spirit of the invention are desired to be protected ."}
|
Is the categorization of this patent accurate?
| 0.25 |
29100f4f34af5f1b11afe606cb14cbdd00d6c6830767f9908c573993335e99fa
| 0.040771 | 0.030273 | 0.160156 | 0.053467 | 0.210938 | 0.0065 |
null |
{"patent": "for the purposes of promoting an understanding of the principles of the invention and presenting its currently understood best mode of operation , reference will now be made to the embodiments illustrated and specific language will be used to describe the same . it will nevertheless be understood that no limitation of the scope of the invention is thereby intended , with such alterations and further modifications in the illustrated device and such further applications of the principles of the invention as illustrated therein being contemplated as would normally occur to one skilled in the art to which the invention relates . the present novel technology was developed to improve urban response to ehe &# 39 ; s . consequently , the following examples and embodiments to reflect and reference a study area consisting of a large , urban area . this study environment was selected such that the large , urban area has experienced an extreme heat event . furthermore , this particular urban area was selected in part because for the time of the extreme heat event there was available associated population data and landsat tm imagery data . as such , ehe &# 39 ; s are naturally occurring and infrequent , and not inherently reproducible in a controlled laboratory environment , this study is frequently referenced herein . however , it should be kept in mind that the present novel technology is broadly applicable beyond the specific details and characteristics of the study embodiment referenced herein . census data are derived at the block group level following the socio - economic characteristics of vulnerability ( such as : hispanic population , black population , asian population , native american population , other race population , age 65 and over , age 65 and living below poverty , age 5 and under , population living below poverty , low education , and the like ) to extreme heat . estimates of population are derived by normalizing the total population by the area of residential land use within each block group . the area of residential land use within a block group may be determined through any number of methods including satellite imagery , aerial survey , and the like . the area of residential land use is typically selected over other possible values , such as total block group area , because it provides a truer indicator of residential density within each block group . this is to say that it provides a more accurate description of the residential density within each block group . however , any convenient relevant value may be chosen heat related fatalities for the area in question is obtained . the data is typically filtered to only include those deaths that occurred during the previously mentioned extreme heat event . the addresses of those qualified decedents are then assigned geographic identifies ( hereafter geocoded ). additionally , the deaths within each block group are totaled to produce a dataset representative of block group level ehe mortality . typically , the landsat thermal mapping ( hereafter tm ) imagery is acquired for the time period in question . the thermal band of the image is then converted to an at - satellite brightness temperature per the following equation , where t is an estimate of land surface temperature in kelvin . k2 is the calibration constant for temperature in kelvin and k1 is the constant for radiance in mwcm 2\u03bcm \u2212 1 . l w is the spectral radiance in mwcm 2 , calculated from the digital number values of the landsat tm thermal band . typically , t is then averaged by block group . this is done to determine the mean estimated land surface temperature per block group . the average t values are then uniformly stratified into the number of different levels desired . the typical spatial analysis method is the standard deviational ellipse ( sde ). it is a well known method and highly suited for point patterns . the end result of this analysis is the assignment of a weight to each of the descriptive variables . for example , in the most simple case , the variables found to be highly descriptive of the data are assigned a weight of 1 ( one ) while those not found so are assigned a weight of 0 ( zero ). the calculation of the sde is reasonably uncomplicated and many current gis applications allow for its use . the sde first requires that the centroid of each block group be calculated and the demographic measures , decedents and t measures respectively be assigned . typically , a weighted mean center of the point set will also be calculated as part of the calculation of the sde . use of the weighted mean center provides a better descriptor of vulnerability than a non - weighted mean center . the weighted mean center is obtained by averaging the coordinates of all the points and providing a weight to each based on an attribute variable of interest . after the mean center is calculated , each point is then transformed into a different metric space referenced from the mean center . the equation for this transformation is x \u2032 j = x j \u2212 x weighted mean center with the y transformation essentially being the same equation . the angle of rotation from the transformed points is calculated by the standard distance on x and y are then calculated as \u03b4 x =\u221a{ square root over (( \u03c3 i \u2212 1 n ( x \u2032 i cos \u03b4 y =\u221a{ square root over (( \u03c3 i \u2212 1 n ( x \u2032 i sin \u03b8 x , \u03b8 y , x weighted mean center , y weighted mean center , and area are used to quantitatively compare the spatial distributions of all the variables . standard t - test and f test are used to determine the levels of spatial similarity . this in turn indicates the importance of the actual variables when describing the real data . additional evaluations of the variables and their importance in describing the data are achieved through evaluation of concentration and eccentricity . a death concentration value is calculated within the spatial distributions of t . eccentricity values are calculated as \u03b8 x / \u03b8 y . the concentration value describes the level of concentration of the spatial phenomena and the eccentricity indicates the polarity of the point distribution within the ellipse . with respect to a variable , the greater the concentration of death and / or smaller eccentricity , the greater coverage of that variable &# 39 ; s descriptive capability with respect to the actual data . typically , a multiple regression modeling technique is used . first , all non - zero weighted variables are interpolated to standard sized cells covering the study area using a kernel density function . after calculation of the kernel density , the mean value per block group residential area is calculated . the non - zero weighted variables are evaluated for multi - collinearity . if needed , any collinearity is removed . a mapping of the kernel density of real death points ( the actual data ) is performed . a multiple regression utilizing the non - zero weighted variables and t as the independent variables is performed , with density of ehe death being the dependent variable . outputs of the regression are generated , forming standardized predictive values of risk . these values are mapped at the census - block group level with decedent locations as the validation layer . the standard r 2 test ( a test that determines what fraction of the total squared error is attributed to the model ) or the like may be used to determine the effectiveness of the model in explaining the variation of the dependent variable . following a significant r 2 value , the models outputs can be viewed as spatially predictive values of future risk . maps depicting spatial variation of risk , typically through a 3 - d map with the y axis representative of relative risk or the like of the city can be created . in the event of an extreme heat event or a predicted extreme heat event , health care professionals concentrate intervention measures into areas denoted as at high risk . while the invention has been illustrated and described in detail in the foregoing description , the same is to be considered as illustrative and not restrictive in character . it is understood that the embodiments have been shown and described in the foregoing specification in satisfaction of the best mode and enablement requirements . it is understood that one of ordinary skill in the art could readily make a nigh - infinite number of insubstantial changes and modifications to the above - described embodiments and that it would be impractical to attempt to describe all such embodiment variations in the present specification . accordingly , it is understood that all changes and modifications that come within the spirit of the invention are desired to be protected .", "category": "Physics"}
|
{"patent": "for the purposes of promoting an understanding of the principles of the invention and presenting its currently understood best mode of operation , reference will now be made to the embodiments illustrated and specific language will be used to describe the same . it will nevertheless be understood that no limitation of the scope of the invention is thereby intended , with such alterations and further modifications in the illustrated device and such further applications of the principles of the invention as illustrated therein being contemplated as would normally occur to one skilled in the art to which the invention relates . the present novel technology was developed to improve urban response to ehe &# 39 ; s . consequently , the following examples and embodiments to reflect and reference a study area consisting of a large , urban area . this study environment was selected such that the large , urban area has experienced an extreme heat event . furthermore , this particular urban area was selected in part because for the time of the extreme heat event there was available associated population data and landsat tm imagery data . as such , ehe &# 39 ; s are naturally occurring and infrequent , and not inherently reproducible in a controlled laboratory environment , this study is frequently referenced herein . however , it should be kept in mind that the present novel technology is broadly applicable beyond the specific details and characteristics of the study embodiment referenced herein . census data are derived at the block group level following the socio - economic characteristics of vulnerability ( such as : hispanic population , black population , asian population , native american population , other race population , age 65 and over , age 65 and living below poverty , age 5 and under , population living below poverty , low education , and the like ) to extreme heat . estimates of population are derived by normalizing the total population by the area of residential land use within each block group . the area of residential land use within a block group may be determined through any number of methods including satellite imagery , aerial survey , and the like . the area of residential land use is typically selected over other possible values , such as total block group area , because it provides a truer indicator of residential density within each block group . this is to say that it provides a more accurate description of the residential density within each block group . however , any convenient relevant value may be chosen heat related fatalities for the area in question is obtained . the data is typically filtered to only include those deaths that occurred during the previously mentioned extreme heat event . the addresses of those qualified decedents are then assigned geographic identifies ( hereafter geocoded ). additionally , the deaths within each block group are totaled to produce a dataset representative of block group level ehe mortality . typically , the landsat thermal mapping ( hereafter tm ) imagery is acquired for the time period in question . the thermal band of the image is then converted to an at - satellite brightness temperature per the following equation , where t is an estimate of land surface temperature in kelvin . k2 is the calibration constant for temperature in kelvin and k1 is the constant for radiance in mwcm 2\u03bcm \u2212 1 . l w is the spectral radiance in mwcm 2 , calculated from the digital number values of the landsat tm thermal band . typically , t is then averaged by block group . this is done to determine the mean estimated land surface temperature per block group . the average t values are then uniformly stratified into the number of different levels desired . the typical spatial analysis method is the standard deviational ellipse ( sde ). it is a well known method and highly suited for point patterns . the end result of this analysis is the assignment of a weight to each of the descriptive variables . for example , in the most simple case , the variables found to be highly descriptive of the data are assigned a weight of 1 ( one ) while those not found so are assigned a weight of 0 ( zero ). the calculation of the sde is reasonably uncomplicated and many current gis applications allow for its use . the sde first requires that the centroid of each block group be calculated and the demographic measures , decedents and t measures respectively be assigned . typically , a weighted mean center of the point set will also be calculated as part of the calculation of the sde . use of the weighted mean center provides a better descriptor of vulnerability than a non - weighted mean center . the weighted mean center is obtained by averaging the coordinates of all the points and providing a weight to each based on an attribute variable of interest . after the mean center is calculated , each point is then transformed into a different metric space referenced from the mean center . the equation for this transformation is x \u2032 j = x j \u2212 x weighted mean center with the y transformation essentially being the same equation . the angle of rotation from the transformed points is calculated by the standard distance on x and y are then calculated as \u03b4 x =\u221a{ square root over (( \u03c3 i \u2212 1 n ( x \u2032 i cos \u03b4 y =\u221a{ square root over (( \u03c3 i \u2212 1 n ( x \u2032 i sin \u03b8 x , \u03b8 y , x weighted mean center , y weighted mean center , and area are used to quantitatively compare the spatial distributions of all the variables . standard t - test and f test are used to determine the levels of spatial similarity . this in turn indicates the importance of the actual variables when describing the real data . additional evaluations of the variables and their importance in describing the data are achieved through evaluation of concentration and eccentricity . a death concentration value is calculated within the spatial distributions of t . eccentricity values are calculated as \u03b8 x / \u03b8 y . the concentration value describes the level of concentration of the spatial phenomena and the eccentricity indicates the polarity of the point distribution within the ellipse . with respect to a variable , the greater the concentration of death and / or smaller eccentricity , the greater coverage of that variable &# 39 ; s descriptive capability with respect to the actual data . typically , a multiple regression modeling technique is used . first , all non - zero weighted variables are interpolated to standard sized cells covering the study area using a kernel density function . after calculation of the kernel density , the mean value per block group residential area is calculated . the non - zero weighted variables are evaluated for multi - collinearity . if needed , any collinearity is removed . a mapping of the kernel density of real death points ( the actual data ) is performed . a multiple regression utilizing the non - zero weighted variables and t as the independent variables is performed , with density of ehe death being the dependent variable . outputs of the regression are generated , forming standardized predictive values of risk . these values are mapped at the census - block group level with decedent locations as the validation layer . the standard r 2 test ( a test that determines what fraction of the total squared error is attributed to the model ) or the like may be used to determine the effectiveness of the model in explaining the variation of the dependent variable . following a significant r 2 value , the models outputs can be viewed as spatially predictive values of future risk . maps depicting spatial variation of risk , typically through a 3 - d map with the y axis representative of relative risk or the like of the city can be created . in the event of an extreme heat event or a predicted extreme heat event , health care professionals concentrate intervention measures into areas denoted as at high risk . while the invention has been illustrated and described in detail in the foregoing description , the same is to be considered as illustrative and not restrictive in character . it is understood that the embodiments have been shown and described in the foregoing specification in satisfaction of the best mode and enablement requirements . it is understood that one of ordinary skill in the art could readily make a nigh - infinite number of insubstantial changes and modifications to the above - described embodiments and that it would be impractical to attempt to describe all such embodiment variations in the present specification . accordingly , it is understood that all changes and modifications that come within the spirit of the invention are desired to be protected .", "category": "Performing Operations; Transporting"}
|
Is the patent correctly categorized?
| 0.25 |
29100f4f34af5f1b11afe606cb14cbdd00d6c6830767f9908c573993335e99fa
| 0.035156 | 0.007111 | 0.020996 | 0.007111 | 0.041992 | 0.094238 |
null |
{"category": "Physics", "patent": "for the purposes of promoting an understanding of the principles of the invention and presenting its currently understood best mode of operation , reference will now be made to the embodiments illustrated and specific language will be used to describe the same . it will nevertheless be understood that no limitation of the scope of the invention is thereby intended , with such alterations and further modifications in the illustrated device and such further applications of the principles of the invention as illustrated therein being contemplated as would normally occur to one skilled in the art to which the invention relates . the present novel technology was developed to improve urban response to ehe &# 39 ; s . consequently , the following examples and embodiments to reflect and reference a study area consisting of a large , urban area . this study environment was selected such that the large , urban area has experienced an extreme heat event . furthermore , this particular urban area was selected in part because for the time of the extreme heat event there was available associated population data and landsat tm imagery data . as such , ehe &# 39 ; s are naturally occurring and infrequent , and not inherently reproducible in a controlled laboratory environment , this study is frequently referenced herein . however , it should be kept in mind that the present novel technology is broadly applicable beyond the specific details and characteristics of the study embodiment referenced herein . census data are derived at the block group level following the socio - economic characteristics of vulnerability ( such as : hispanic population , black population , asian population , native american population , other race population , age 65 and over , age 65 and living below poverty , age 5 and under , population living below poverty , low education , and the like ) to extreme heat . estimates of population are derived by normalizing the total population by the area of residential land use within each block group . the area of residential land use within a block group may be determined through any number of methods including satellite imagery , aerial survey , and the like . the area of residential land use is typically selected over other possible values , such as total block group area , because it provides a truer indicator of residential density within each block group . this is to say that it provides a more accurate description of the residential density within each block group . however , any convenient relevant value may be chosen heat related fatalities for the area in question is obtained . the data is typically filtered to only include those deaths that occurred during the previously mentioned extreme heat event . the addresses of those qualified decedents are then assigned geographic identifies ( hereafter geocoded ). additionally , the deaths within each block group are totaled to produce a dataset representative of block group level ehe mortality . typically , the landsat thermal mapping ( hereafter tm ) imagery is acquired for the time period in question . the thermal band of the image is then converted to an at - satellite brightness temperature per the following equation , where t is an estimate of land surface temperature in kelvin . k2 is the calibration constant for temperature in kelvin and k1 is the constant for radiance in mwcm 2\u03bcm \u2212 1 . l w is the spectral radiance in mwcm 2 , calculated from the digital number values of the landsat tm thermal band . typically , t is then averaged by block group . this is done to determine the mean estimated land surface temperature per block group . the average t values are then uniformly stratified into the number of different levels desired . the typical spatial analysis method is the standard deviational ellipse ( sde ). it is a well known method and highly suited for point patterns . the end result of this analysis is the assignment of a weight to each of the descriptive variables . for example , in the most simple case , the variables found to be highly descriptive of the data are assigned a weight of 1 ( one ) while those not found so are assigned a weight of 0 ( zero ). the calculation of the sde is reasonably uncomplicated and many current gis applications allow for its use . the sde first requires that the centroid of each block group be calculated and the demographic measures , decedents and t measures respectively be assigned . typically , a weighted mean center of the point set will also be calculated as part of the calculation of the sde . use of the weighted mean center provides a better descriptor of vulnerability than a non - weighted mean center . the weighted mean center is obtained by averaging the coordinates of all the points and providing a weight to each based on an attribute variable of interest . after the mean center is calculated , each point is then transformed into a different metric space referenced from the mean center . the equation for this transformation is x \u2032 j = x j \u2212 x weighted mean center with the y transformation essentially being the same equation . the angle of rotation from the transformed points is calculated by the standard distance on x and y are then calculated as \u03b4 x =\u221a{ square root over (( \u03c3 i \u2212 1 n ( x \u2032 i cos \u03b4 y =\u221a{ square root over (( \u03c3 i \u2212 1 n ( x \u2032 i sin \u03b8 x , \u03b8 y , x weighted mean center , y weighted mean center , and area are used to quantitatively compare the spatial distributions of all the variables . standard t - test and f test are used to determine the levels of spatial similarity . this in turn indicates the importance of the actual variables when describing the real data . additional evaluations of the variables and their importance in describing the data are achieved through evaluation of concentration and eccentricity . a death concentration value is calculated within the spatial distributions of t . eccentricity values are calculated as \u03b8 x / \u03b8 y . the concentration value describes the level of concentration of the spatial phenomena and the eccentricity indicates the polarity of the point distribution within the ellipse . with respect to a variable , the greater the concentration of death and / or smaller eccentricity , the greater coverage of that variable &# 39 ; s descriptive capability with respect to the actual data . typically , a multiple regression modeling technique is used . first , all non - zero weighted variables are interpolated to standard sized cells covering the study area using a kernel density function . after calculation of the kernel density , the mean value per block group residential area is calculated . the non - zero weighted variables are evaluated for multi - collinearity . if needed , any collinearity is removed . a mapping of the kernel density of real death points ( the actual data ) is performed . a multiple regression utilizing the non - zero weighted variables and t as the independent variables is performed , with density of ehe death being the dependent variable . outputs of the regression are generated , forming standardized predictive values of risk . these values are mapped at the census - block group level with decedent locations as the validation layer . the standard r 2 test ( a test that determines what fraction of the total squared error is attributed to the model ) or the like may be used to determine the effectiveness of the model in explaining the variation of the dependent variable . following a significant r 2 value , the models outputs can be viewed as spatially predictive values of future risk . maps depicting spatial variation of risk , typically through a 3 - d map with the y axis representative of relative risk or the like of the city can be created . in the event of an extreme heat event or a predicted extreme heat event , health care professionals concentrate intervention measures into areas denoted as at high risk . while the invention has been illustrated and described in detail in the foregoing description , the same is to be considered as illustrative and not restrictive in character . it is understood that the embodiments have been shown and described in the foregoing specification in satisfaction of the best mode and enablement requirements . it is understood that one of ordinary skill in the art could readily make a nigh - infinite number of insubstantial changes and modifications to the above - described embodiments and that it would be impractical to attempt to describe all such embodiment variations in the present specification . accordingly , it is understood that all changes and modifications that come within the spirit of the invention are desired to be protected ."}
|
{"patent": "for the purposes of promoting an understanding of the principles of the invention and presenting its currently understood best mode of operation , reference will now be made to the embodiments illustrated and specific language will be used to describe the same . it will nevertheless be understood that no limitation of the scope of the invention is thereby intended , with such alterations and further modifications in the illustrated device and such further applications of the principles of the invention as illustrated therein being contemplated as would normally occur to one skilled in the art to which the invention relates . the present novel technology was developed to improve urban response to ehe &# 39 ; s . consequently , the following examples and embodiments to reflect and reference a study area consisting of a large , urban area . this study environment was selected such that the large , urban area has experienced an extreme heat event . furthermore , this particular urban area was selected in part because for the time of the extreme heat event there was available associated population data and landsat tm imagery data . as such , ehe &# 39 ; s are naturally occurring and infrequent , and not inherently reproducible in a controlled laboratory environment , this study is frequently referenced herein . however , it should be kept in mind that the present novel technology is broadly applicable beyond the specific details and characteristics of the study embodiment referenced herein . census data are derived at the block group level following the socio - economic characteristics of vulnerability ( such as : hispanic population , black population , asian population , native american population , other race population , age 65 and over , age 65 and living below poverty , age 5 and under , population living below poverty , low education , and the like ) to extreme heat . estimates of population are derived by normalizing the total population by the area of residential land use within each block group . the area of residential land use within a block group may be determined through any number of methods including satellite imagery , aerial survey , and the like . the area of residential land use is typically selected over other possible values , such as total block group area , because it provides a truer indicator of residential density within each block group . this is to say that it provides a more accurate description of the residential density within each block group . however , any convenient relevant value may be chosen heat related fatalities for the area in question is obtained . the data is typically filtered to only include those deaths that occurred during the previously mentioned extreme heat event . the addresses of those qualified decedents are then assigned geographic identifies ( hereafter geocoded ). additionally , the deaths within each block group are totaled to produce a dataset representative of block group level ehe mortality . typically , the landsat thermal mapping ( hereafter tm ) imagery is acquired for the time period in question . the thermal band of the image is then converted to an at - satellite brightness temperature per the following equation , where t is an estimate of land surface temperature in kelvin . k2 is the calibration constant for temperature in kelvin and k1 is the constant for radiance in mwcm 2\u03bcm \u2212 1 . l w is the spectral radiance in mwcm 2 , calculated from the digital number values of the landsat tm thermal band . typically , t is then averaged by block group . this is done to determine the mean estimated land surface temperature per block group . the average t values are then uniformly stratified into the number of different levels desired . the typical spatial analysis method is the standard deviational ellipse ( sde ). it is a well known method and highly suited for point patterns . the end result of this analysis is the assignment of a weight to each of the descriptive variables . for example , in the most simple case , the variables found to be highly descriptive of the data are assigned a weight of 1 ( one ) while those not found so are assigned a weight of 0 ( zero ). the calculation of the sde is reasonably uncomplicated and many current gis applications allow for its use . the sde first requires that the centroid of each block group be calculated and the demographic measures , decedents and t measures respectively be assigned . typically , a weighted mean center of the point set will also be calculated as part of the calculation of the sde . use of the weighted mean center provides a better descriptor of vulnerability than a non - weighted mean center . the weighted mean center is obtained by averaging the coordinates of all the points and providing a weight to each based on an attribute variable of interest . after the mean center is calculated , each point is then transformed into a different metric space referenced from the mean center . the equation for this transformation is x \u2032 j = x j \u2212 x weighted mean center with the y transformation essentially being the same equation . the angle of rotation from the transformed points is calculated by the standard distance on x and y are then calculated as \u03b4 x =\u221a{ square root over (( \u03c3 i \u2212 1 n ( x \u2032 i cos \u03b4 y =\u221a{ square root over (( \u03c3 i \u2212 1 n ( x \u2032 i sin \u03b8 x , \u03b8 y , x weighted mean center , y weighted mean center , and area are used to quantitatively compare the spatial distributions of all the variables . standard t - test and f test are used to determine the levels of spatial similarity . this in turn indicates the importance of the actual variables when describing the real data . additional evaluations of the variables and their importance in describing the data are achieved through evaluation of concentration and eccentricity . a death concentration value is calculated within the spatial distributions of t . eccentricity values are calculated as \u03b8 x / \u03b8 y . the concentration value describes the level of concentration of the spatial phenomena and the eccentricity indicates the polarity of the point distribution within the ellipse . with respect to a variable , the greater the concentration of death and / or smaller eccentricity , the greater coverage of that variable &# 39 ; s descriptive capability with respect to the actual data . typically , a multiple regression modeling technique is used . first , all non - zero weighted variables are interpolated to standard sized cells covering the study area using a kernel density function . after calculation of the kernel density , the mean value per block group residential area is calculated . the non - zero weighted variables are evaluated for multi - collinearity . if needed , any collinearity is removed . a mapping of the kernel density of real death points ( the actual data ) is performed . a multiple regression utilizing the non - zero weighted variables and t as the independent variables is performed , with density of ehe death being the dependent variable . outputs of the regression are generated , forming standardized predictive values of risk . these values are mapped at the census - block group level with decedent locations as the validation layer . the standard r 2 test ( a test that determines what fraction of the total squared error is attributed to the model ) or the like may be used to determine the effectiveness of the model in explaining the variation of the dependent variable . following a significant r 2 value , the models outputs can be viewed as spatially predictive values of future risk . maps depicting spatial variation of risk , typically through a 3 - d map with the y axis representative of relative risk or the like of the city can be created . in the event of an extreme heat event or a predicted extreme heat event , health care professionals concentrate intervention measures into areas denoted as at high risk . while the invention has been illustrated and described in detail in the foregoing description , the same is to be considered as illustrative and not restrictive in character . it is understood that the embodiments have been shown and described in the foregoing specification in satisfaction of the best mode and enablement requirements . it is understood that one of ordinary skill in the art could readily make a nigh - infinite number of insubstantial changes and modifications to the above - described embodiments and that it would be impractical to attempt to describe all such embodiment variations in the present specification . accordingly , it is understood that all changes and modifications that come within the spirit of the invention are desired to be protected .", "category": "Chemistry; Metallurgy"}
|
Is the category the most suitable category for the given patent?
| 0.25 |
29100f4f34af5f1b11afe606cb14cbdd00d6c6830767f9908c573993335e99fa
| 0.032471 | 0.000179 | 0.041992 | 0.001205 | 0.392578 | 0.001282 |
null |
{"category": "Physics", "patent": "for the purposes of promoting an understanding of the principles of the invention and presenting its currently understood best mode of operation , reference will now be made to the embodiments illustrated and specific language will be used to describe the same . it will nevertheless be understood that no limitation of the scope of the invention is thereby intended , with such alterations and further modifications in the illustrated device and such further applications of the principles of the invention as illustrated therein being contemplated as would normally occur to one skilled in the art to which the invention relates . the present novel technology was developed to improve urban response to ehe &# 39 ; s . consequently , the following examples and embodiments to reflect and reference a study area consisting of a large , urban area . this study environment was selected such that the large , urban area has experienced an extreme heat event . furthermore , this particular urban area was selected in part because for the time of the extreme heat event there was available associated population data and landsat tm imagery data . as such , ehe &# 39 ; s are naturally occurring and infrequent , and not inherently reproducible in a controlled laboratory environment , this study is frequently referenced herein . however , it should be kept in mind that the present novel technology is broadly applicable beyond the specific details and characteristics of the study embodiment referenced herein . census data are derived at the block group level following the socio - economic characteristics of vulnerability ( such as : hispanic population , black population , asian population , native american population , other race population , age 65 and over , age 65 and living below poverty , age 5 and under , population living below poverty , low education , and the like ) to extreme heat . estimates of population are derived by normalizing the total population by the area of residential land use within each block group . the area of residential land use within a block group may be determined through any number of methods including satellite imagery , aerial survey , and the like . the area of residential land use is typically selected over other possible values , such as total block group area , because it provides a truer indicator of residential density within each block group . this is to say that it provides a more accurate description of the residential density within each block group . however , any convenient relevant value may be chosen heat related fatalities for the area in question is obtained . the data is typically filtered to only include those deaths that occurred during the previously mentioned extreme heat event . the addresses of those qualified decedents are then assigned geographic identifies ( hereafter geocoded ). additionally , the deaths within each block group are totaled to produce a dataset representative of block group level ehe mortality . typically , the landsat thermal mapping ( hereafter tm ) imagery is acquired for the time period in question . the thermal band of the image is then converted to an at - satellite brightness temperature per the following equation , where t is an estimate of land surface temperature in kelvin . k2 is the calibration constant for temperature in kelvin and k1 is the constant for radiance in mwcm 2\u03bcm \u2212 1 . l w is the spectral radiance in mwcm 2 , calculated from the digital number values of the landsat tm thermal band . typically , t is then averaged by block group . this is done to determine the mean estimated land surface temperature per block group . the average t values are then uniformly stratified into the number of different levels desired . the typical spatial analysis method is the standard deviational ellipse ( sde ). it is a well known method and highly suited for point patterns . the end result of this analysis is the assignment of a weight to each of the descriptive variables . for example , in the most simple case , the variables found to be highly descriptive of the data are assigned a weight of 1 ( one ) while those not found so are assigned a weight of 0 ( zero ). the calculation of the sde is reasonably uncomplicated and many current gis applications allow for its use . the sde first requires that the centroid of each block group be calculated and the demographic measures , decedents and t measures respectively be assigned . typically , a weighted mean center of the point set will also be calculated as part of the calculation of the sde . use of the weighted mean center provides a better descriptor of vulnerability than a non - weighted mean center . the weighted mean center is obtained by averaging the coordinates of all the points and providing a weight to each based on an attribute variable of interest . after the mean center is calculated , each point is then transformed into a different metric space referenced from the mean center . the equation for this transformation is x \u2032 j = x j \u2212 x weighted mean center with the y transformation essentially being the same equation . the angle of rotation from the transformed points is calculated by the standard distance on x and y are then calculated as \u03b4 x =\u221a{ square root over (( \u03c3 i \u2212 1 n ( x \u2032 i cos \u03b4 y =\u221a{ square root over (( \u03c3 i \u2212 1 n ( x \u2032 i sin \u03b8 x , \u03b8 y , x weighted mean center , y weighted mean center , and area are used to quantitatively compare the spatial distributions of all the variables . standard t - test and f test are used to determine the levels of spatial similarity . this in turn indicates the importance of the actual variables when describing the real data . additional evaluations of the variables and their importance in describing the data are achieved through evaluation of concentration and eccentricity . a death concentration value is calculated within the spatial distributions of t . eccentricity values are calculated as \u03b8 x / \u03b8 y . the concentration value describes the level of concentration of the spatial phenomena and the eccentricity indicates the polarity of the point distribution within the ellipse . with respect to a variable , the greater the concentration of death and / or smaller eccentricity , the greater coverage of that variable &# 39 ; s descriptive capability with respect to the actual data . typically , a multiple regression modeling technique is used . first , all non - zero weighted variables are interpolated to standard sized cells covering the study area using a kernel density function . after calculation of the kernel density , the mean value per block group residential area is calculated . the non - zero weighted variables are evaluated for multi - collinearity . if needed , any collinearity is removed . a mapping of the kernel density of real death points ( the actual data ) is performed . a multiple regression utilizing the non - zero weighted variables and t as the independent variables is performed , with density of ehe death being the dependent variable . outputs of the regression are generated , forming standardized predictive values of risk . these values are mapped at the census - block group level with decedent locations as the validation layer . the standard r 2 test ( a test that determines what fraction of the total squared error is attributed to the model ) or the like may be used to determine the effectiveness of the model in explaining the variation of the dependent variable . following a significant r 2 value , the models outputs can be viewed as spatially predictive values of future risk . maps depicting spatial variation of risk , typically through a 3 - d map with the y axis representative of relative risk or the like of the city can be created . in the event of an extreme heat event or a predicted extreme heat event , health care professionals concentrate intervention measures into areas denoted as at high risk . while the invention has been illustrated and described in detail in the foregoing description , the same is to be considered as illustrative and not restrictive in character . it is understood that the embodiments have been shown and described in the foregoing specification in satisfaction of the best mode and enablement requirements . it is understood that one of ordinary skill in the art could readily make a nigh - infinite number of insubstantial changes and modifications to the above - described embodiments and that it would be impractical to attempt to describe all such embodiment variations in the present specification . accordingly , it is understood that all changes and modifications that come within the spirit of the invention are desired to be protected ."}
|
{"category": "Textiles; Paper", "patent": "for the purposes of promoting an understanding of the principles of the invention and presenting its currently understood best mode of operation , reference will now be made to the embodiments illustrated and specific language will be used to describe the same . it will nevertheless be understood that no limitation of the scope of the invention is thereby intended , with such alterations and further modifications in the illustrated device and such further applications of the principles of the invention as illustrated therein being contemplated as would normally occur to one skilled in the art to which the invention relates . the present novel technology was developed to improve urban response to ehe &# 39 ; s . consequently , the following examples and embodiments to reflect and reference a study area consisting of a large , urban area . this study environment was selected such that the large , urban area has experienced an extreme heat event . furthermore , this particular urban area was selected in part because for the time of the extreme heat event there was available associated population data and landsat tm imagery data . as such , ehe &# 39 ; s are naturally occurring and infrequent , and not inherently reproducible in a controlled laboratory environment , this study is frequently referenced herein . however , it should be kept in mind that the present novel technology is broadly applicable beyond the specific details and characteristics of the study embodiment referenced herein . census data are derived at the block group level following the socio - economic characteristics of vulnerability ( such as : hispanic population , black population , asian population , native american population , other race population , age 65 and over , age 65 and living below poverty , age 5 and under , population living below poverty , low education , and the like ) to extreme heat . estimates of population are derived by normalizing the total population by the area of residential land use within each block group . the area of residential land use within a block group may be determined through any number of methods including satellite imagery , aerial survey , and the like . the area of residential land use is typically selected over other possible values , such as total block group area , because it provides a truer indicator of residential density within each block group . this is to say that it provides a more accurate description of the residential density within each block group . however , any convenient relevant value may be chosen heat related fatalities for the area in question is obtained . the data is typically filtered to only include those deaths that occurred during the previously mentioned extreme heat event . the addresses of those qualified decedents are then assigned geographic identifies ( hereafter geocoded ). additionally , the deaths within each block group are totaled to produce a dataset representative of block group level ehe mortality . typically , the landsat thermal mapping ( hereafter tm ) imagery is acquired for the time period in question . the thermal band of the image is then converted to an at - satellite brightness temperature per the following equation , where t is an estimate of land surface temperature in kelvin . k2 is the calibration constant for temperature in kelvin and k1 is the constant for radiance in mwcm 2\u03bcm \u2212 1 . l w is the spectral radiance in mwcm 2 , calculated from the digital number values of the landsat tm thermal band . typically , t is then averaged by block group . this is done to determine the mean estimated land surface temperature per block group . the average t values are then uniformly stratified into the number of different levels desired . the typical spatial analysis method is the standard deviational ellipse ( sde ). it is a well known method and highly suited for point patterns . the end result of this analysis is the assignment of a weight to each of the descriptive variables . for example , in the most simple case , the variables found to be highly descriptive of the data are assigned a weight of 1 ( one ) while those not found so are assigned a weight of 0 ( zero ). the calculation of the sde is reasonably uncomplicated and many current gis applications allow for its use . the sde first requires that the centroid of each block group be calculated and the demographic measures , decedents and t measures respectively be assigned . typically , a weighted mean center of the point set will also be calculated as part of the calculation of the sde . use of the weighted mean center provides a better descriptor of vulnerability than a non - weighted mean center . the weighted mean center is obtained by averaging the coordinates of all the points and providing a weight to each based on an attribute variable of interest . after the mean center is calculated , each point is then transformed into a different metric space referenced from the mean center . the equation for this transformation is x \u2032 j = x j \u2212 x weighted mean center with the y transformation essentially being the same equation . the angle of rotation from the transformed points is calculated by the standard distance on x and y are then calculated as \u03b4 x =\u221a{ square root over (( \u03c3 i \u2212 1 n ( x \u2032 i cos \u03b4 y =\u221a{ square root over (( \u03c3 i \u2212 1 n ( x \u2032 i sin \u03b8 x , \u03b8 y , x weighted mean center , y weighted mean center , and area are used to quantitatively compare the spatial distributions of all the variables . standard t - test and f test are used to determine the levels of spatial similarity . this in turn indicates the importance of the actual variables when describing the real data . additional evaluations of the variables and their importance in describing the data are achieved through evaluation of concentration and eccentricity . a death concentration value is calculated within the spatial distributions of t . eccentricity values are calculated as \u03b8 x / \u03b8 y . the concentration value describes the level of concentration of the spatial phenomena and the eccentricity indicates the polarity of the point distribution within the ellipse . with respect to a variable , the greater the concentration of death and / or smaller eccentricity , the greater coverage of that variable &# 39 ; s descriptive capability with respect to the actual data . typically , a multiple regression modeling technique is used . first , all non - zero weighted variables are interpolated to standard sized cells covering the study area using a kernel density function . after calculation of the kernel density , the mean value per block group residential area is calculated . the non - zero weighted variables are evaluated for multi - collinearity . if needed , any collinearity is removed . a mapping of the kernel density of real death points ( the actual data ) is performed . a multiple regression utilizing the non - zero weighted variables and t as the independent variables is performed , with density of ehe death being the dependent variable . outputs of the regression are generated , forming standardized predictive values of risk . these values are mapped at the census - block group level with decedent locations as the validation layer . the standard r 2 test ( a test that determines what fraction of the total squared error is attributed to the model ) or the like may be used to determine the effectiveness of the model in explaining the variation of the dependent variable . following a significant r 2 value , the models outputs can be viewed as spatially predictive values of future risk . maps depicting spatial variation of risk , typically through a 3 - d map with the y axis representative of relative risk or the like of the city can be created . in the event of an extreme heat event or a predicted extreme heat event , health care professionals concentrate intervention measures into areas denoted as at high risk . while the invention has been illustrated and described in detail in the foregoing description , the same is to be considered as illustrative and not restrictive in character . it is understood that the embodiments have been shown and described in the foregoing specification in satisfaction of the best mode and enablement requirements . it is understood that one of ordinary skill in the art could readily make a nigh - infinite number of insubstantial changes and modifications to the above - described embodiments and that it would be impractical to attempt to describe all such embodiment variations in the present specification . accordingly , it is understood that all changes and modifications that come within the spirit of the invention are desired to be protected ."}
|
Does the category match the content of the patent?
| 0.25 |
29100f4f34af5f1b11afe606cb14cbdd00d6c6830767f9908c573993335e99fa
| 0.150391 | 0.015442 | 0.400391 | 0.001366 | 0.5 | 0.001457 |
null |
{"category": "Physics", "patent": "for the purposes of promoting an understanding of the principles of the invention and presenting its currently understood best mode of operation , reference will now be made to the embodiments illustrated and specific language will be used to describe the same . it will nevertheless be understood that no limitation of the scope of the invention is thereby intended , with such alterations and further modifications in the illustrated device and such further applications of the principles of the invention as illustrated therein being contemplated as would normally occur to one skilled in the art to which the invention relates . the present novel technology was developed to improve urban response to ehe &# 39 ; s . consequently , the following examples and embodiments to reflect and reference a study area consisting of a large , urban area . this study environment was selected such that the large , urban area has experienced an extreme heat event . furthermore , this particular urban area was selected in part because for the time of the extreme heat event there was available associated population data and landsat tm imagery data . as such , ehe &# 39 ; s are naturally occurring and infrequent , and not inherently reproducible in a controlled laboratory environment , this study is frequently referenced herein . however , it should be kept in mind that the present novel technology is broadly applicable beyond the specific details and characteristics of the study embodiment referenced herein . census data are derived at the block group level following the socio - economic characteristics of vulnerability ( such as : hispanic population , black population , asian population , native american population , other race population , age 65 and over , age 65 and living below poverty , age 5 and under , population living below poverty , low education , and the like ) to extreme heat . estimates of population are derived by normalizing the total population by the area of residential land use within each block group . the area of residential land use within a block group may be determined through any number of methods including satellite imagery , aerial survey , and the like . the area of residential land use is typically selected over other possible values , such as total block group area , because it provides a truer indicator of residential density within each block group . this is to say that it provides a more accurate description of the residential density within each block group . however , any convenient relevant value may be chosen heat related fatalities for the area in question is obtained . the data is typically filtered to only include those deaths that occurred during the previously mentioned extreme heat event . the addresses of those qualified decedents are then assigned geographic identifies ( hereafter geocoded ). additionally , the deaths within each block group are totaled to produce a dataset representative of block group level ehe mortality . typically , the landsat thermal mapping ( hereafter tm ) imagery is acquired for the time period in question . the thermal band of the image is then converted to an at - satellite brightness temperature per the following equation , where t is an estimate of land surface temperature in kelvin . k2 is the calibration constant for temperature in kelvin and k1 is the constant for radiance in mwcm 2\u03bcm \u2212 1 . l w is the spectral radiance in mwcm 2 , calculated from the digital number values of the landsat tm thermal band . typically , t is then averaged by block group . this is done to determine the mean estimated land surface temperature per block group . the average t values are then uniformly stratified into the number of different levels desired . the typical spatial analysis method is the standard deviational ellipse ( sde ). it is a well known method and highly suited for point patterns . the end result of this analysis is the assignment of a weight to each of the descriptive variables . for example , in the most simple case , the variables found to be highly descriptive of the data are assigned a weight of 1 ( one ) while those not found so are assigned a weight of 0 ( zero ). the calculation of the sde is reasonably uncomplicated and many current gis applications allow for its use . the sde first requires that the centroid of each block group be calculated and the demographic measures , decedents and t measures respectively be assigned . typically , a weighted mean center of the point set will also be calculated as part of the calculation of the sde . use of the weighted mean center provides a better descriptor of vulnerability than a non - weighted mean center . the weighted mean center is obtained by averaging the coordinates of all the points and providing a weight to each based on an attribute variable of interest . after the mean center is calculated , each point is then transformed into a different metric space referenced from the mean center . the equation for this transformation is x \u2032 j = x j \u2212 x weighted mean center with the y transformation essentially being the same equation . the angle of rotation from the transformed points is calculated by the standard distance on x and y are then calculated as \u03b4 x =\u221a{ square root over (( \u03c3 i \u2212 1 n ( x \u2032 i cos \u03b4 y =\u221a{ square root over (( \u03c3 i \u2212 1 n ( x \u2032 i sin \u03b8 x , \u03b8 y , x weighted mean center , y weighted mean center , and area are used to quantitatively compare the spatial distributions of all the variables . standard t - test and f test are used to determine the levels of spatial similarity . this in turn indicates the importance of the actual variables when describing the real data . additional evaluations of the variables and their importance in describing the data are achieved through evaluation of concentration and eccentricity . a death concentration value is calculated within the spatial distributions of t . eccentricity values are calculated as \u03b8 x / \u03b8 y . the concentration value describes the level of concentration of the spatial phenomena and the eccentricity indicates the polarity of the point distribution within the ellipse . with respect to a variable , the greater the concentration of death and / or smaller eccentricity , the greater coverage of that variable &# 39 ; s descriptive capability with respect to the actual data . typically , a multiple regression modeling technique is used . first , all non - zero weighted variables are interpolated to standard sized cells covering the study area using a kernel density function . after calculation of the kernel density , the mean value per block group residential area is calculated . the non - zero weighted variables are evaluated for multi - collinearity . if needed , any collinearity is removed . a mapping of the kernel density of real death points ( the actual data ) is performed . a multiple regression utilizing the non - zero weighted variables and t as the independent variables is performed , with density of ehe death being the dependent variable . outputs of the regression are generated , forming standardized predictive values of risk . these values are mapped at the census - block group level with decedent locations as the validation layer . the standard r 2 test ( a test that determines what fraction of the total squared error is attributed to the model ) or the like may be used to determine the effectiveness of the model in explaining the variation of the dependent variable . following a significant r 2 value , the models outputs can be viewed as spatially predictive values of future risk . maps depicting spatial variation of risk , typically through a 3 - d map with the y axis representative of relative risk or the like of the city can be created . in the event of an extreme heat event or a predicted extreme heat event , health care professionals concentrate intervention measures into areas denoted as at high risk . while the invention has been illustrated and described in detail in the foregoing description , the same is to be considered as illustrative and not restrictive in character . it is understood that the embodiments have been shown and described in the foregoing specification in satisfaction of the best mode and enablement requirements . it is understood that one of ordinary skill in the art could readily make a nigh - infinite number of insubstantial changes and modifications to the above - described embodiments and that it would be impractical to attempt to describe all such embodiment variations in the present specification . accordingly , it is understood that all changes and modifications that come within the spirit of the invention are desired to be protected ."}
|
{"patent": "for the purposes of promoting an understanding of the principles of the invention and presenting its currently understood best mode of operation , reference will now be made to the embodiments illustrated and specific language will be used to describe the same . it will nevertheless be understood that no limitation of the scope of the invention is thereby intended , with such alterations and further modifications in the illustrated device and such further applications of the principles of the invention as illustrated therein being contemplated as would normally occur to one skilled in the art to which the invention relates . the present novel technology was developed to improve urban response to ehe &# 39 ; s . consequently , the following examples and embodiments to reflect and reference a study area consisting of a large , urban area . this study environment was selected such that the large , urban area has experienced an extreme heat event . furthermore , this particular urban area was selected in part because for the time of the extreme heat event there was available associated population data and landsat tm imagery data . as such , ehe &# 39 ; s are naturally occurring and infrequent , and not inherently reproducible in a controlled laboratory environment , this study is frequently referenced herein . however , it should be kept in mind that the present novel technology is broadly applicable beyond the specific details and characteristics of the study embodiment referenced herein . census data are derived at the block group level following the socio - economic characteristics of vulnerability ( such as : hispanic population , black population , asian population , native american population , other race population , age 65 and over , age 65 and living below poverty , age 5 and under , population living below poverty , low education , and the like ) to extreme heat . estimates of population are derived by normalizing the total population by the area of residential land use within each block group . the area of residential land use within a block group may be determined through any number of methods including satellite imagery , aerial survey , and the like . the area of residential land use is typically selected over other possible values , such as total block group area , because it provides a truer indicator of residential density within each block group . this is to say that it provides a more accurate description of the residential density within each block group . however , any convenient relevant value may be chosen heat related fatalities for the area in question is obtained . the data is typically filtered to only include those deaths that occurred during the previously mentioned extreme heat event . the addresses of those qualified decedents are then assigned geographic identifies ( hereafter geocoded ). additionally , the deaths within each block group are totaled to produce a dataset representative of block group level ehe mortality . typically , the landsat thermal mapping ( hereafter tm ) imagery is acquired for the time period in question . the thermal band of the image is then converted to an at - satellite brightness temperature per the following equation , where t is an estimate of land surface temperature in kelvin . k2 is the calibration constant for temperature in kelvin and k1 is the constant for radiance in mwcm 2\u03bcm \u2212 1 . l w is the spectral radiance in mwcm 2 , calculated from the digital number values of the landsat tm thermal band . typically , t is then averaged by block group . this is done to determine the mean estimated land surface temperature per block group . the average t values are then uniformly stratified into the number of different levels desired . the typical spatial analysis method is the standard deviational ellipse ( sde ). it is a well known method and highly suited for point patterns . the end result of this analysis is the assignment of a weight to each of the descriptive variables . for example , in the most simple case , the variables found to be highly descriptive of the data are assigned a weight of 1 ( one ) while those not found so are assigned a weight of 0 ( zero ). the calculation of the sde is reasonably uncomplicated and many current gis applications allow for its use . the sde first requires that the centroid of each block group be calculated and the demographic measures , decedents and t measures respectively be assigned . typically , a weighted mean center of the point set will also be calculated as part of the calculation of the sde . use of the weighted mean center provides a better descriptor of vulnerability than a non - weighted mean center . the weighted mean center is obtained by averaging the coordinates of all the points and providing a weight to each based on an attribute variable of interest . after the mean center is calculated , each point is then transformed into a different metric space referenced from the mean center . the equation for this transformation is x \u2032 j = x j \u2212 x weighted mean center with the y transformation essentially being the same equation . the angle of rotation from the transformed points is calculated by the standard distance on x and y are then calculated as \u03b4 x =\u221a{ square root over (( \u03c3 i \u2212 1 n ( x \u2032 i cos \u03b4 y =\u221a{ square root over (( \u03c3 i \u2212 1 n ( x \u2032 i sin \u03b8 x , \u03b8 y , x weighted mean center , y weighted mean center , and area are used to quantitatively compare the spatial distributions of all the variables . standard t - test and f test are used to determine the levels of spatial similarity . this in turn indicates the importance of the actual variables when describing the real data . additional evaluations of the variables and their importance in describing the data are achieved through evaluation of concentration and eccentricity . a death concentration value is calculated within the spatial distributions of t . eccentricity values are calculated as \u03b8 x / \u03b8 y . the concentration value describes the level of concentration of the spatial phenomena and the eccentricity indicates the polarity of the point distribution within the ellipse . with respect to a variable , the greater the concentration of death and / or smaller eccentricity , the greater coverage of that variable &# 39 ; s descriptive capability with respect to the actual data . typically , a multiple regression modeling technique is used . first , all non - zero weighted variables are interpolated to standard sized cells covering the study area using a kernel density function . after calculation of the kernel density , the mean value per block group residential area is calculated . the non - zero weighted variables are evaluated for multi - collinearity . if needed , any collinearity is removed . a mapping of the kernel density of real death points ( the actual data ) is performed . a multiple regression utilizing the non - zero weighted variables and t as the independent variables is performed , with density of ehe death being the dependent variable . outputs of the regression are generated , forming standardized predictive values of risk . these values are mapped at the census - block group level with decedent locations as the validation layer . the standard r 2 test ( a test that determines what fraction of the total squared error is attributed to the model ) or the like may be used to determine the effectiveness of the model in explaining the variation of the dependent variable . following a significant r 2 value , the models outputs can be viewed as spatially predictive values of future risk . maps depicting spatial variation of risk , typically through a 3 - d map with the y axis representative of relative risk or the like of the city can be created . in the event of an extreme heat event or a predicted extreme heat event , health care professionals concentrate intervention measures into areas denoted as at high risk . while the invention has been illustrated and described in detail in the foregoing description , the same is to be considered as illustrative and not restrictive in character . it is understood that the embodiments have been shown and described in the foregoing specification in satisfaction of the best mode and enablement requirements . it is understood that one of ordinary skill in the art could readily make a nigh - infinite number of insubstantial changes and modifications to the above - described embodiments and that it would be impractical to attempt to describe all such embodiment variations in the present specification . accordingly , it is understood that all changes and modifications that come within the spirit of the invention are desired to be protected .", "category": "Fixed Constructions"}
|
Is the category the most suitable category for the given patent?
| 0.25 |
29100f4f34af5f1b11afe606cb14cbdd00d6c6830767f9908c573993335e99fa
| 0.032471 | 0.126953 | 0.041992 | 0.028931 | 0.392578 | 0.269531 |
null |
{"patent": "for the purposes of promoting an understanding of the principles of the invention and presenting its currently understood best mode of operation , reference will now be made to the embodiments illustrated and specific language will be used to describe the same . it will nevertheless be understood that no limitation of the scope of the invention is thereby intended , with such alterations and further modifications in the illustrated device and such further applications of the principles of the invention as illustrated therein being contemplated as would normally occur to one skilled in the art to which the invention relates . the present novel technology was developed to improve urban response to ehe &# 39 ; s . consequently , the following examples and embodiments to reflect and reference a study area consisting of a large , urban area . this study environment was selected such that the large , urban area has experienced an extreme heat event . furthermore , this particular urban area was selected in part because for the time of the extreme heat event there was available associated population data and landsat tm imagery data . as such , ehe &# 39 ; s are naturally occurring and infrequent , and not inherently reproducible in a controlled laboratory environment , this study is frequently referenced herein . however , it should be kept in mind that the present novel technology is broadly applicable beyond the specific details and characteristics of the study embodiment referenced herein . census data are derived at the block group level following the socio - economic characteristics of vulnerability ( such as : hispanic population , black population , asian population , native american population , other race population , age 65 and over , age 65 and living below poverty , age 5 and under , population living below poverty , low education , and the like ) to extreme heat . estimates of population are derived by normalizing the total population by the area of residential land use within each block group . the area of residential land use within a block group may be determined through any number of methods including satellite imagery , aerial survey , and the like . the area of residential land use is typically selected over other possible values , such as total block group area , because it provides a truer indicator of residential density within each block group . this is to say that it provides a more accurate description of the residential density within each block group . however , any convenient relevant value may be chosen heat related fatalities for the area in question is obtained . the data is typically filtered to only include those deaths that occurred during the previously mentioned extreme heat event . the addresses of those qualified decedents are then assigned geographic identifies ( hereafter geocoded ). additionally , the deaths within each block group are totaled to produce a dataset representative of block group level ehe mortality . typically , the landsat thermal mapping ( hereafter tm ) imagery is acquired for the time period in question . the thermal band of the image is then converted to an at - satellite brightness temperature per the following equation , where t is an estimate of land surface temperature in kelvin . k2 is the calibration constant for temperature in kelvin and k1 is the constant for radiance in mwcm 2\u03bcm \u2212 1 . l w is the spectral radiance in mwcm 2 , calculated from the digital number values of the landsat tm thermal band . typically , t is then averaged by block group . this is done to determine the mean estimated land surface temperature per block group . the average t values are then uniformly stratified into the number of different levels desired . the typical spatial analysis method is the standard deviational ellipse ( sde ). it is a well known method and highly suited for point patterns . the end result of this analysis is the assignment of a weight to each of the descriptive variables . for example , in the most simple case , the variables found to be highly descriptive of the data are assigned a weight of 1 ( one ) while those not found so are assigned a weight of 0 ( zero ). the calculation of the sde is reasonably uncomplicated and many current gis applications allow for its use . the sde first requires that the centroid of each block group be calculated and the demographic measures , decedents and t measures respectively be assigned . typically , a weighted mean center of the point set will also be calculated as part of the calculation of the sde . use of the weighted mean center provides a better descriptor of vulnerability than a non - weighted mean center . the weighted mean center is obtained by averaging the coordinates of all the points and providing a weight to each based on an attribute variable of interest . after the mean center is calculated , each point is then transformed into a different metric space referenced from the mean center . the equation for this transformation is x \u2032 j = x j \u2212 x weighted mean center with the y transformation essentially being the same equation . the angle of rotation from the transformed points is calculated by the standard distance on x and y are then calculated as \u03b4 x =\u221a{ square root over (( \u03c3 i \u2212 1 n ( x \u2032 i cos \u03b4 y =\u221a{ square root over (( \u03c3 i \u2212 1 n ( x \u2032 i sin \u03b8 x , \u03b8 y , x weighted mean center , y weighted mean center , and area are used to quantitatively compare the spatial distributions of all the variables . standard t - test and f test are used to determine the levels of spatial similarity . this in turn indicates the importance of the actual variables when describing the real data . additional evaluations of the variables and their importance in describing the data are achieved through evaluation of concentration and eccentricity . a death concentration value is calculated within the spatial distributions of t . eccentricity values are calculated as \u03b8 x / \u03b8 y . the concentration value describes the level of concentration of the spatial phenomena and the eccentricity indicates the polarity of the point distribution within the ellipse . with respect to a variable , the greater the concentration of death and / or smaller eccentricity , the greater coverage of that variable &# 39 ; s descriptive capability with respect to the actual data . typically , a multiple regression modeling technique is used . first , all non - zero weighted variables are interpolated to standard sized cells covering the study area using a kernel density function . after calculation of the kernel density , the mean value per block group residential area is calculated . the non - zero weighted variables are evaluated for multi - collinearity . if needed , any collinearity is removed . a mapping of the kernel density of real death points ( the actual data ) is performed . a multiple regression utilizing the non - zero weighted variables and t as the independent variables is performed , with density of ehe death being the dependent variable . outputs of the regression are generated , forming standardized predictive values of risk . these values are mapped at the census - block group level with decedent locations as the validation layer . the standard r 2 test ( a test that determines what fraction of the total squared error is attributed to the model ) or the like may be used to determine the effectiveness of the model in explaining the variation of the dependent variable . following a significant r 2 value , the models outputs can be viewed as spatially predictive values of future risk . maps depicting spatial variation of risk , typically through a 3 - d map with the y axis representative of relative risk or the like of the city can be created . in the event of an extreme heat event or a predicted extreme heat event , health care professionals concentrate intervention measures into areas denoted as at high risk . while the invention has been illustrated and described in detail in the foregoing description , the same is to be considered as illustrative and not restrictive in character . it is understood that the embodiments have been shown and described in the foregoing specification in satisfaction of the best mode and enablement requirements . it is understood that one of ordinary skill in the art could readily make a nigh - infinite number of insubstantial changes and modifications to the above - described embodiments and that it would be impractical to attempt to describe all such embodiment variations in the present specification . accordingly , it is understood that all changes and modifications that come within the spirit of the invention are desired to be protected .", "category": "Physics"}
|
{"patent": "for the purposes of promoting an understanding of the principles of the invention and presenting its currently understood best mode of operation , reference will now be made to the embodiments illustrated and specific language will be used to describe the same . it will nevertheless be understood that no limitation of the scope of the invention is thereby intended , with such alterations and further modifications in the illustrated device and such further applications of the principles of the invention as illustrated therein being contemplated as would normally occur to one skilled in the art to which the invention relates . the present novel technology was developed to improve urban response to ehe &# 39 ; s . consequently , the following examples and embodiments to reflect and reference a study area consisting of a large , urban area . this study environment was selected such that the large , urban area has experienced an extreme heat event . furthermore , this particular urban area was selected in part because for the time of the extreme heat event there was available associated population data and landsat tm imagery data . as such , ehe &# 39 ; s are naturally occurring and infrequent , and not inherently reproducible in a controlled laboratory environment , this study is frequently referenced herein . however , it should be kept in mind that the present novel technology is broadly applicable beyond the specific details and characteristics of the study embodiment referenced herein . census data are derived at the block group level following the socio - economic characteristics of vulnerability ( such as : hispanic population , black population , asian population , native american population , other race population , age 65 and over , age 65 and living below poverty , age 5 and under , population living below poverty , low education , and the like ) to extreme heat . estimates of population are derived by normalizing the total population by the area of residential land use within each block group . the area of residential land use within a block group may be determined through any number of methods including satellite imagery , aerial survey , and the like . the area of residential land use is typically selected over other possible values , such as total block group area , because it provides a truer indicator of residential density within each block group . this is to say that it provides a more accurate description of the residential density within each block group . however , any convenient relevant value may be chosen heat related fatalities for the area in question is obtained . the data is typically filtered to only include those deaths that occurred during the previously mentioned extreme heat event . the addresses of those qualified decedents are then assigned geographic identifies ( hereafter geocoded ). additionally , the deaths within each block group are totaled to produce a dataset representative of block group level ehe mortality . typically , the landsat thermal mapping ( hereafter tm ) imagery is acquired for the time period in question . the thermal band of the image is then converted to an at - satellite brightness temperature per the following equation , where t is an estimate of land surface temperature in kelvin . k2 is the calibration constant for temperature in kelvin and k1 is the constant for radiance in mwcm 2\u03bcm \u2212 1 . l w is the spectral radiance in mwcm 2 , calculated from the digital number values of the landsat tm thermal band . typically , t is then averaged by block group . this is done to determine the mean estimated land surface temperature per block group . the average t values are then uniformly stratified into the number of different levels desired . the typical spatial analysis method is the standard deviational ellipse ( sde ). it is a well known method and highly suited for point patterns . the end result of this analysis is the assignment of a weight to each of the descriptive variables . for example , in the most simple case , the variables found to be highly descriptive of the data are assigned a weight of 1 ( one ) while those not found so are assigned a weight of 0 ( zero ). the calculation of the sde is reasonably uncomplicated and many current gis applications allow for its use . the sde first requires that the centroid of each block group be calculated and the demographic measures , decedents and t measures respectively be assigned . typically , a weighted mean center of the point set will also be calculated as part of the calculation of the sde . use of the weighted mean center provides a better descriptor of vulnerability than a non - weighted mean center . the weighted mean center is obtained by averaging the coordinates of all the points and providing a weight to each based on an attribute variable of interest . after the mean center is calculated , each point is then transformed into a different metric space referenced from the mean center . the equation for this transformation is x \u2032 j = x j \u2212 x weighted mean center with the y transformation essentially being the same equation . the angle of rotation from the transformed points is calculated by the standard distance on x and y are then calculated as \u03b4 x =\u221a{ square root over (( \u03c3 i \u2212 1 n ( x \u2032 i cos \u03b4 y =\u221a{ square root over (( \u03c3 i \u2212 1 n ( x \u2032 i sin \u03b8 x , \u03b8 y , x weighted mean center , y weighted mean center , and area are used to quantitatively compare the spatial distributions of all the variables . standard t - test and f test are used to determine the levels of spatial similarity . this in turn indicates the importance of the actual variables when describing the real data . additional evaluations of the variables and their importance in describing the data are achieved through evaluation of concentration and eccentricity . a death concentration value is calculated within the spatial distributions of t . eccentricity values are calculated as \u03b8 x / \u03b8 y . the concentration value describes the level of concentration of the spatial phenomena and the eccentricity indicates the polarity of the point distribution within the ellipse . with respect to a variable , the greater the concentration of death and / or smaller eccentricity , the greater coverage of that variable &# 39 ; s descriptive capability with respect to the actual data . typically , a multiple regression modeling technique is used . first , all non - zero weighted variables are interpolated to standard sized cells covering the study area using a kernel density function . after calculation of the kernel density , the mean value per block group residential area is calculated . the non - zero weighted variables are evaluated for multi - collinearity . if needed , any collinearity is removed . a mapping of the kernel density of real death points ( the actual data ) is performed . a multiple regression utilizing the non - zero weighted variables and t as the independent variables is performed , with density of ehe death being the dependent variable . outputs of the regression are generated , forming standardized predictive values of risk . these values are mapped at the census - block group level with decedent locations as the validation layer . the standard r 2 test ( a test that determines what fraction of the total squared error is attributed to the model ) or the like may be used to determine the effectiveness of the model in explaining the variation of the dependent variable . following a significant r 2 value , the models outputs can be viewed as spatially predictive values of future risk . maps depicting spatial variation of risk , typically through a 3 - d map with the y axis representative of relative risk or the like of the city can be created . in the event of an extreme heat event or a predicted extreme heat event , health care professionals concentrate intervention measures into areas denoted as at high risk . while the invention has been illustrated and described in detail in the foregoing description , the same is to be considered as illustrative and not restrictive in character . it is understood that the embodiments have been shown and described in the foregoing specification in satisfaction of the best mode and enablement requirements . it is understood that one of ordinary skill in the art could readily make a nigh - infinite number of insubstantial changes and modifications to the above - described embodiments and that it would be impractical to attempt to describe all such embodiment variations in the present specification . accordingly , it is understood that all changes and modifications that come within the spirit of the invention are desired to be protected .", "category": "Mechanical Engineering; Lightning; Heating; Weapons; Blasting"}
|
Does the patent belong in this category?
| 0.25 |
29100f4f34af5f1b11afe606cb14cbdd00d6c6830767f9908c573993335e99fa
| 0.057373 | 0.003601 | 0.040771 | 0.003479 | 0.129883 | 0.005219 |
null |
{"patent": "for the purposes of promoting an understanding of the principles of the invention and presenting its currently understood best mode of operation , reference will now be made to the embodiments illustrated and specific language will be used to describe the same . it will nevertheless be understood that no limitation of the scope of the invention is thereby intended , with such alterations and further modifications in the illustrated device and such further applications of the principles of the invention as illustrated therein being contemplated as would normally occur to one skilled in the art to which the invention relates . the present novel technology was developed to improve urban response to ehe &# 39 ; s . consequently , the following examples and embodiments to reflect and reference a study area consisting of a large , urban area . this study environment was selected such that the large , urban area has experienced an extreme heat event . furthermore , this particular urban area was selected in part because for the time of the extreme heat event there was available associated population data and landsat tm imagery data . as such , ehe &# 39 ; s are naturally occurring and infrequent , and not inherently reproducible in a controlled laboratory environment , this study is frequently referenced herein . however , it should be kept in mind that the present novel technology is broadly applicable beyond the specific details and characteristics of the study embodiment referenced herein . census data are derived at the block group level following the socio - economic characteristics of vulnerability ( such as : hispanic population , black population , asian population , native american population , other race population , age 65 and over , age 65 and living below poverty , age 5 and under , population living below poverty , low education , and the like ) to extreme heat . estimates of population are derived by normalizing the total population by the area of residential land use within each block group . the area of residential land use within a block group may be determined through any number of methods including satellite imagery , aerial survey , and the like . the area of residential land use is typically selected over other possible values , such as total block group area , because it provides a truer indicator of residential density within each block group . this is to say that it provides a more accurate description of the residential density within each block group . however , any convenient relevant value may be chosen heat related fatalities for the area in question is obtained . the data is typically filtered to only include those deaths that occurred during the previously mentioned extreme heat event . the addresses of those qualified decedents are then assigned geographic identifies ( hereafter geocoded ). additionally , the deaths within each block group are totaled to produce a dataset representative of block group level ehe mortality . typically , the landsat thermal mapping ( hereafter tm ) imagery is acquired for the time period in question . the thermal band of the image is then converted to an at - satellite brightness temperature per the following equation , where t is an estimate of land surface temperature in kelvin . k2 is the calibration constant for temperature in kelvin and k1 is the constant for radiance in mwcm 2\u03bcm \u2212 1 . l w is the spectral radiance in mwcm 2 , calculated from the digital number values of the landsat tm thermal band . typically , t is then averaged by block group . this is done to determine the mean estimated land surface temperature per block group . the average t values are then uniformly stratified into the number of different levels desired . the typical spatial analysis method is the standard deviational ellipse ( sde ). it is a well known method and highly suited for point patterns . the end result of this analysis is the assignment of a weight to each of the descriptive variables . for example , in the most simple case , the variables found to be highly descriptive of the data are assigned a weight of 1 ( one ) while those not found so are assigned a weight of 0 ( zero ). the calculation of the sde is reasonably uncomplicated and many current gis applications allow for its use . the sde first requires that the centroid of each block group be calculated and the demographic measures , decedents and t measures respectively be assigned . typically , a weighted mean center of the point set will also be calculated as part of the calculation of the sde . use of the weighted mean center provides a better descriptor of vulnerability than a non - weighted mean center . the weighted mean center is obtained by averaging the coordinates of all the points and providing a weight to each based on an attribute variable of interest . after the mean center is calculated , each point is then transformed into a different metric space referenced from the mean center . the equation for this transformation is x \u2032 j = x j \u2212 x weighted mean center with the y transformation essentially being the same equation . the angle of rotation from the transformed points is calculated by the standard distance on x and y are then calculated as \u03b4 x =\u221a{ square root over (( \u03c3 i \u2212 1 n ( x \u2032 i cos \u03b4 y =\u221a{ square root over (( \u03c3 i \u2212 1 n ( x \u2032 i sin \u03b8 x , \u03b8 y , x weighted mean center , y weighted mean center , and area are used to quantitatively compare the spatial distributions of all the variables . standard t - test and f test are used to determine the levels of spatial similarity . this in turn indicates the importance of the actual variables when describing the real data . additional evaluations of the variables and their importance in describing the data are achieved through evaluation of concentration and eccentricity . a death concentration value is calculated within the spatial distributions of t . eccentricity values are calculated as \u03b8 x / \u03b8 y . the concentration value describes the level of concentration of the spatial phenomena and the eccentricity indicates the polarity of the point distribution within the ellipse . with respect to a variable , the greater the concentration of death and / or smaller eccentricity , the greater coverage of that variable &# 39 ; s descriptive capability with respect to the actual data . typically , a multiple regression modeling technique is used . first , all non - zero weighted variables are interpolated to standard sized cells covering the study area using a kernel density function . after calculation of the kernel density , the mean value per block group residential area is calculated . the non - zero weighted variables are evaluated for multi - collinearity . if needed , any collinearity is removed . a mapping of the kernel density of real death points ( the actual data ) is performed . a multiple regression utilizing the non - zero weighted variables and t as the independent variables is performed , with density of ehe death being the dependent variable . outputs of the regression are generated , forming standardized predictive values of risk . these values are mapped at the census - block group level with decedent locations as the validation layer . the standard r 2 test ( a test that determines what fraction of the total squared error is attributed to the model ) or the like may be used to determine the effectiveness of the model in explaining the variation of the dependent variable . following a significant r 2 value , the models outputs can be viewed as spatially predictive values of future risk . maps depicting spatial variation of risk , typically through a 3 - d map with the y axis representative of relative risk or the like of the city can be created . in the event of an extreme heat event or a predicted extreme heat event , health care professionals concentrate intervention measures into areas denoted as at high risk . while the invention has been illustrated and described in detail in the foregoing description , the same is to be considered as illustrative and not restrictive in character . it is understood that the embodiments have been shown and described in the foregoing specification in satisfaction of the best mode and enablement requirements . it is understood that one of ordinary skill in the art could readily make a nigh - infinite number of insubstantial changes and modifications to the above - described embodiments and that it would be impractical to attempt to describe all such embodiment variations in the present specification . accordingly , it is understood that all changes and modifications that come within the spirit of the invention are desired to be protected .", "category": "Physics"}
|
{"category": "Electricity", "patent": "for the purposes of promoting an understanding of the principles of the invention and presenting its currently understood best mode of operation , reference will now be made to the embodiments illustrated and specific language will be used to describe the same . it will nevertheless be understood that no limitation of the scope of the invention is thereby intended , with such alterations and further modifications in the illustrated device and such further applications of the principles of the invention as illustrated therein being contemplated as would normally occur to one skilled in the art to which the invention relates . the present novel technology was developed to improve urban response to ehe &# 39 ; s . consequently , the following examples and embodiments to reflect and reference a study area consisting of a large , urban area . this study environment was selected such that the large , urban area has experienced an extreme heat event . furthermore , this particular urban area was selected in part because for the time of the extreme heat event there was available associated population data and landsat tm imagery data . as such , ehe &# 39 ; s are naturally occurring and infrequent , and not inherently reproducible in a controlled laboratory environment , this study is frequently referenced herein . however , it should be kept in mind that the present novel technology is broadly applicable beyond the specific details and characteristics of the study embodiment referenced herein . census data are derived at the block group level following the socio - economic characteristics of vulnerability ( such as : hispanic population , black population , asian population , native american population , other race population , age 65 and over , age 65 and living below poverty , age 5 and under , population living below poverty , low education , and the like ) to extreme heat . estimates of population are derived by normalizing the total population by the area of residential land use within each block group . the area of residential land use within a block group may be determined through any number of methods including satellite imagery , aerial survey , and the like . the area of residential land use is typically selected over other possible values , such as total block group area , because it provides a truer indicator of residential density within each block group . this is to say that it provides a more accurate description of the residential density within each block group . however , any convenient relevant value may be chosen heat related fatalities for the area in question is obtained . the data is typically filtered to only include those deaths that occurred during the previously mentioned extreme heat event . the addresses of those qualified decedents are then assigned geographic identifies ( hereafter geocoded ). additionally , the deaths within each block group are totaled to produce a dataset representative of block group level ehe mortality . typically , the landsat thermal mapping ( hereafter tm ) imagery is acquired for the time period in question . the thermal band of the image is then converted to an at - satellite brightness temperature per the following equation , where t is an estimate of land surface temperature in kelvin . k2 is the calibration constant for temperature in kelvin and k1 is the constant for radiance in mwcm 2\u03bcm \u2212 1 . l w is the spectral radiance in mwcm 2 , calculated from the digital number values of the landsat tm thermal band . typically , t is then averaged by block group . this is done to determine the mean estimated land surface temperature per block group . the average t values are then uniformly stratified into the number of different levels desired . the typical spatial analysis method is the standard deviational ellipse ( sde ). it is a well known method and highly suited for point patterns . the end result of this analysis is the assignment of a weight to each of the descriptive variables . for example , in the most simple case , the variables found to be highly descriptive of the data are assigned a weight of 1 ( one ) while those not found so are assigned a weight of 0 ( zero ). the calculation of the sde is reasonably uncomplicated and many current gis applications allow for its use . the sde first requires that the centroid of each block group be calculated and the demographic measures , decedents and t measures respectively be assigned . typically , a weighted mean center of the point set will also be calculated as part of the calculation of the sde . use of the weighted mean center provides a better descriptor of vulnerability than a non - weighted mean center . the weighted mean center is obtained by averaging the coordinates of all the points and providing a weight to each based on an attribute variable of interest . after the mean center is calculated , each point is then transformed into a different metric space referenced from the mean center . the equation for this transformation is x \u2032 j = x j \u2212 x weighted mean center with the y transformation essentially being the same equation . the angle of rotation from the transformed points is calculated by the standard distance on x and y are then calculated as \u03b4 x =\u221a{ square root over (( \u03c3 i \u2212 1 n ( x \u2032 i cos \u03b4 y =\u221a{ square root over (( \u03c3 i \u2212 1 n ( x \u2032 i sin \u03b8 x , \u03b8 y , x weighted mean center , y weighted mean center , and area are used to quantitatively compare the spatial distributions of all the variables . standard t - test and f test are used to determine the levels of spatial similarity . this in turn indicates the importance of the actual variables when describing the real data . additional evaluations of the variables and their importance in describing the data are achieved through evaluation of concentration and eccentricity . a death concentration value is calculated within the spatial distributions of t . eccentricity values are calculated as \u03b8 x / \u03b8 y . the concentration value describes the level of concentration of the spatial phenomena and the eccentricity indicates the polarity of the point distribution within the ellipse . with respect to a variable , the greater the concentration of death and / or smaller eccentricity , the greater coverage of that variable &# 39 ; s descriptive capability with respect to the actual data . typically , a multiple regression modeling technique is used . first , all non - zero weighted variables are interpolated to standard sized cells covering the study area using a kernel density function . after calculation of the kernel density , the mean value per block group residential area is calculated . the non - zero weighted variables are evaluated for multi - collinearity . if needed , any collinearity is removed . a mapping of the kernel density of real death points ( the actual data ) is performed . a multiple regression utilizing the non - zero weighted variables and t as the independent variables is performed , with density of ehe death being the dependent variable . outputs of the regression are generated , forming standardized predictive values of risk . these values are mapped at the census - block group level with decedent locations as the validation layer . the standard r 2 test ( a test that determines what fraction of the total squared error is attributed to the model ) or the like may be used to determine the effectiveness of the model in explaining the variation of the dependent variable . following a significant r 2 value , the models outputs can be viewed as spatially predictive values of future risk . maps depicting spatial variation of risk , typically through a 3 - d map with the y axis representative of relative risk or the like of the city can be created . in the event of an extreme heat event or a predicted extreme heat event , health care professionals concentrate intervention measures into areas denoted as at high risk . while the invention has been illustrated and described in detail in the foregoing description , the same is to be considered as illustrative and not restrictive in character . it is understood that the embodiments have been shown and described in the foregoing specification in satisfaction of the best mode and enablement requirements . it is understood that one of ordinary skill in the art could readily make a nigh - infinite number of insubstantial changes and modifications to the above - described embodiments and that it would be impractical to attempt to describe all such embodiment variations in the present specification . accordingly , it is understood that all changes and modifications that come within the spirit of the invention are desired to be protected ."}
|
Does the patent belong in this category?
| 0.25 |
29100f4f34af5f1b11afe606cb14cbdd00d6c6830767f9908c573993335e99fa
| 0.057373 | 0.014526 | 0.040771 | 0.00193 | 0.129883 | 0.039551 |
null |
{"patent": "for the purposes of promoting an understanding of the principles of the invention and presenting its currently understood best mode of operation , reference will now be made to the embodiments illustrated and specific language will be used to describe the same . it will nevertheless be understood that no limitation of the scope of the invention is thereby intended , with such alterations and further modifications in the illustrated device and such further applications of the principles of the invention as illustrated therein being contemplated as would normally occur to one skilled in the art to which the invention relates . the present novel technology was developed to improve urban response to ehe &# 39 ; s . consequently , the following examples and embodiments to reflect and reference a study area consisting of a large , urban area . this study environment was selected such that the large , urban area has experienced an extreme heat event . furthermore , this particular urban area was selected in part because for the time of the extreme heat event there was available associated population data and landsat tm imagery data . as such , ehe &# 39 ; s are naturally occurring and infrequent , and not inherently reproducible in a controlled laboratory environment , this study is frequently referenced herein . however , it should be kept in mind that the present novel technology is broadly applicable beyond the specific details and characteristics of the study embodiment referenced herein . census data are derived at the block group level following the socio - economic characteristics of vulnerability ( such as : hispanic population , black population , asian population , native american population , other race population , age 65 and over , age 65 and living below poverty , age 5 and under , population living below poverty , low education , and the like ) to extreme heat . estimates of population are derived by normalizing the total population by the area of residential land use within each block group . the area of residential land use within a block group may be determined through any number of methods including satellite imagery , aerial survey , and the like . the area of residential land use is typically selected over other possible values , such as total block group area , because it provides a truer indicator of residential density within each block group . this is to say that it provides a more accurate description of the residential density within each block group . however , any convenient relevant value may be chosen heat related fatalities for the area in question is obtained . the data is typically filtered to only include those deaths that occurred during the previously mentioned extreme heat event . the addresses of those qualified decedents are then assigned geographic identifies ( hereafter geocoded ). additionally , the deaths within each block group are totaled to produce a dataset representative of block group level ehe mortality . typically , the landsat thermal mapping ( hereafter tm ) imagery is acquired for the time period in question . the thermal band of the image is then converted to an at - satellite brightness temperature per the following equation , where t is an estimate of land surface temperature in kelvin . k2 is the calibration constant for temperature in kelvin and k1 is the constant for radiance in mwcm 2\u03bcm \u2212 1 . l w is the spectral radiance in mwcm 2 , calculated from the digital number values of the landsat tm thermal band . typically , t is then averaged by block group . this is done to determine the mean estimated land surface temperature per block group . the average t values are then uniformly stratified into the number of different levels desired . the typical spatial analysis method is the standard deviational ellipse ( sde ). it is a well known method and highly suited for point patterns . the end result of this analysis is the assignment of a weight to each of the descriptive variables . for example , in the most simple case , the variables found to be highly descriptive of the data are assigned a weight of 1 ( one ) while those not found so are assigned a weight of 0 ( zero ). the calculation of the sde is reasonably uncomplicated and many current gis applications allow for its use . the sde first requires that the centroid of each block group be calculated and the demographic measures , decedents and t measures respectively be assigned . typically , a weighted mean center of the point set will also be calculated as part of the calculation of the sde . use of the weighted mean center provides a better descriptor of vulnerability than a non - weighted mean center . the weighted mean center is obtained by averaging the coordinates of all the points and providing a weight to each based on an attribute variable of interest . after the mean center is calculated , each point is then transformed into a different metric space referenced from the mean center . the equation for this transformation is x \u2032 j = x j \u2212 x weighted mean center with the y transformation essentially being the same equation . the angle of rotation from the transformed points is calculated by the standard distance on x and y are then calculated as \u03b4 x =\u221a{ square root over (( \u03c3 i \u2212 1 n ( x \u2032 i cos \u03b4 y =\u221a{ square root over (( \u03c3 i \u2212 1 n ( x \u2032 i sin \u03b8 x , \u03b8 y , x weighted mean center , y weighted mean center , and area are used to quantitatively compare the spatial distributions of all the variables . standard t - test and f test are used to determine the levels of spatial similarity . this in turn indicates the importance of the actual variables when describing the real data . additional evaluations of the variables and their importance in describing the data are achieved through evaluation of concentration and eccentricity . a death concentration value is calculated within the spatial distributions of t . eccentricity values are calculated as \u03b8 x / \u03b8 y . the concentration value describes the level of concentration of the spatial phenomena and the eccentricity indicates the polarity of the point distribution within the ellipse . with respect to a variable , the greater the concentration of death and / or smaller eccentricity , the greater coverage of that variable &# 39 ; s descriptive capability with respect to the actual data . typically , a multiple regression modeling technique is used . first , all non - zero weighted variables are interpolated to standard sized cells covering the study area using a kernel density function . after calculation of the kernel density , the mean value per block group residential area is calculated . the non - zero weighted variables are evaluated for multi - collinearity . if needed , any collinearity is removed . a mapping of the kernel density of real death points ( the actual data ) is performed . a multiple regression utilizing the non - zero weighted variables and t as the independent variables is performed , with density of ehe death being the dependent variable . outputs of the regression are generated , forming standardized predictive values of risk . these values are mapped at the census - block group level with decedent locations as the validation layer . the standard r 2 test ( a test that determines what fraction of the total squared error is attributed to the model ) or the like may be used to determine the effectiveness of the model in explaining the variation of the dependent variable . following a significant r 2 value , the models outputs can be viewed as spatially predictive values of future risk . maps depicting spatial variation of risk , typically through a 3 - d map with the y axis representative of relative risk or the like of the city can be created . in the event of an extreme heat event or a predicted extreme heat event , health care professionals concentrate intervention measures into areas denoted as at high risk . while the invention has been illustrated and described in detail in the foregoing description , the same is to be considered as illustrative and not restrictive in character . it is understood that the embodiments have been shown and described in the foregoing specification in satisfaction of the best mode and enablement requirements . it is understood that one of ordinary skill in the art could readily make a nigh - infinite number of insubstantial changes and modifications to the above - described embodiments and that it would be impractical to attempt to describe all such embodiment variations in the present specification . accordingly , it is understood that all changes and modifications that come within the spirit of the invention are desired to be protected .", "category": "Physics"}
|
{"patent": "for the purposes of promoting an understanding of the principles of the invention and presenting its currently understood best mode of operation , reference will now be made to the embodiments illustrated and specific language will be used to describe the same . it will nevertheless be understood that no limitation of the scope of the invention is thereby intended , with such alterations and further modifications in the illustrated device and such further applications of the principles of the invention as illustrated therein being contemplated as would normally occur to one skilled in the art to which the invention relates . the present novel technology was developed to improve urban response to ehe &# 39 ; s . consequently , the following examples and embodiments to reflect and reference a study area consisting of a large , urban area . this study environment was selected such that the large , urban area has experienced an extreme heat event . furthermore , this particular urban area was selected in part because for the time of the extreme heat event there was available associated population data and landsat tm imagery data . as such , ehe &# 39 ; s are naturally occurring and infrequent , and not inherently reproducible in a controlled laboratory environment , this study is frequently referenced herein . however , it should be kept in mind that the present novel technology is broadly applicable beyond the specific details and characteristics of the study embodiment referenced herein . census data are derived at the block group level following the socio - economic characteristics of vulnerability ( such as : hispanic population , black population , asian population , native american population , other race population , age 65 and over , age 65 and living below poverty , age 5 and under , population living below poverty , low education , and the like ) to extreme heat . estimates of population are derived by normalizing the total population by the area of residential land use within each block group . the area of residential land use within a block group may be determined through any number of methods including satellite imagery , aerial survey , and the like . the area of residential land use is typically selected over other possible values , such as total block group area , because it provides a truer indicator of residential density within each block group . this is to say that it provides a more accurate description of the residential density within each block group . however , any convenient relevant value may be chosen heat related fatalities for the area in question is obtained . the data is typically filtered to only include those deaths that occurred during the previously mentioned extreme heat event . the addresses of those qualified decedents are then assigned geographic identifies ( hereafter geocoded ). additionally , the deaths within each block group are totaled to produce a dataset representative of block group level ehe mortality . typically , the landsat thermal mapping ( hereafter tm ) imagery is acquired for the time period in question . the thermal band of the image is then converted to an at - satellite brightness temperature per the following equation , where t is an estimate of land surface temperature in kelvin . k2 is the calibration constant for temperature in kelvin and k1 is the constant for radiance in mwcm 2\u03bcm \u2212 1 . l w is the spectral radiance in mwcm 2 , calculated from the digital number values of the landsat tm thermal band . typically , t is then averaged by block group . this is done to determine the mean estimated land surface temperature per block group . the average t values are then uniformly stratified into the number of different levels desired . the typical spatial analysis method is the standard deviational ellipse ( sde ). it is a well known method and highly suited for point patterns . the end result of this analysis is the assignment of a weight to each of the descriptive variables . for example , in the most simple case , the variables found to be highly descriptive of the data are assigned a weight of 1 ( one ) while those not found so are assigned a weight of 0 ( zero ). the calculation of the sde is reasonably uncomplicated and many current gis applications allow for its use . the sde first requires that the centroid of each block group be calculated and the demographic measures , decedents and t measures respectively be assigned . typically , a weighted mean center of the point set will also be calculated as part of the calculation of the sde . use of the weighted mean center provides a better descriptor of vulnerability than a non - weighted mean center . the weighted mean center is obtained by averaging the coordinates of all the points and providing a weight to each based on an attribute variable of interest . after the mean center is calculated , each point is then transformed into a different metric space referenced from the mean center . the equation for this transformation is x \u2032 j = x j \u2212 x weighted mean center with the y transformation essentially being the same equation . the angle of rotation from the transformed points is calculated by the standard distance on x and y are then calculated as \u03b4 x =\u221a{ square root over (( \u03c3 i \u2212 1 n ( x \u2032 i cos \u03b4 y =\u221a{ square root over (( \u03c3 i \u2212 1 n ( x \u2032 i sin \u03b8 x , \u03b8 y , x weighted mean center , y weighted mean center , and area are used to quantitatively compare the spatial distributions of all the variables . standard t - test and f test are used to determine the levels of spatial similarity . this in turn indicates the importance of the actual variables when describing the real data . additional evaluations of the variables and their importance in describing the data are achieved through evaluation of concentration and eccentricity . a death concentration value is calculated within the spatial distributions of t . eccentricity values are calculated as \u03b8 x / \u03b8 y . the concentration value describes the level of concentration of the spatial phenomena and the eccentricity indicates the polarity of the point distribution within the ellipse . with respect to a variable , the greater the concentration of death and / or smaller eccentricity , the greater coverage of that variable &# 39 ; s descriptive capability with respect to the actual data . typically , a multiple regression modeling technique is used . first , all non - zero weighted variables are interpolated to standard sized cells covering the study area using a kernel density function . after calculation of the kernel density , the mean value per block group residential area is calculated . the non - zero weighted variables are evaluated for multi - collinearity . if needed , any collinearity is removed . a mapping of the kernel density of real death points ( the actual data ) is performed . a multiple regression utilizing the non - zero weighted variables and t as the independent variables is performed , with density of ehe death being the dependent variable . outputs of the regression are generated , forming standardized predictive values of risk . these values are mapped at the census - block group level with decedent locations as the validation layer . the standard r 2 test ( a test that determines what fraction of the total squared error is attributed to the model ) or the like may be used to determine the effectiveness of the model in explaining the variation of the dependent variable . following a significant r 2 value , the models outputs can be viewed as spatially predictive values of future risk . maps depicting spatial variation of risk , typically through a 3 - d map with the y axis representative of relative risk or the like of the city can be created . in the event of an extreme heat event or a predicted extreme heat event , health care professionals concentrate intervention measures into areas denoted as at high risk . while the invention has been illustrated and described in detail in the foregoing description , the same is to be considered as illustrative and not restrictive in character . it is understood that the embodiments have been shown and described in the foregoing specification in satisfaction of the best mode and enablement requirements . it is understood that one of ordinary skill in the art could readily make a nigh - infinite number of insubstantial changes and modifications to the above - described embodiments and that it would be impractical to attempt to describe all such embodiment variations in the present specification . accordingly , it is understood that all changes and modifications that come within the spirit of the invention are desired to be protected .", "category": "General tagging of new or cross-sectional technology"}
|
Does the patent belong in this category?
| 0.25 |
29100f4f34af5f1b11afe606cb14cbdd00d6c6830767f9908c573993335e99fa
| 0.057373 | 0.080566 | 0.040771 | 0.359375 | 0.129883 | 0.144531 |
null |
{"patent": "the term \u201c integrated circuit \u201d ( ic ) is defined as an integrated circuit chip and a package or module containing the ic . the term \u201c integrated circuit chip \u201d is defined as the semiconductor ( e . g ., silicon ) die containing devices such as transistors , diodes , capacitors , resisters and inductors and the wiring layers built up on the die that interconnect the devices into circuits . the term micro - electronic device includes an ic or an integrated circuit chip . in magnetic resonance imaging ( also called nuclear magnetic resonance ( nmr ) imaging )) a sample is positioned in a static magnetic field and subjected to a pulsed radio frequency ( rf ) signal to place the sample in an excited state . the magnetic field may be turned off or adjusted and a rf return signal is produced by the sample returning to a normal from the excited state is then recorded . in order to allow spatial encoding , the static magnetic field is superimposed with a gradient magnetic field . the term validation means the process by which an unknown micro - electronic device is \u201c imaged \u201d by nmr and the resulting data is compared to data from a known good or trusted micro - electronic that was nmr \u201c imaged \u201d and that the unknown ic micro - electronic device should be essentially identical ( i . e ., of identical design and within fabrication specification limits ) to . known good micro - electronic devices include those fabricated under secure conditions , those from trusted sources and those thoroughly tested and physically inspected after an nmr signature has been obtained . fig1 is a diagram of an exemplary magnetic resonance inspection system according of the present invention . in fig1 , a magnetic resonance imaging ( mri ) system 100 includes a magnet unit 105 , a signal processing unit 110 and a computer 115 . magnet unit 105 of mri system 100 includes a gradient magnetic field coil 120 , a transmit coil 125 , a sample chamber 130 , a receive coil and a high gauss magnet 140 ( e . g ., a permanent magnet , coil magnet or super - conductive magnet ). signal processing unit 110 includes a driving circuit 145 , an rf power amplifier 150 , a preamplifier 155 , a sequence memory 160 , a modulation circuit 165 , an rf oscillator 170 , a phase detector 175 and an analog - to - digital ( a / d ) converter 180 . computer 115 includes a display 185 , an input ( e . g ., a keyboard , mouse , disk drive ) and an output ( e . g ., a display unit , a printer , a disk drive ). gradient magnetic field coil 120 , transmit coil 125 , receive coil 135 and high gauss magnet 140 are disposed so as to substantially surround sample chamber 130 . high gauss magnet 140 applies a static magnetic field having a constant strength to a sample in chamber 130 . gradient magnetic field coil 120 applies gradient magnetic fields selectively to mutually orthogonal x , y and z directions ( in imaging parlance , to a slice axis , phase axis and frequency axis ). transmit coil 125 supplies a pulsed rf signal for exciting spins of atomic nuclei within a micro - electronic device 200 in sample chamber 130 . receive coil 135 detects returned rf signals from the sample in chamber 130 generated when spins of atomic nuclei within micro - electronic device 200 return to a normal state from an exited state . gradient magnetic field coil 120 , transmit coil 125 , and receive coil 135 are operatively associated with driving circuit 145 , an rf power amplifier 150 , and a preamplifier 155 , respectively . sequence memory 160 operates driving circuit 145 based on a stored pulse sequence in response to instructions from computer 115 to thereby apply gradient fields from gradient magnetic field coil 120 in specific directions . sequence memory 160 also operates a modulation circuit 165 to modulate a carrier output signal from rf oscillator 170 into a pulsed rf signal of predefined timing and envelope shape . the pulsed rf signal is applied to rf power amplifier 150 and then the amplified pulsed rf signal is applied to transmit coil 125 . preamplifier 155 amplifies the return rf signal from micro - electronic device 200 in sample chamber 130 detected at receive coil 135 . preamplifier 155 amplifies the received rf signal and sends an amplified rf signal to phase detector 175 . phase detector 175 generates an analog phase - detect signal from the amplified rf signal using a carrier output signal from rf oscillator 170 as a reference signal , and supplies the phase - detected signal to a / d converter 180 . a / d converter 180 converts the phase - detected analog signal into a digital signal , which is supplied to the computer 115 . computer 115 reads and / or processes the data from a / d converter 1801 , and includes algorithms in the form of computer instructions which when executed perform various signal analyses and statistical analyses on the stored data as described infra . results of these analyses may be displayed on output unit 195 . computer 115 can also be responsible for overall control such as receiving information supplied from input 195 from an operator . mri system 100 includes an optional tester 205 , which is connected between a socket , or probe card ( not shown ) in sample chamber 130 and computer 185 . this allows voltage bias , analog signals , digital data patterns or combinations thereof to be applied to micro - electronic device 200 during the mri process . biasing , applying signals and test patterns serves two purposes . first it results in more complex return rf signals . second , if the biasing analog signals , digital data patterns are kept secure , it is very difficult for a an unauthorized party to place a masking circuit into an unauthorized micro electronic device , the purpose of the masking circuit being is to mask the presence of the unauthorized circuit and the masking circuit in the unauthorized micro - electronic device by altering the nmr image of the unauthorized micro - electronic device to mimic that of an authorized micro - electronic device . mri system 100 shown in fig1 is provided as an example , and it will be understood that embodiments of the invention are not limited to mri system 100 shown in fig1 . it will be understood that an mri system 100 according to aspects of the invention can include additional components to those shown in fig1 or may not include every component shown in fig1 . fig2 is a cross - section through an exemplary first type of integrated circuit . in fig2 , an ic 200 a includes an integrated circuit chip 210 physically and electrically connected to a module 215 by solder bumps 220 . wires 225 in module 225 connect solder bumps 220 to solder balls 230 . module 215 may be organic ( e . g ., fiberglass ) or ceramic and wires 25 may comprise one or more wiring layers . solder balls 230 are designed for surface mounting ic 200 a directly to a printed circuit board ( pcb ). solder balls 230 may be replaced by solder columns . solder balls 230 may be replaced with pins , which can be used to mount ics into sockets , which are mounted on a pcb or mounted directly to a pcb . ic 200 a is thus an example of an ic that uses flip - chip ( or c4 ) technology . fig3 is a cross - section through an exemplary second type of integrated circuit . in fig3 , an ic 200 b includes and integrated circuit chip contained within a plastic package 240 . wire bonds 245 electrically connect integrated circuit chip 235 to leads 250 . leads 250 are designed for surface mounting ic 200 b directly to a pcb . leads 250 may be replaced with pins , which can be used to mount ics into sockets , which are mounted on a pcb or mounted directly to a pcb . the leads on some plastic packages are designed to be mounted in sockets on a pcb . ic 200 b is thus an example of plastic packaging technology . fig4 is a top view of an exemplary integrated circuit chip for enhanced comparison according embodiments of the present invention . in fig4 , an exemplary integrated circuit chip 255 includes regions 260 a , 260 b , 260 c , 260 d , 260 e , 260 f , 260 g , 260 h , 2601 , 269 j and 260 k . there may be more or less regions than illustrated in fig4 . these regions often correspond to cores , which are pre - designed circuit functions . for example , a microprocessor may contain multiple processing cores , memory cores , arithmetic cores etc . formed , by way of example , in cores 260 b , 260 g , 260 i and 260 j are serpentine signal enhancing structures 265 which are not electrically connected to any wire or device ( e . g ., transistor , diode , resistor , capacitor or inductor ) of integrated circuit chip 255 . signal enhancing structures 265 may comprise an electrical conductor , a magnetic material , or an electrically conductive magnetic material . signal enhancing structures 265 are designed to interact with the magnetic fields from gradient magnetic field coil 120 and high gauss magnet 140 of mri system 100 ( see fig1 ) to generate a more complex return rf signal . fig5 is a top view of an exemplary integrated circuit chip for preventing or detecting comparison according embodiments of the present invention . in fig5 , an integrated circuit chip 255 a ( similar to integrated circuit chip 255 of fig4 ) includes a serpentine structure 265 a connected to a destruct circuit 270 . serpentine structure 265 a acts as an inductor which generates a current when subjected to a varying magnetic field . the current may be used by destruct circuit 270 to program fuses or activate transistors to render integrated circuit 155 a inoperable or to leave a signature that can later be read to indicate if an attempt at nmr imaging \u201d has been performed on integrated circuit chip 255 a . in one example , serpentine structure 265 a is designed to not be detectable by x - ray imaging . fig6 a is a cross - sectional view of an integrated circuit containing devices to flag an unauthorized comparison attempt according embodiments of the present invention . in fig6 a , an ic 200 c includes an integrated circuit chip 235 a in a plastic package body 240 a . a set ( two sets are shown in the example of fig6 a ) of three \u201c horseshoe \u201d magnets 275 a , 275 b and 275 c are placed in body 240 a . they are aligned so respective lines ( dashed lines ) passing through the poles of each magnet are mutually orthogonal . in fig6 a , the line passing through the poles of magnets 275 a and 275 c are in planes parallel to the top surface 272 of integrated circuit chip 235 a . other pole orientations are possible . in one example , only a single horseshoe magnet is used . horseshoe shaped magnets may be replaced by other shaped magnets such as bar magnets and disc magnets . fig6 b is a cross - sectional view of the integrated circuit of fig6 a after an unauthorized comparison attempt according embodiments of the present invention . when magnets 275 a , 275 b and 275 c are subjected to the intense magnetic field generated in an nmr machine , magnets 275 a , 275 b and 275 c are pulled / pushed by that field so strongly that cracks 280 are formed in body 240 a . fig7 is a flowchart of methods of comparison according to embodiments of the present invention . the steps of the flowchart of fig7 are performed first on a known or trusted micro - electronic device and then on an unknown ( or suspect ) micro - electronic device that and essentially identical to the known micro - electronic device ( i . e ., identically designed and within fabrication specification limits ). for the integrated circuit chip specification limits define allowable variations in material and structure and include , for example , the allowable differences in metal line widths and thickness differences in metal and insulating layers . for the package of the ic specification limits define allowable variations in material and structure and include , for example , package dimensions , size of solder bumps , positions and size of wire bonds , widths and thickness of land in modules . in step 300 an ic ( or integrated circuit is placed in the mri chamber . in step 305 the high gauss magnetic field is turned on . in one example , the high gauss magnetic field has a field strength of between about a 0 . 5 tesla and about 10 tesla and is applied in the z direction . the method can know follow one of two mutually exclusive paths . the first is the path ( 1 ) through steps 310 , 315 , 320 , 325 and 330 to step 335 . the second path ( 2 ) is through steps 315 a , 320 a and 325 a to step 335 . the unknown micro - electronic device advantageously follows the same path and is subjected to the same nmr conditions as the known micro - electronic device . in step 310 , a direction ( x , y or z ) is selected . in step 315 , a gradient magnetic field of , for example , 1 tesla is applied in the selected direction and in step 320 a rf pulse is directed to the ic . in step 325 the emitted ( returned ) rf signal is detected and information describing the return rf signal ( which is also a pulse ) is stored . the duration and time of reception of the return rf signal will vary . in step 330 , if another direction of the three possible directions ( x , y , z ) is selected and the method loops back to step 310 . this loop will repeat three times , once for each direction . in path ( 2 ) in step 315 a , gradient magnetic fields are applied in the x , y , and z directions simultaneously . each gradient field may have a same or a different field strength of between about 0 . 5 tesla and about 10 tesla . steps 320 a and 325 a are similar to steps 320 and 325 respectively . in steps 315 and 315 a , optional test conditions ( i . e ., voltage bias , analog signal , digital data pattern or combinations thereof ) may be applied to the micro - electronic device . in a first example , optional test conditions are applied only during step 315 ( or 315 a ). in a second example , optional test conditions are applied the only during steps 315 and 320 ( or 315 a and 320 a ). in a third example , the optional test conditions are applied the only during steps 305 , 310 , 315 and 320 ( or 305 , 315 a and 320 a ). in a fourth example , the optional test conditions are applied during steps 305 , 310 , 315 , 320 and 325 ( or 305 , 315 a , 320 a and 325 a ). in step 335 , the ic is removed from the mri chamber and it is determined whether the micro - electronic device was a known micro - electronic device or an unknown micro - electronic device . if the micro - electronic device was a known micro - electronic device then in step 340 analyses are performed and the analyses data is stored . if the micro - electronic device was an unknown micro - electronic device then in step 345 analyses are performed and the analyses is compared to analyses previously stored from a known micro - electronic device mri \u201c images \u201d under the same mri ( and bias / pattern ) conditions . if the compare is within predefined limits the unknown micro - electronic device is validated . it does not matter whether the known or unknown micro - electronic device is run first , but the comparison cannot be performed until both known or unknown micro - electronic devices have been run and the respective data analyzed . fig8 a and 8b illustrate a first method of data analysis for comparison according to embodiments of the present invention . fig8 a is a plot of the transverse component of the returned rf signal for a known micro - electronic device 350 and an unknown micro - electronic device 355 as relative rf strength versus time . direct comparison or statistical analysis of the difference between curves 350 and 355 is performed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig8 a , the difference between curve 355 relative to curve 350 is shown as a positive time shift . the shift may be negative . fig8 b is a plot of the longitudinal component of the returned rf signal for a known micro - electronic device 360 and an unknown micro - electronic device 365 as relative rf strength versus time . direct comparison or statistical analysis of the difference between curves 360 and 365 is performed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig8 b , the difference of curve 365 relative to curve 360 is shown as a negative time shift . the shift may be positive . fig9 a and 9b illustrate a second method of data analysis for comparison according to embodiments of the present invention . fig9 a is a plot of a fast fourier transform of the returned rf signal for a known ic ( or integrated circuit ) as rf amplitude versus rf frequency . element 370 is the returned signal and elements 370 a and 370 b are harmonic artifacts of element 370 . fig9 b is a plot of a fast fourier transform of the returned rf signal for an unknown ic ( or integrated circuit ) as rf amplitude versus rf frequency . element 375 is the returned signal and elements 375 a and 3750 b are harmonic artifacts of element 370 . element 370 and 375 are statistically compared in terms of amplitude of a given frequency range to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig9 a and 9b , the amplitude of the returned rf signal has been transferred from a time domain to a frequency domain . fig1 a , 10 b and 10 c illustrate a third method of data analysis for comparison according to embodiments of the present invention . fig1 a is a plot of the returned rf signal as relative rf signal strength versus time for a known micro - electronic device . rf signal 380 has a fixed amplitude between times a and b measured from when the rf pulse signal terminated ( or other convenient time reference ). an unknown micro - electronic device may exhibit a signal that is offset ion amplitude , phase , frequency or combinations thereof . two simple examples are given in fig1 b and 10c . fig1 b is a plot of the returned rf signal as relative rf signal strength versus time for an unknown micro - electronic device having only a frequency shift . rf signal 385 a has fixed amplitude between times a \u2032 and b \u2032 measured from the same reference as a and b of fig1 a . the shift between a and a \u2032 and b and b \u2032 is statistically to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . fig1 c is a plot of the returned rf signal as relative rf signal strength versus time for an unknown micro - electronic device having only an amplitude shift . rf signal 385 b has a fixed amplitude between times a and b measured from the same reference as a and b of fig1 a . the change in amplitude between curve 380 of fig1 a and curve 385 b of fig1 c is statistically to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . it will be appreciated that a complete statistical analysis would compare time , amplitude , frequency and phase differences of the curves described in fig1 a , 10 b and 10 c . fig1 a , 11 b and 11 c illustrate a fourth method of data analysis for comparison according to embodiments of the present invention . fig1 a is a plot of the returned rf signal xy space ( at a selected z ) as a function of frequency for a known micro - electronic device . as such fig1 a is closer to what is may be considered an \u201c image .\u201d fig1 b is a plot of the returned rf signal xy space ( at the selected z ) as a function of frequency for an unknown micro - electronic device . fig1 c is a plot of the delta between the plot of fig1 a and that of fig1 b . the amount of structure ( and optionally the position of the structures ) is statistically analyzed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . fig1 c is fig1 a \u201c subtracted \u201d from fig1 a . thus the embodiments of the present invention allow for relatively quick and inexpensive validation of integrated circuit chips . the description of the embodiments of the present invention is given above for the understanding of the present invention . it will be understood that the invention is not limited to the particular embodiments described herein , but is capable of various modifications , rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention . therefore , it is intended that the following claims cover all such modifications and changes as fall within the true spirit and scope of the invention .", "category": "Physics"}
|
{"category": "Human Necessities", "patent": "the term \u201c integrated circuit \u201d ( ic ) is defined as an integrated circuit chip and a package or module containing the ic . the term \u201c integrated circuit chip \u201d is defined as the semiconductor ( e . g ., silicon ) die containing devices such as transistors , diodes , capacitors , resisters and inductors and the wiring layers built up on the die that interconnect the devices into circuits . the term micro - electronic device includes an ic or an integrated circuit chip . in magnetic resonance imaging ( also called nuclear magnetic resonance ( nmr ) imaging )) a sample is positioned in a static magnetic field and subjected to a pulsed radio frequency ( rf ) signal to place the sample in an excited state . the magnetic field may be turned off or adjusted and a rf return signal is produced by the sample returning to a normal from the excited state is then recorded . in order to allow spatial encoding , the static magnetic field is superimposed with a gradient magnetic field . the term validation means the process by which an unknown micro - electronic device is \u201c imaged \u201d by nmr and the resulting data is compared to data from a known good or trusted micro - electronic that was nmr \u201c imaged \u201d and that the unknown ic micro - electronic device should be essentially identical ( i . e ., of identical design and within fabrication specification limits ) to . known good micro - electronic devices include those fabricated under secure conditions , those from trusted sources and those thoroughly tested and physically inspected after an nmr signature has been obtained . fig1 is a diagram of an exemplary magnetic resonance inspection system according of the present invention . in fig1 , a magnetic resonance imaging ( mri ) system 100 includes a magnet unit 105 , a signal processing unit 110 and a computer 115 . magnet unit 105 of mri system 100 includes a gradient magnetic field coil 120 , a transmit coil 125 , a sample chamber 130 , a receive coil and a high gauss magnet 140 ( e . g ., a permanent magnet , coil magnet or super - conductive magnet ). signal processing unit 110 includes a driving circuit 145 , an rf power amplifier 150 , a preamplifier 155 , a sequence memory 160 , a modulation circuit 165 , an rf oscillator 170 , a phase detector 175 and an analog - to - digital ( a / d ) converter 180 . computer 115 includes a display 185 , an input ( e . g ., a keyboard , mouse , disk drive ) and an output ( e . g ., a display unit , a printer , a disk drive ). gradient magnetic field coil 120 , transmit coil 125 , receive coil 135 and high gauss magnet 140 are disposed so as to substantially surround sample chamber 130 . high gauss magnet 140 applies a static magnetic field having a constant strength to a sample in chamber 130 . gradient magnetic field coil 120 applies gradient magnetic fields selectively to mutually orthogonal x , y and z directions ( in imaging parlance , to a slice axis , phase axis and frequency axis ). transmit coil 125 supplies a pulsed rf signal for exciting spins of atomic nuclei within a micro - electronic device 200 in sample chamber 130 . receive coil 135 detects returned rf signals from the sample in chamber 130 generated when spins of atomic nuclei within micro - electronic device 200 return to a normal state from an exited state . gradient magnetic field coil 120 , transmit coil 125 , and receive coil 135 are operatively associated with driving circuit 145 , an rf power amplifier 150 , and a preamplifier 155 , respectively . sequence memory 160 operates driving circuit 145 based on a stored pulse sequence in response to instructions from computer 115 to thereby apply gradient fields from gradient magnetic field coil 120 in specific directions . sequence memory 160 also operates a modulation circuit 165 to modulate a carrier output signal from rf oscillator 170 into a pulsed rf signal of predefined timing and envelope shape . the pulsed rf signal is applied to rf power amplifier 150 and then the amplified pulsed rf signal is applied to transmit coil 125 . preamplifier 155 amplifies the return rf signal from micro - electronic device 200 in sample chamber 130 detected at receive coil 135 . preamplifier 155 amplifies the received rf signal and sends an amplified rf signal to phase detector 175 . phase detector 175 generates an analog phase - detect signal from the amplified rf signal using a carrier output signal from rf oscillator 170 as a reference signal , and supplies the phase - detected signal to a / d converter 180 . a / d converter 180 converts the phase - detected analog signal into a digital signal , which is supplied to the computer 115 . computer 115 reads and / or processes the data from a / d converter 1801 , and includes algorithms in the form of computer instructions which when executed perform various signal analyses and statistical analyses on the stored data as described infra . results of these analyses may be displayed on output unit 195 . computer 115 can also be responsible for overall control such as receiving information supplied from input 195 from an operator . mri system 100 includes an optional tester 205 , which is connected between a socket , or probe card ( not shown ) in sample chamber 130 and computer 185 . this allows voltage bias , analog signals , digital data patterns or combinations thereof to be applied to micro - electronic device 200 during the mri process . biasing , applying signals and test patterns serves two purposes . first it results in more complex return rf signals . second , if the biasing analog signals , digital data patterns are kept secure , it is very difficult for a an unauthorized party to place a masking circuit into an unauthorized micro electronic device , the purpose of the masking circuit being is to mask the presence of the unauthorized circuit and the masking circuit in the unauthorized micro - electronic device by altering the nmr image of the unauthorized micro - electronic device to mimic that of an authorized micro - electronic device . mri system 100 shown in fig1 is provided as an example , and it will be understood that embodiments of the invention are not limited to mri system 100 shown in fig1 . it will be understood that an mri system 100 according to aspects of the invention can include additional components to those shown in fig1 or may not include every component shown in fig1 . fig2 is a cross - section through an exemplary first type of integrated circuit . in fig2 , an ic 200 a includes an integrated circuit chip 210 physically and electrically connected to a module 215 by solder bumps 220 . wires 225 in module 225 connect solder bumps 220 to solder balls 230 . module 215 may be organic ( e . g ., fiberglass ) or ceramic and wires 25 may comprise one or more wiring layers . solder balls 230 are designed for surface mounting ic 200 a directly to a printed circuit board ( pcb ). solder balls 230 may be replaced by solder columns . solder balls 230 may be replaced with pins , which can be used to mount ics into sockets , which are mounted on a pcb or mounted directly to a pcb . ic 200 a is thus an example of an ic that uses flip - chip ( or c4 ) technology . fig3 is a cross - section through an exemplary second type of integrated circuit . in fig3 , an ic 200 b includes and integrated circuit chip contained within a plastic package 240 . wire bonds 245 electrically connect integrated circuit chip 235 to leads 250 . leads 250 are designed for surface mounting ic 200 b directly to a pcb . leads 250 may be replaced with pins , which can be used to mount ics into sockets , which are mounted on a pcb or mounted directly to a pcb . the leads on some plastic packages are designed to be mounted in sockets on a pcb . ic 200 b is thus an example of plastic packaging technology . fig4 is a top view of an exemplary integrated circuit chip for enhanced comparison according embodiments of the present invention . in fig4 , an exemplary integrated circuit chip 255 includes regions 260 a , 260 b , 260 c , 260 d , 260 e , 260 f , 260 g , 260 h , 2601 , 269 j and 260 k . there may be more or less regions than illustrated in fig4 . these regions often correspond to cores , which are pre - designed circuit functions . for example , a microprocessor may contain multiple processing cores , memory cores , arithmetic cores etc . formed , by way of example , in cores 260 b , 260 g , 260 i and 260 j are serpentine signal enhancing structures 265 which are not electrically connected to any wire or device ( e . g ., transistor , diode , resistor , capacitor or inductor ) of integrated circuit chip 255 . signal enhancing structures 265 may comprise an electrical conductor , a magnetic material , or an electrically conductive magnetic material . signal enhancing structures 265 are designed to interact with the magnetic fields from gradient magnetic field coil 120 and high gauss magnet 140 of mri system 100 ( see fig1 ) to generate a more complex return rf signal . fig5 is a top view of an exemplary integrated circuit chip for preventing or detecting comparison according embodiments of the present invention . in fig5 , an integrated circuit chip 255 a ( similar to integrated circuit chip 255 of fig4 ) includes a serpentine structure 265 a connected to a destruct circuit 270 . serpentine structure 265 a acts as an inductor which generates a current when subjected to a varying magnetic field . the current may be used by destruct circuit 270 to program fuses or activate transistors to render integrated circuit 155 a inoperable or to leave a signature that can later be read to indicate if an attempt at nmr imaging \u201d has been performed on integrated circuit chip 255 a . in one example , serpentine structure 265 a is designed to not be detectable by x - ray imaging . fig6 a is a cross - sectional view of an integrated circuit containing devices to flag an unauthorized comparison attempt according embodiments of the present invention . in fig6 a , an ic 200 c includes an integrated circuit chip 235 a in a plastic package body 240 a . a set ( two sets are shown in the example of fig6 a ) of three \u201c horseshoe \u201d magnets 275 a , 275 b and 275 c are placed in body 240 a . they are aligned so respective lines ( dashed lines ) passing through the poles of each magnet are mutually orthogonal . in fig6 a , the line passing through the poles of magnets 275 a and 275 c are in planes parallel to the top surface 272 of integrated circuit chip 235 a . other pole orientations are possible . in one example , only a single horseshoe magnet is used . horseshoe shaped magnets may be replaced by other shaped magnets such as bar magnets and disc magnets . fig6 b is a cross - sectional view of the integrated circuit of fig6 a after an unauthorized comparison attempt according embodiments of the present invention . when magnets 275 a , 275 b and 275 c are subjected to the intense magnetic field generated in an nmr machine , magnets 275 a , 275 b and 275 c are pulled / pushed by that field so strongly that cracks 280 are formed in body 240 a . fig7 is a flowchart of methods of comparison according to embodiments of the present invention . the steps of the flowchart of fig7 are performed first on a known or trusted micro - electronic device and then on an unknown ( or suspect ) micro - electronic device that and essentially identical to the known micro - electronic device ( i . e ., identically designed and within fabrication specification limits ). for the integrated circuit chip specification limits define allowable variations in material and structure and include , for example , the allowable differences in metal line widths and thickness differences in metal and insulating layers . for the package of the ic specification limits define allowable variations in material and structure and include , for example , package dimensions , size of solder bumps , positions and size of wire bonds , widths and thickness of land in modules . in step 300 an ic ( or integrated circuit is placed in the mri chamber . in step 305 the high gauss magnetic field is turned on . in one example , the high gauss magnetic field has a field strength of between about a 0 . 5 tesla and about 10 tesla and is applied in the z direction . the method can know follow one of two mutually exclusive paths . the first is the path ( 1 ) through steps 310 , 315 , 320 , 325 and 330 to step 335 . the second path ( 2 ) is through steps 315 a , 320 a and 325 a to step 335 . the unknown micro - electronic device advantageously follows the same path and is subjected to the same nmr conditions as the known micro - electronic device . in step 310 , a direction ( x , y or z ) is selected . in step 315 , a gradient magnetic field of , for example , 1 tesla is applied in the selected direction and in step 320 a rf pulse is directed to the ic . in step 325 the emitted ( returned ) rf signal is detected and information describing the return rf signal ( which is also a pulse ) is stored . the duration and time of reception of the return rf signal will vary . in step 330 , if another direction of the three possible directions ( x , y , z ) is selected and the method loops back to step 310 . this loop will repeat three times , once for each direction . in path ( 2 ) in step 315 a , gradient magnetic fields are applied in the x , y , and z directions simultaneously . each gradient field may have a same or a different field strength of between about 0 . 5 tesla and about 10 tesla . steps 320 a and 325 a are similar to steps 320 and 325 respectively . in steps 315 and 315 a , optional test conditions ( i . e ., voltage bias , analog signal , digital data pattern or combinations thereof ) may be applied to the micro - electronic device . in a first example , optional test conditions are applied only during step 315 ( or 315 a ). in a second example , optional test conditions are applied the only during steps 315 and 320 ( or 315 a and 320 a ). in a third example , the optional test conditions are applied the only during steps 305 , 310 , 315 and 320 ( or 305 , 315 a and 320 a ). in a fourth example , the optional test conditions are applied during steps 305 , 310 , 315 , 320 and 325 ( or 305 , 315 a , 320 a and 325 a ). in step 335 , the ic is removed from the mri chamber and it is determined whether the micro - electronic device was a known micro - electronic device or an unknown micro - electronic device . if the micro - electronic device was a known micro - electronic device then in step 340 analyses are performed and the analyses data is stored . if the micro - electronic device was an unknown micro - electronic device then in step 345 analyses are performed and the analyses is compared to analyses previously stored from a known micro - electronic device mri \u201c images \u201d under the same mri ( and bias / pattern ) conditions . if the compare is within predefined limits the unknown micro - electronic device is validated . it does not matter whether the known or unknown micro - electronic device is run first , but the comparison cannot be performed until both known or unknown micro - electronic devices have been run and the respective data analyzed . fig8 a and 8b illustrate a first method of data analysis for comparison according to embodiments of the present invention . fig8 a is a plot of the transverse component of the returned rf signal for a known micro - electronic device 350 and an unknown micro - electronic device 355 as relative rf strength versus time . direct comparison or statistical analysis of the difference between curves 350 and 355 is performed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig8 a , the difference between curve 355 relative to curve 350 is shown as a positive time shift . the shift may be negative . fig8 b is a plot of the longitudinal component of the returned rf signal for a known micro - electronic device 360 and an unknown micro - electronic device 365 as relative rf strength versus time . direct comparison or statistical analysis of the difference between curves 360 and 365 is performed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig8 b , the difference of curve 365 relative to curve 360 is shown as a negative time shift . the shift may be positive . fig9 a and 9b illustrate a second method of data analysis for comparison according to embodiments of the present invention . fig9 a is a plot of a fast fourier transform of the returned rf signal for a known ic ( or integrated circuit ) as rf amplitude versus rf frequency . element 370 is the returned signal and elements 370 a and 370 b are harmonic artifacts of element 370 . fig9 b is a plot of a fast fourier transform of the returned rf signal for an unknown ic ( or integrated circuit ) as rf amplitude versus rf frequency . element 375 is the returned signal and elements 375 a and 3750 b are harmonic artifacts of element 370 . element 370 and 375 are statistically compared in terms of amplitude of a given frequency range to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig9 a and 9b , the amplitude of the returned rf signal has been transferred from a time domain to a frequency domain . fig1 a , 10 b and 10 c illustrate a third method of data analysis for comparison according to embodiments of the present invention . fig1 a is a plot of the returned rf signal as relative rf signal strength versus time for a known micro - electronic device . rf signal 380 has a fixed amplitude between times a and b measured from when the rf pulse signal terminated ( or other convenient time reference ). an unknown micro - electronic device may exhibit a signal that is offset ion amplitude , phase , frequency or combinations thereof . two simple examples are given in fig1 b and 10c . fig1 b is a plot of the returned rf signal as relative rf signal strength versus time for an unknown micro - electronic device having only a frequency shift . rf signal 385 a has fixed amplitude between times a \u2032 and b \u2032 measured from the same reference as a and b of fig1 a . the shift between a and a \u2032 and b and b \u2032 is statistically to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . fig1 c is a plot of the returned rf signal as relative rf signal strength versus time for an unknown micro - electronic device having only an amplitude shift . rf signal 385 b has a fixed amplitude between times a and b measured from the same reference as a and b of fig1 a . the change in amplitude between curve 380 of fig1 a and curve 385 b of fig1 c is statistically to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . it will be appreciated that a complete statistical analysis would compare time , amplitude , frequency and phase differences of the curves described in fig1 a , 10 b and 10 c . fig1 a , 11 b and 11 c illustrate a fourth method of data analysis for comparison according to embodiments of the present invention . fig1 a is a plot of the returned rf signal xy space ( at a selected z ) as a function of frequency for a known micro - electronic device . as such fig1 a is closer to what is may be considered an \u201c image .\u201d fig1 b is a plot of the returned rf signal xy space ( at the selected z ) as a function of frequency for an unknown micro - electronic device . fig1 c is a plot of the delta between the plot of fig1 a and that of fig1 b . the amount of structure ( and optionally the position of the structures ) is statistically analyzed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . fig1 c is fig1 a \u201c subtracted \u201d from fig1 a . thus the embodiments of the present invention allow for relatively quick and inexpensive validation of integrated circuit chips . the description of the embodiments of the present invention is given above for the understanding of the present invention . it will be understood that the invention is not limited to the particular embodiments described herein , but is capable of various modifications , rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention . therefore , it is intended that the following claims cover all such modifications and changes as fall within the true spirit and scope of the invention ."}
|
Does the patent belong in this category?
| 0.25 |
d2a533e20b88cf12a66089384d35063da6ac708b8421ee9a63cae12490153828
| 0.012024 | 0.003372 | 0.046143 | 0.001328 | 0.228516 | 0.006287 |
null |
{"category": "Physics", "patent": "the term \u201c integrated circuit \u201d ( ic ) is defined as an integrated circuit chip and a package or module containing the ic . the term \u201c integrated circuit chip \u201d is defined as the semiconductor ( e . g ., silicon ) die containing devices such as transistors , diodes , capacitors , resisters and inductors and the wiring layers built up on the die that interconnect the devices into circuits . the term micro - electronic device includes an ic or an integrated circuit chip . in magnetic resonance imaging ( also called nuclear magnetic resonance ( nmr ) imaging )) a sample is positioned in a static magnetic field and subjected to a pulsed radio frequency ( rf ) signal to place the sample in an excited state . the magnetic field may be turned off or adjusted and a rf return signal is produced by the sample returning to a normal from the excited state is then recorded . in order to allow spatial encoding , the static magnetic field is superimposed with a gradient magnetic field . the term validation means the process by which an unknown micro - electronic device is \u201c imaged \u201d by nmr and the resulting data is compared to data from a known good or trusted micro - electronic that was nmr \u201c imaged \u201d and that the unknown ic micro - electronic device should be essentially identical ( i . e ., of identical design and within fabrication specification limits ) to . known good micro - electronic devices include those fabricated under secure conditions , those from trusted sources and those thoroughly tested and physically inspected after an nmr signature has been obtained . fig1 is a diagram of an exemplary magnetic resonance inspection system according of the present invention . in fig1 , a magnetic resonance imaging ( mri ) system 100 includes a magnet unit 105 , a signal processing unit 110 and a computer 115 . magnet unit 105 of mri system 100 includes a gradient magnetic field coil 120 , a transmit coil 125 , a sample chamber 130 , a receive coil and a high gauss magnet 140 ( e . g ., a permanent magnet , coil magnet or super - conductive magnet ). signal processing unit 110 includes a driving circuit 145 , an rf power amplifier 150 , a preamplifier 155 , a sequence memory 160 , a modulation circuit 165 , an rf oscillator 170 , a phase detector 175 and an analog - to - digital ( a / d ) converter 180 . computer 115 includes a display 185 , an input ( e . g ., a keyboard , mouse , disk drive ) and an output ( e . g ., a display unit , a printer , a disk drive ). gradient magnetic field coil 120 , transmit coil 125 , receive coil 135 and high gauss magnet 140 are disposed so as to substantially surround sample chamber 130 . high gauss magnet 140 applies a static magnetic field having a constant strength to a sample in chamber 130 . gradient magnetic field coil 120 applies gradient magnetic fields selectively to mutually orthogonal x , y and z directions ( in imaging parlance , to a slice axis , phase axis and frequency axis ). transmit coil 125 supplies a pulsed rf signal for exciting spins of atomic nuclei within a micro - electronic device 200 in sample chamber 130 . receive coil 135 detects returned rf signals from the sample in chamber 130 generated when spins of atomic nuclei within micro - electronic device 200 return to a normal state from an exited state . gradient magnetic field coil 120 , transmit coil 125 , and receive coil 135 are operatively associated with driving circuit 145 , an rf power amplifier 150 , and a preamplifier 155 , respectively . sequence memory 160 operates driving circuit 145 based on a stored pulse sequence in response to instructions from computer 115 to thereby apply gradient fields from gradient magnetic field coil 120 in specific directions . sequence memory 160 also operates a modulation circuit 165 to modulate a carrier output signal from rf oscillator 170 into a pulsed rf signal of predefined timing and envelope shape . the pulsed rf signal is applied to rf power amplifier 150 and then the amplified pulsed rf signal is applied to transmit coil 125 . preamplifier 155 amplifies the return rf signal from micro - electronic device 200 in sample chamber 130 detected at receive coil 135 . preamplifier 155 amplifies the received rf signal and sends an amplified rf signal to phase detector 175 . phase detector 175 generates an analog phase - detect signal from the amplified rf signal using a carrier output signal from rf oscillator 170 as a reference signal , and supplies the phase - detected signal to a / d converter 180 . a / d converter 180 converts the phase - detected analog signal into a digital signal , which is supplied to the computer 115 . computer 115 reads and / or processes the data from a / d converter 1801 , and includes algorithms in the form of computer instructions which when executed perform various signal analyses and statistical analyses on the stored data as described infra . results of these analyses may be displayed on output unit 195 . computer 115 can also be responsible for overall control such as receiving information supplied from input 195 from an operator . mri system 100 includes an optional tester 205 , which is connected between a socket , or probe card ( not shown ) in sample chamber 130 and computer 185 . this allows voltage bias , analog signals , digital data patterns or combinations thereof to be applied to micro - electronic device 200 during the mri process . biasing , applying signals and test patterns serves two purposes . first it results in more complex return rf signals . second , if the biasing analog signals , digital data patterns are kept secure , it is very difficult for a an unauthorized party to place a masking circuit into an unauthorized micro electronic device , the purpose of the masking circuit being is to mask the presence of the unauthorized circuit and the masking circuit in the unauthorized micro - electronic device by altering the nmr image of the unauthorized micro - electronic device to mimic that of an authorized micro - electronic device . mri system 100 shown in fig1 is provided as an example , and it will be understood that embodiments of the invention are not limited to mri system 100 shown in fig1 . it will be understood that an mri system 100 according to aspects of the invention can include additional components to those shown in fig1 or may not include every component shown in fig1 . fig2 is a cross - section through an exemplary first type of integrated circuit . in fig2 , an ic 200 a includes an integrated circuit chip 210 physically and electrically connected to a module 215 by solder bumps 220 . wires 225 in module 225 connect solder bumps 220 to solder balls 230 . module 215 may be organic ( e . g ., fiberglass ) or ceramic and wires 25 may comprise one or more wiring layers . solder balls 230 are designed for surface mounting ic 200 a directly to a printed circuit board ( pcb ). solder balls 230 may be replaced by solder columns . solder balls 230 may be replaced with pins , which can be used to mount ics into sockets , which are mounted on a pcb or mounted directly to a pcb . ic 200 a is thus an example of an ic that uses flip - chip ( or c4 ) technology . fig3 is a cross - section through an exemplary second type of integrated circuit . in fig3 , an ic 200 b includes and integrated circuit chip contained within a plastic package 240 . wire bonds 245 electrically connect integrated circuit chip 235 to leads 250 . leads 250 are designed for surface mounting ic 200 b directly to a pcb . leads 250 may be replaced with pins , which can be used to mount ics into sockets , which are mounted on a pcb or mounted directly to a pcb . the leads on some plastic packages are designed to be mounted in sockets on a pcb . ic 200 b is thus an example of plastic packaging technology . fig4 is a top view of an exemplary integrated circuit chip for enhanced comparison according embodiments of the present invention . in fig4 , an exemplary integrated circuit chip 255 includes regions 260 a , 260 b , 260 c , 260 d , 260 e , 260 f , 260 g , 260 h , 2601 , 269 j and 260 k . there may be more or less regions than illustrated in fig4 . these regions often correspond to cores , which are pre - designed circuit functions . for example , a microprocessor may contain multiple processing cores , memory cores , arithmetic cores etc . formed , by way of example , in cores 260 b , 260 g , 260 i and 260 j are serpentine signal enhancing structures 265 which are not electrically connected to any wire or device ( e . g ., transistor , diode , resistor , capacitor or inductor ) of integrated circuit chip 255 . signal enhancing structures 265 may comprise an electrical conductor , a magnetic material , or an electrically conductive magnetic material . signal enhancing structures 265 are designed to interact with the magnetic fields from gradient magnetic field coil 120 and high gauss magnet 140 of mri system 100 ( see fig1 ) to generate a more complex return rf signal . fig5 is a top view of an exemplary integrated circuit chip for preventing or detecting comparison according embodiments of the present invention . in fig5 , an integrated circuit chip 255 a ( similar to integrated circuit chip 255 of fig4 ) includes a serpentine structure 265 a connected to a destruct circuit 270 . serpentine structure 265 a acts as an inductor which generates a current when subjected to a varying magnetic field . the current may be used by destruct circuit 270 to program fuses or activate transistors to render integrated circuit 155 a inoperable or to leave a signature that can later be read to indicate if an attempt at nmr imaging \u201d has been performed on integrated circuit chip 255 a . in one example , serpentine structure 265 a is designed to not be detectable by x - ray imaging . fig6 a is a cross - sectional view of an integrated circuit containing devices to flag an unauthorized comparison attempt according embodiments of the present invention . in fig6 a , an ic 200 c includes an integrated circuit chip 235 a in a plastic package body 240 a . a set ( two sets are shown in the example of fig6 a ) of three \u201c horseshoe \u201d magnets 275 a , 275 b and 275 c are placed in body 240 a . they are aligned so respective lines ( dashed lines ) passing through the poles of each magnet are mutually orthogonal . in fig6 a , the line passing through the poles of magnets 275 a and 275 c are in planes parallel to the top surface 272 of integrated circuit chip 235 a . other pole orientations are possible . in one example , only a single horseshoe magnet is used . horseshoe shaped magnets may be replaced by other shaped magnets such as bar magnets and disc magnets . fig6 b is a cross - sectional view of the integrated circuit of fig6 a after an unauthorized comparison attempt according embodiments of the present invention . when magnets 275 a , 275 b and 275 c are subjected to the intense magnetic field generated in an nmr machine , magnets 275 a , 275 b and 275 c are pulled / pushed by that field so strongly that cracks 280 are formed in body 240 a . fig7 is a flowchart of methods of comparison according to embodiments of the present invention . the steps of the flowchart of fig7 are performed first on a known or trusted micro - electronic device and then on an unknown ( or suspect ) micro - electronic device that and essentially identical to the known micro - electronic device ( i . e ., identically designed and within fabrication specification limits ). for the integrated circuit chip specification limits define allowable variations in material and structure and include , for example , the allowable differences in metal line widths and thickness differences in metal and insulating layers . for the package of the ic specification limits define allowable variations in material and structure and include , for example , package dimensions , size of solder bumps , positions and size of wire bonds , widths and thickness of land in modules . in step 300 an ic ( or integrated circuit is placed in the mri chamber . in step 305 the high gauss magnetic field is turned on . in one example , the high gauss magnetic field has a field strength of between about a 0 . 5 tesla and about 10 tesla and is applied in the z direction . the method can know follow one of two mutually exclusive paths . the first is the path ( 1 ) through steps 310 , 315 , 320 , 325 and 330 to step 335 . the second path ( 2 ) is through steps 315 a , 320 a and 325 a to step 335 . the unknown micro - electronic device advantageously follows the same path and is subjected to the same nmr conditions as the known micro - electronic device . in step 310 , a direction ( x , y or z ) is selected . in step 315 , a gradient magnetic field of , for example , 1 tesla is applied in the selected direction and in step 320 a rf pulse is directed to the ic . in step 325 the emitted ( returned ) rf signal is detected and information describing the return rf signal ( which is also a pulse ) is stored . the duration and time of reception of the return rf signal will vary . in step 330 , if another direction of the three possible directions ( x , y , z ) is selected and the method loops back to step 310 . this loop will repeat three times , once for each direction . in path ( 2 ) in step 315 a , gradient magnetic fields are applied in the x , y , and z directions simultaneously . each gradient field may have a same or a different field strength of between about 0 . 5 tesla and about 10 tesla . steps 320 a and 325 a are similar to steps 320 and 325 respectively . in steps 315 and 315 a , optional test conditions ( i . e ., voltage bias , analog signal , digital data pattern or combinations thereof ) may be applied to the micro - electronic device . in a first example , optional test conditions are applied only during step 315 ( or 315 a ). in a second example , optional test conditions are applied the only during steps 315 and 320 ( or 315 a and 320 a ). in a third example , the optional test conditions are applied the only during steps 305 , 310 , 315 and 320 ( or 305 , 315 a and 320 a ). in a fourth example , the optional test conditions are applied during steps 305 , 310 , 315 , 320 and 325 ( or 305 , 315 a , 320 a and 325 a ). in step 335 , the ic is removed from the mri chamber and it is determined whether the micro - electronic device was a known micro - electronic device or an unknown micro - electronic device . if the micro - electronic device was a known micro - electronic device then in step 340 analyses are performed and the analyses data is stored . if the micro - electronic device was an unknown micro - electronic device then in step 345 analyses are performed and the analyses is compared to analyses previously stored from a known micro - electronic device mri \u201c images \u201d under the same mri ( and bias / pattern ) conditions . if the compare is within predefined limits the unknown micro - electronic device is validated . it does not matter whether the known or unknown micro - electronic device is run first , but the comparison cannot be performed until both known or unknown micro - electronic devices have been run and the respective data analyzed . fig8 a and 8b illustrate a first method of data analysis for comparison according to embodiments of the present invention . fig8 a is a plot of the transverse component of the returned rf signal for a known micro - electronic device 350 and an unknown micro - electronic device 355 as relative rf strength versus time . direct comparison or statistical analysis of the difference between curves 350 and 355 is performed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig8 a , the difference between curve 355 relative to curve 350 is shown as a positive time shift . the shift may be negative . fig8 b is a plot of the longitudinal component of the returned rf signal for a known micro - electronic device 360 and an unknown micro - electronic device 365 as relative rf strength versus time . direct comparison or statistical analysis of the difference between curves 360 and 365 is performed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig8 b , the difference of curve 365 relative to curve 360 is shown as a negative time shift . the shift may be positive . fig9 a and 9b illustrate a second method of data analysis for comparison according to embodiments of the present invention . fig9 a is a plot of a fast fourier transform of the returned rf signal for a known ic ( or integrated circuit ) as rf amplitude versus rf frequency . element 370 is the returned signal and elements 370 a and 370 b are harmonic artifacts of element 370 . fig9 b is a plot of a fast fourier transform of the returned rf signal for an unknown ic ( or integrated circuit ) as rf amplitude versus rf frequency . element 375 is the returned signal and elements 375 a and 3750 b are harmonic artifacts of element 370 . element 370 and 375 are statistically compared in terms of amplitude of a given frequency range to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig9 a and 9b , the amplitude of the returned rf signal has been transferred from a time domain to a frequency domain . fig1 a , 10 b and 10 c illustrate a third method of data analysis for comparison according to embodiments of the present invention . fig1 a is a plot of the returned rf signal as relative rf signal strength versus time for a known micro - electronic device . rf signal 380 has a fixed amplitude between times a and b measured from when the rf pulse signal terminated ( or other convenient time reference ). an unknown micro - electronic device may exhibit a signal that is offset ion amplitude , phase , frequency or combinations thereof . two simple examples are given in fig1 b and 10c . fig1 b is a plot of the returned rf signal as relative rf signal strength versus time for an unknown micro - electronic device having only a frequency shift . rf signal 385 a has fixed amplitude between times a \u2032 and b \u2032 measured from the same reference as a and b of fig1 a . the shift between a and a \u2032 and b and b \u2032 is statistically to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . fig1 c is a plot of the returned rf signal as relative rf signal strength versus time for an unknown micro - electronic device having only an amplitude shift . rf signal 385 b has a fixed amplitude between times a and b measured from the same reference as a and b of fig1 a . the change in amplitude between curve 380 of fig1 a and curve 385 b of fig1 c is statistically to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . it will be appreciated that a complete statistical analysis would compare time , amplitude , frequency and phase differences of the curves described in fig1 a , 10 b and 10 c . fig1 a , 11 b and 11 c illustrate a fourth method of data analysis for comparison according to embodiments of the present invention . fig1 a is a plot of the returned rf signal xy space ( at a selected z ) as a function of frequency for a known micro - electronic device . as such fig1 a is closer to what is may be considered an \u201c image .\u201d fig1 b is a plot of the returned rf signal xy space ( at the selected z ) as a function of frequency for an unknown micro - electronic device . fig1 c is a plot of the delta between the plot of fig1 a and that of fig1 b . the amount of structure ( and optionally the position of the structures ) is statistically analyzed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . fig1 c is fig1 a \u201c subtracted \u201d from fig1 a . thus the embodiments of the present invention allow for relatively quick and inexpensive validation of integrated circuit chips . the description of the embodiments of the present invention is given above for the understanding of the present invention . it will be understood that the invention is not limited to the particular embodiments described herein , but is capable of various modifications , rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention . therefore , it is intended that the following claims cover all such modifications and changes as fall within the true spirit and scope of the invention ."}
|
{"patent": "the term \u201c integrated circuit \u201d ( ic ) is defined as an integrated circuit chip and a package or module containing the ic . the term \u201c integrated circuit chip \u201d is defined as the semiconductor ( e . g ., silicon ) die containing devices such as transistors , diodes , capacitors , resisters and inductors and the wiring layers built up on the die that interconnect the devices into circuits . the term micro - electronic device includes an ic or an integrated circuit chip . in magnetic resonance imaging ( also called nuclear magnetic resonance ( nmr ) imaging )) a sample is positioned in a static magnetic field and subjected to a pulsed radio frequency ( rf ) signal to place the sample in an excited state . the magnetic field may be turned off or adjusted and a rf return signal is produced by the sample returning to a normal from the excited state is then recorded . in order to allow spatial encoding , the static magnetic field is superimposed with a gradient magnetic field . the term validation means the process by which an unknown micro - electronic device is \u201c imaged \u201d by nmr and the resulting data is compared to data from a known good or trusted micro - electronic that was nmr \u201c imaged \u201d and that the unknown ic micro - electronic device should be essentially identical ( i . e ., of identical design and within fabrication specification limits ) to . known good micro - electronic devices include those fabricated under secure conditions , those from trusted sources and those thoroughly tested and physically inspected after an nmr signature has been obtained . fig1 is a diagram of an exemplary magnetic resonance inspection system according of the present invention . in fig1 , a magnetic resonance imaging ( mri ) system 100 includes a magnet unit 105 , a signal processing unit 110 and a computer 115 . magnet unit 105 of mri system 100 includes a gradient magnetic field coil 120 , a transmit coil 125 , a sample chamber 130 , a receive coil and a high gauss magnet 140 ( e . g ., a permanent magnet , coil magnet or super - conductive magnet ). signal processing unit 110 includes a driving circuit 145 , an rf power amplifier 150 , a preamplifier 155 , a sequence memory 160 , a modulation circuit 165 , an rf oscillator 170 , a phase detector 175 and an analog - to - digital ( a / d ) converter 180 . computer 115 includes a display 185 , an input ( e . g ., a keyboard , mouse , disk drive ) and an output ( e . g ., a display unit , a printer , a disk drive ). gradient magnetic field coil 120 , transmit coil 125 , receive coil 135 and high gauss magnet 140 are disposed so as to substantially surround sample chamber 130 . high gauss magnet 140 applies a static magnetic field having a constant strength to a sample in chamber 130 . gradient magnetic field coil 120 applies gradient magnetic fields selectively to mutually orthogonal x , y and z directions ( in imaging parlance , to a slice axis , phase axis and frequency axis ). transmit coil 125 supplies a pulsed rf signal for exciting spins of atomic nuclei within a micro - electronic device 200 in sample chamber 130 . receive coil 135 detects returned rf signals from the sample in chamber 130 generated when spins of atomic nuclei within micro - electronic device 200 return to a normal state from an exited state . gradient magnetic field coil 120 , transmit coil 125 , and receive coil 135 are operatively associated with driving circuit 145 , an rf power amplifier 150 , and a preamplifier 155 , respectively . sequence memory 160 operates driving circuit 145 based on a stored pulse sequence in response to instructions from computer 115 to thereby apply gradient fields from gradient magnetic field coil 120 in specific directions . sequence memory 160 also operates a modulation circuit 165 to modulate a carrier output signal from rf oscillator 170 into a pulsed rf signal of predefined timing and envelope shape . the pulsed rf signal is applied to rf power amplifier 150 and then the amplified pulsed rf signal is applied to transmit coil 125 . preamplifier 155 amplifies the return rf signal from micro - electronic device 200 in sample chamber 130 detected at receive coil 135 . preamplifier 155 amplifies the received rf signal and sends an amplified rf signal to phase detector 175 . phase detector 175 generates an analog phase - detect signal from the amplified rf signal using a carrier output signal from rf oscillator 170 as a reference signal , and supplies the phase - detected signal to a / d converter 180 . a / d converter 180 converts the phase - detected analog signal into a digital signal , which is supplied to the computer 115 . computer 115 reads and / or processes the data from a / d converter 1801 , and includes algorithms in the form of computer instructions which when executed perform various signal analyses and statistical analyses on the stored data as described infra . results of these analyses may be displayed on output unit 195 . computer 115 can also be responsible for overall control such as receiving information supplied from input 195 from an operator . mri system 100 includes an optional tester 205 , which is connected between a socket , or probe card ( not shown ) in sample chamber 130 and computer 185 . this allows voltage bias , analog signals , digital data patterns or combinations thereof to be applied to micro - electronic device 200 during the mri process . biasing , applying signals and test patterns serves two purposes . first it results in more complex return rf signals . second , if the biasing analog signals , digital data patterns are kept secure , it is very difficult for a an unauthorized party to place a masking circuit into an unauthorized micro electronic device , the purpose of the masking circuit being is to mask the presence of the unauthorized circuit and the masking circuit in the unauthorized micro - electronic device by altering the nmr image of the unauthorized micro - electronic device to mimic that of an authorized micro - electronic device . mri system 100 shown in fig1 is provided as an example , and it will be understood that embodiments of the invention are not limited to mri system 100 shown in fig1 . it will be understood that an mri system 100 according to aspects of the invention can include additional components to those shown in fig1 or may not include every component shown in fig1 . fig2 is a cross - section through an exemplary first type of integrated circuit . in fig2 , an ic 200 a includes an integrated circuit chip 210 physically and electrically connected to a module 215 by solder bumps 220 . wires 225 in module 225 connect solder bumps 220 to solder balls 230 . module 215 may be organic ( e . g ., fiberglass ) or ceramic and wires 25 may comprise one or more wiring layers . solder balls 230 are designed for surface mounting ic 200 a directly to a printed circuit board ( pcb ). solder balls 230 may be replaced by solder columns . solder balls 230 may be replaced with pins , which can be used to mount ics into sockets , which are mounted on a pcb or mounted directly to a pcb . ic 200 a is thus an example of an ic that uses flip - chip ( or c4 ) technology . fig3 is a cross - section through an exemplary second type of integrated circuit . in fig3 , an ic 200 b includes and integrated circuit chip contained within a plastic package 240 . wire bonds 245 electrically connect integrated circuit chip 235 to leads 250 . leads 250 are designed for surface mounting ic 200 b directly to a pcb . leads 250 may be replaced with pins , which can be used to mount ics into sockets , which are mounted on a pcb or mounted directly to a pcb . the leads on some plastic packages are designed to be mounted in sockets on a pcb . ic 200 b is thus an example of plastic packaging technology . fig4 is a top view of an exemplary integrated circuit chip for enhanced comparison according embodiments of the present invention . in fig4 , an exemplary integrated circuit chip 255 includes regions 260 a , 260 b , 260 c , 260 d , 260 e , 260 f , 260 g , 260 h , 2601 , 269 j and 260 k . there may be more or less regions than illustrated in fig4 . these regions often correspond to cores , which are pre - designed circuit functions . for example , a microprocessor may contain multiple processing cores , memory cores , arithmetic cores etc . formed , by way of example , in cores 260 b , 260 g , 260 i and 260 j are serpentine signal enhancing structures 265 which are not electrically connected to any wire or device ( e . g ., transistor , diode , resistor , capacitor or inductor ) of integrated circuit chip 255 . signal enhancing structures 265 may comprise an electrical conductor , a magnetic material , or an electrically conductive magnetic material . signal enhancing structures 265 are designed to interact with the magnetic fields from gradient magnetic field coil 120 and high gauss magnet 140 of mri system 100 ( see fig1 ) to generate a more complex return rf signal . fig5 is a top view of an exemplary integrated circuit chip for preventing or detecting comparison according embodiments of the present invention . in fig5 , an integrated circuit chip 255 a ( similar to integrated circuit chip 255 of fig4 ) includes a serpentine structure 265 a connected to a destruct circuit 270 . serpentine structure 265 a acts as an inductor which generates a current when subjected to a varying magnetic field . the current may be used by destruct circuit 270 to program fuses or activate transistors to render integrated circuit 155 a inoperable or to leave a signature that can later be read to indicate if an attempt at nmr imaging \u201d has been performed on integrated circuit chip 255 a . in one example , serpentine structure 265 a is designed to not be detectable by x - ray imaging . fig6 a is a cross - sectional view of an integrated circuit containing devices to flag an unauthorized comparison attempt according embodiments of the present invention . in fig6 a , an ic 200 c includes an integrated circuit chip 235 a in a plastic package body 240 a . a set ( two sets are shown in the example of fig6 a ) of three \u201c horseshoe \u201d magnets 275 a , 275 b and 275 c are placed in body 240 a . they are aligned so respective lines ( dashed lines ) passing through the poles of each magnet are mutually orthogonal . in fig6 a , the line passing through the poles of magnets 275 a and 275 c are in planes parallel to the top surface 272 of integrated circuit chip 235 a . other pole orientations are possible . in one example , only a single horseshoe magnet is used . horseshoe shaped magnets may be replaced by other shaped magnets such as bar magnets and disc magnets . fig6 b is a cross - sectional view of the integrated circuit of fig6 a after an unauthorized comparison attempt according embodiments of the present invention . when magnets 275 a , 275 b and 275 c are subjected to the intense magnetic field generated in an nmr machine , magnets 275 a , 275 b and 275 c are pulled / pushed by that field so strongly that cracks 280 are formed in body 240 a . fig7 is a flowchart of methods of comparison according to embodiments of the present invention . the steps of the flowchart of fig7 are performed first on a known or trusted micro - electronic device and then on an unknown ( or suspect ) micro - electronic device that and essentially identical to the known micro - electronic device ( i . e ., identically designed and within fabrication specification limits ). for the integrated circuit chip specification limits define allowable variations in material and structure and include , for example , the allowable differences in metal line widths and thickness differences in metal and insulating layers . for the package of the ic specification limits define allowable variations in material and structure and include , for example , package dimensions , size of solder bumps , positions and size of wire bonds , widths and thickness of land in modules . in step 300 an ic ( or integrated circuit is placed in the mri chamber . in step 305 the high gauss magnetic field is turned on . in one example , the high gauss magnetic field has a field strength of between about a 0 . 5 tesla and about 10 tesla and is applied in the z direction . the method can know follow one of two mutually exclusive paths . the first is the path ( 1 ) through steps 310 , 315 , 320 , 325 and 330 to step 335 . the second path ( 2 ) is through steps 315 a , 320 a and 325 a to step 335 . the unknown micro - electronic device advantageously follows the same path and is subjected to the same nmr conditions as the known micro - electronic device . in step 310 , a direction ( x , y or z ) is selected . in step 315 , a gradient magnetic field of , for example , 1 tesla is applied in the selected direction and in step 320 a rf pulse is directed to the ic . in step 325 the emitted ( returned ) rf signal is detected and information describing the return rf signal ( which is also a pulse ) is stored . the duration and time of reception of the return rf signal will vary . in step 330 , if another direction of the three possible directions ( x , y , z ) is selected and the method loops back to step 310 . this loop will repeat three times , once for each direction . in path ( 2 ) in step 315 a , gradient magnetic fields are applied in the x , y , and z directions simultaneously . each gradient field may have a same or a different field strength of between about 0 . 5 tesla and about 10 tesla . steps 320 a and 325 a are similar to steps 320 and 325 respectively . in steps 315 and 315 a , optional test conditions ( i . e ., voltage bias , analog signal , digital data pattern or combinations thereof ) may be applied to the micro - electronic device . in a first example , optional test conditions are applied only during step 315 ( or 315 a ). in a second example , optional test conditions are applied the only during steps 315 and 320 ( or 315 a and 320 a ). in a third example , the optional test conditions are applied the only during steps 305 , 310 , 315 and 320 ( or 305 , 315 a and 320 a ). in a fourth example , the optional test conditions are applied during steps 305 , 310 , 315 , 320 and 325 ( or 305 , 315 a , 320 a and 325 a ). in step 335 , the ic is removed from the mri chamber and it is determined whether the micro - electronic device was a known micro - electronic device or an unknown micro - electronic device . if the micro - electronic device was a known micro - electronic device then in step 340 analyses are performed and the analyses data is stored . if the micro - electronic device was an unknown micro - electronic device then in step 345 analyses are performed and the analyses is compared to analyses previously stored from a known micro - electronic device mri \u201c images \u201d under the same mri ( and bias / pattern ) conditions . if the compare is within predefined limits the unknown micro - electronic device is validated . it does not matter whether the known or unknown micro - electronic device is run first , but the comparison cannot be performed until both known or unknown micro - electronic devices have been run and the respective data analyzed . fig8 a and 8b illustrate a first method of data analysis for comparison according to embodiments of the present invention . fig8 a is a plot of the transverse component of the returned rf signal for a known micro - electronic device 350 and an unknown micro - electronic device 355 as relative rf strength versus time . direct comparison or statistical analysis of the difference between curves 350 and 355 is performed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig8 a , the difference between curve 355 relative to curve 350 is shown as a positive time shift . the shift may be negative . fig8 b is a plot of the longitudinal component of the returned rf signal for a known micro - electronic device 360 and an unknown micro - electronic device 365 as relative rf strength versus time . direct comparison or statistical analysis of the difference between curves 360 and 365 is performed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig8 b , the difference of curve 365 relative to curve 360 is shown as a negative time shift . the shift may be positive . fig9 a and 9b illustrate a second method of data analysis for comparison according to embodiments of the present invention . fig9 a is a plot of a fast fourier transform of the returned rf signal for a known ic ( or integrated circuit ) as rf amplitude versus rf frequency . element 370 is the returned signal and elements 370 a and 370 b are harmonic artifacts of element 370 . fig9 b is a plot of a fast fourier transform of the returned rf signal for an unknown ic ( or integrated circuit ) as rf amplitude versus rf frequency . element 375 is the returned signal and elements 375 a and 3750 b are harmonic artifacts of element 370 . element 370 and 375 are statistically compared in terms of amplitude of a given frequency range to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig9 a and 9b , the amplitude of the returned rf signal has been transferred from a time domain to a frequency domain . fig1 a , 10 b and 10 c illustrate a third method of data analysis for comparison according to embodiments of the present invention . fig1 a is a plot of the returned rf signal as relative rf signal strength versus time for a known micro - electronic device . rf signal 380 has a fixed amplitude between times a and b measured from when the rf pulse signal terminated ( or other convenient time reference ). an unknown micro - electronic device may exhibit a signal that is offset ion amplitude , phase , frequency or combinations thereof . two simple examples are given in fig1 b and 10c . fig1 b is a plot of the returned rf signal as relative rf signal strength versus time for an unknown micro - electronic device having only a frequency shift . rf signal 385 a has fixed amplitude between times a \u2032 and b \u2032 measured from the same reference as a and b of fig1 a . the shift between a and a \u2032 and b and b \u2032 is statistically to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . fig1 c is a plot of the returned rf signal as relative rf signal strength versus time for an unknown micro - electronic device having only an amplitude shift . rf signal 385 b has a fixed amplitude between times a and b measured from the same reference as a and b of fig1 a . the change in amplitude between curve 380 of fig1 a and curve 385 b of fig1 c is statistically to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . it will be appreciated that a complete statistical analysis would compare time , amplitude , frequency and phase differences of the curves described in fig1 a , 10 b and 10 c . fig1 a , 11 b and 11 c illustrate a fourth method of data analysis for comparison according to embodiments of the present invention . fig1 a is a plot of the returned rf signal xy space ( at a selected z ) as a function of frequency for a known micro - electronic device . as such fig1 a is closer to what is may be considered an \u201c image .\u201d fig1 b is a plot of the returned rf signal xy space ( at the selected z ) as a function of frequency for an unknown micro - electronic device . fig1 c is a plot of the delta between the plot of fig1 a and that of fig1 b . the amount of structure ( and optionally the position of the structures ) is statistically analyzed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . fig1 c is fig1 a \u201c subtracted \u201d from fig1 a . thus the embodiments of the present invention allow for relatively quick and inexpensive validation of integrated circuit chips . the description of the embodiments of the present invention is given above for the understanding of the present invention . it will be understood that the invention is not limited to the particular embodiments described herein , but is capable of various modifications , rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention . therefore , it is intended that the following claims cover all such modifications and changes as fall within the true spirit and scope of the invention .", "category": "Performing Operations; Transporting"}
|
Is the patent correctly categorized?
| 0.25 |
d2a533e20b88cf12a66089384d35063da6ac708b8421ee9a63cae12490153828
| 0.1875 | 0.000296 | 0.255859 | 0.004059 | 0.652344 | 0.022583 |
null |
{"patent": "the term \u201c integrated circuit \u201d ( ic ) is defined as an integrated circuit chip and a package or module containing the ic . the term \u201c integrated circuit chip \u201d is defined as the semiconductor ( e . g ., silicon ) die containing devices such as transistors , diodes , capacitors , resisters and inductors and the wiring layers built up on the die that interconnect the devices into circuits . the term micro - electronic device includes an ic or an integrated circuit chip . in magnetic resonance imaging ( also called nuclear magnetic resonance ( nmr ) imaging )) a sample is positioned in a static magnetic field and subjected to a pulsed radio frequency ( rf ) signal to place the sample in an excited state . the magnetic field may be turned off or adjusted and a rf return signal is produced by the sample returning to a normal from the excited state is then recorded . in order to allow spatial encoding , the static magnetic field is superimposed with a gradient magnetic field . the term validation means the process by which an unknown micro - electronic device is \u201c imaged \u201d by nmr and the resulting data is compared to data from a known good or trusted micro - electronic that was nmr \u201c imaged \u201d and that the unknown ic micro - electronic device should be essentially identical ( i . e ., of identical design and within fabrication specification limits ) to . known good micro - electronic devices include those fabricated under secure conditions , those from trusted sources and those thoroughly tested and physically inspected after an nmr signature has been obtained . fig1 is a diagram of an exemplary magnetic resonance inspection system according of the present invention . in fig1 , a magnetic resonance imaging ( mri ) system 100 includes a magnet unit 105 , a signal processing unit 110 and a computer 115 . magnet unit 105 of mri system 100 includes a gradient magnetic field coil 120 , a transmit coil 125 , a sample chamber 130 , a receive coil and a high gauss magnet 140 ( e . g ., a permanent magnet , coil magnet or super - conductive magnet ). signal processing unit 110 includes a driving circuit 145 , an rf power amplifier 150 , a preamplifier 155 , a sequence memory 160 , a modulation circuit 165 , an rf oscillator 170 , a phase detector 175 and an analog - to - digital ( a / d ) converter 180 . computer 115 includes a display 185 , an input ( e . g ., a keyboard , mouse , disk drive ) and an output ( e . g ., a display unit , a printer , a disk drive ). gradient magnetic field coil 120 , transmit coil 125 , receive coil 135 and high gauss magnet 140 are disposed so as to substantially surround sample chamber 130 . high gauss magnet 140 applies a static magnetic field having a constant strength to a sample in chamber 130 . gradient magnetic field coil 120 applies gradient magnetic fields selectively to mutually orthogonal x , y and z directions ( in imaging parlance , to a slice axis , phase axis and frequency axis ). transmit coil 125 supplies a pulsed rf signal for exciting spins of atomic nuclei within a micro - electronic device 200 in sample chamber 130 . receive coil 135 detects returned rf signals from the sample in chamber 130 generated when spins of atomic nuclei within micro - electronic device 200 return to a normal state from an exited state . gradient magnetic field coil 120 , transmit coil 125 , and receive coil 135 are operatively associated with driving circuit 145 , an rf power amplifier 150 , and a preamplifier 155 , respectively . sequence memory 160 operates driving circuit 145 based on a stored pulse sequence in response to instructions from computer 115 to thereby apply gradient fields from gradient magnetic field coil 120 in specific directions . sequence memory 160 also operates a modulation circuit 165 to modulate a carrier output signal from rf oscillator 170 into a pulsed rf signal of predefined timing and envelope shape . the pulsed rf signal is applied to rf power amplifier 150 and then the amplified pulsed rf signal is applied to transmit coil 125 . preamplifier 155 amplifies the return rf signal from micro - electronic device 200 in sample chamber 130 detected at receive coil 135 . preamplifier 155 amplifies the received rf signal and sends an amplified rf signal to phase detector 175 . phase detector 175 generates an analog phase - detect signal from the amplified rf signal using a carrier output signal from rf oscillator 170 as a reference signal , and supplies the phase - detected signal to a / d converter 180 . a / d converter 180 converts the phase - detected analog signal into a digital signal , which is supplied to the computer 115 . computer 115 reads and / or processes the data from a / d converter 1801 , and includes algorithms in the form of computer instructions which when executed perform various signal analyses and statistical analyses on the stored data as described infra . results of these analyses may be displayed on output unit 195 . computer 115 can also be responsible for overall control such as receiving information supplied from input 195 from an operator . mri system 100 includes an optional tester 205 , which is connected between a socket , or probe card ( not shown ) in sample chamber 130 and computer 185 . this allows voltage bias , analog signals , digital data patterns or combinations thereof to be applied to micro - electronic device 200 during the mri process . biasing , applying signals and test patterns serves two purposes . first it results in more complex return rf signals . second , if the biasing analog signals , digital data patterns are kept secure , it is very difficult for a an unauthorized party to place a masking circuit into an unauthorized micro electronic device , the purpose of the masking circuit being is to mask the presence of the unauthorized circuit and the masking circuit in the unauthorized micro - electronic device by altering the nmr image of the unauthorized micro - electronic device to mimic that of an authorized micro - electronic device . mri system 100 shown in fig1 is provided as an example , and it will be understood that embodiments of the invention are not limited to mri system 100 shown in fig1 . it will be understood that an mri system 100 according to aspects of the invention can include additional components to those shown in fig1 or may not include every component shown in fig1 . fig2 is a cross - section through an exemplary first type of integrated circuit . in fig2 , an ic 200 a includes an integrated circuit chip 210 physically and electrically connected to a module 215 by solder bumps 220 . wires 225 in module 225 connect solder bumps 220 to solder balls 230 . module 215 may be organic ( e . g ., fiberglass ) or ceramic and wires 25 may comprise one or more wiring layers . solder balls 230 are designed for surface mounting ic 200 a directly to a printed circuit board ( pcb ). solder balls 230 may be replaced by solder columns . solder balls 230 may be replaced with pins , which can be used to mount ics into sockets , which are mounted on a pcb or mounted directly to a pcb . ic 200 a is thus an example of an ic that uses flip - chip ( or c4 ) technology . fig3 is a cross - section through an exemplary second type of integrated circuit . in fig3 , an ic 200 b includes and integrated circuit chip contained within a plastic package 240 . wire bonds 245 electrically connect integrated circuit chip 235 to leads 250 . leads 250 are designed for surface mounting ic 200 b directly to a pcb . leads 250 may be replaced with pins , which can be used to mount ics into sockets , which are mounted on a pcb or mounted directly to a pcb . the leads on some plastic packages are designed to be mounted in sockets on a pcb . ic 200 b is thus an example of plastic packaging technology . fig4 is a top view of an exemplary integrated circuit chip for enhanced comparison according embodiments of the present invention . in fig4 , an exemplary integrated circuit chip 255 includes regions 260 a , 260 b , 260 c , 260 d , 260 e , 260 f , 260 g , 260 h , 2601 , 269 j and 260 k . there may be more or less regions than illustrated in fig4 . these regions often correspond to cores , which are pre - designed circuit functions . for example , a microprocessor may contain multiple processing cores , memory cores , arithmetic cores etc . formed , by way of example , in cores 260 b , 260 g , 260 i and 260 j are serpentine signal enhancing structures 265 which are not electrically connected to any wire or device ( e . g ., transistor , diode , resistor , capacitor or inductor ) of integrated circuit chip 255 . signal enhancing structures 265 may comprise an electrical conductor , a magnetic material , or an electrically conductive magnetic material . signal enhancing structures 265 are designed to interact with the magnetic fields from gradient magnetic field coil 120 and high gauss magnet 140 of mri system 100 ( see fig1 ) to generate a more complex return rf signal . fig5 is a top view of an exemplary integrated circuit chip for preventing or detecting comparison according embodiments of the present invention . in fig5 , an integrated circuit chip 255 a ( similar to integrated circuit chip 255 of fig4 ) includes a serpentine structure 265 a connected to a destruct circuit 270 . serpentine structure 265 a acts as an inductor which generates a current when subjected to a varying magnetic field . the current may be used by destruct circuit 270 to program fuses or activate transistors to render integrated circuit 155 a inoperable or to leave a signature that can later be read to indicate if an attempt at nmr imaging \u201d has been performed on integrated circuit chip 255 a . in one example , serpentine structure 265 a is designed to not be detectable by x - ray imaging . fig6 a is a cross - sectional view of an integrated circuit containing devices to flag an unauthorized comparison attempt according embodiments of the present invention . in fig6 a , an ic 200 c includes an integrated circuit chip 235 a in a plastic package body 240 a . a set ( two sets are shown in the example of fig6 a ) of three \u201c horseshoe \u201d magnets 275 a , 275 b and 275 c are placed in body 240 a . they are aligned so respective lines ( dashed lines ) passing through the poles of each magnet are mutually orthogonal . in fig6 a , the line passing through the poles of magnets 275 a and 275 c are in planes parallel to the top surface 272 of integrated circuit chip 235 a . other pole orientations are possible . in one example , only a single horseshoe magnet is used . horseshoe shaped magnets may be replaced by other shaped magnets such as bar magnets and disc magnets . fig6 b is a cross - sectional view of the integrated circuit of fig6 a after an unauthorized comparison attempt according embodiments of the present invention . when magnets 275 a , 275 b and 275 c are subjected to the intense magnetic field generated in an nmr machine , magnets 275 a , 275 b and 275 c are pulled / pushed by that field so strongly that cracks 280 are formed in body 240 a . fig7 is a flowchart of methods of comparison according to embodiments of the present invention . the steps of the flowchart of fig7 are performed first on a known or trusted micro - electronic device and then on an unknown ( or suspect ) micro - electronic device that and essentially identical to the known micro - electronic device ( i . e ., identically designed and within fabrication specification limits ). for the integrated circuit chip specification limits define allowable variations in material and structure and include , for example , the allowable differences in metal line widths and thickness differences in metal and insulating layers . for the package of the ic specification limits define allowable variations in material and structure and include , for example , package dimensions , size of solder bumps , positions and size of wire bonds , widths and thickness of land in modules . in step 300 an ic ( or integrated circuit is placed in the mri chamber . in step 305 the high gauss magnetic field is turned on . in one example , the high gauss magnetic field has a field strength of between about a 0 . 5 tesla and about 10 tesla and is applied in the z direction . the method can know follow one of two mutually exclusive paths . the first is the path ( 1 ) through steps 310 , 315 , 320 , 325 and 330 to step 335 . the second path ( 2 ) is through steps 315 a , 320 a and 325 a to step 335 . the unknown micro - electronic device advantageously follows the same path and is subjected to the same nmr conditions as the known micro - electronic device . in step 310 , a direction ( x , y or z ) is selected . in step 315 , a gradient magnetic field of , for example , 1 tesla is applied in the selected direction and in step 320 a rf pulse is directed to the ic . in step 325 the emitted ( returned ) rf signal is detected and information describing the return rf signal ( which is also a pulse ) is stored . the duration and time of reception of the return rf signal will vary . in step 330 , if another direction of the three possible directions ( x , y , z ) is selected and the method loops back to step 310 . this loop will repeat three times , once for each direction . in path ( 2 ) in step 315 a , gradient magnetic fields are applied in the x , y , and z directions simultaneously . each gradient field may have a same or a different field strength of between about 0 . 5 tesla and about 10 tesla . steps 320 a and 325 a are similar to steps 320 and 325 respectively . in steps 315 and 315 a , optional test conditions ( i . e ., voltage bias , analog signal , digital data pattern or combinations thereof ) may be applied to the micro - electronic device . in a first example , optional test conditions are applied only during step 315 ( or 315 a ). in a second example , optional test conditions are applied the only during steps 315 and 320 ( or 315 a and 320 a ). in a third example , the optional test conditions are applied the only during steps 305 , 310 , 315 and 320 ( or 305 , 315 a and 320 a ). in a fourth example , the optional test conditions are applied during steps 305 , 310 , 315 , 320 and 325 ( or 305 , 315 a , 320 a and 325 a ). in step 335 , the ic is removed from the mri chamber and it is determined whether the micro - electronic device was a known micro - electronic device or an unknown micro - electronic device . if the micro - electronic device was a known micro - electronic device then in step 340 analyses are performed and the analyses data is stored . if the micro - electronic device was an unknown micro - electronic device then in step 345 analyses are performed and the analyses is compared to analyses previously stored from a known micro - electronic device mri \u201c images \u201d under the same mri ( and bias / pattern ) conditions . if the compare is within predefined limits the unknown micro - electronic device is validated . it does not matter whether the known or unknown micro - electronic device is run first , but the comparison cannot be performed until both known or unknown micro - electronic devices have been run and the respective data analyzed . fig8 a and 8b illustrate a first method of data analysis for comparison according to embodiments of the present invention . fig8 a is a plot of the transverse component of the returned rf signal for a known micro - electronic device 350 and an unknown micro - electronic device 355 as relative rf strength versus time . direct comparison or statistical analysis of the difference between curves 350 and 355 is performed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig8 a , the difference between curve 355 relative to curve 350 is shown as a positive time shift . the shift may be negative . fig8 b is a plot of the longitudinal component of the returned rf signal for a known micro - electronic device 360 and an unknown micro - electronic device 365 as relative rf strength versus time . direct comparison or statistical analysis of the difference between curves 360 and 365 is performed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig8 b , the difference of curve 365 relative to curve 360 is shown as a negative time shift . the shift may be positive . fig9 a and 9b illustrate a second method of data analysis for comparison according to embodiments of the present invention . fig9 a is a plot of a fast fourier transform of the returned rf signal for a known ic ( or integrated circuit ) as rf amplitude versus rf frequency . element 370 is the returned signal and elements 370 a and 370 b are harmonic artifacts of element 370 . fig9 b is a plot of a fast fourier transform of the returned rf signal for an unknown ic ( or integrated circuit ) as rf amplitude versus rf frequency . element 375 is the returned signal and elements 375 a and 3750 b are harmonic artifacts of element 370 . element 370 and 375 are statistically compared in terms of amplitude of a given frequency range to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig9 a and 9b , the amplitude of the returned rf signal has been transferred from a time domain to a frequency domain . fig1 a , 10 b and 10 c illustrate a third method of data analysis for comparison according to embodiments of the present invention . fig1 a is a plot of the returned rf signal as relative rf signal strength versus time for a known micro - electronic device . rf signal 380 has a fixed amplitude between times a and b measured from when the rf pulse signal terminated ( or other convenient time reference ). an unknown micro - electronic device may exhibit a signal that is offset ion amplitude , phase , frequency or combinations thereof . two simple examples are given in fig1 b and 10c . fig1 b is a plot of the returned rf signal as relative rf signal strength versus time for an unknown micro - electronic device having only a frequency shift . rf signal 385 a has fixed amplitude between times a \u2032 and b \u2032 measured from the same reference as a and b of fig1 a . the shift between a and a \u2032 and b and b \u2032 is statistically to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . fig1 c is a plot of the returned rf signal as relative rf signal strength versus time for an unknown micro - electronic device having only an amplitude shift . rf signal 385 b has a fixed amplitude between times a and b measured from the same reference as a and b of fig1 a . the change in amplitude between curve 380 of fig1 a and curve 385 b of fig1 c is statistically to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . it will be appreciated that a complete statistical analysis would compare time , amplitude , frequency and phase differences of the curves described in fig1 a , 10 b and 10 c . fig1 a , 11 b and 11 c illustrate a fourth method of data analysis for comparison according to embodiments of the present invention . fig1 a is a plot of the returned rf signal xy space ( at a selected z ) as a function of frequency for a known micro - electronic device . as such fig1 a is closer to what is may be considered an \u201c image .\u201d fig1 b is a plot of the returned rf signal xy space ( at the selected z ) as a function of frequency for an unknown micro - electronic device . fig1 c is a plot of the delta between the plot of fig1 a and that of fig1 b . the amount of structure ( and optionally the position of the structures ) is statistically analyzed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . fig1 c is fig1 a \u201c subtracted \u201d from fig1 a . thus the embodiments of the present invention allow for relatively quick and inexpensive validation of integrated circuit chips . the description of the embodiments of the present invention is given above for the understanding of the present invention . it will be understood that the invention is not limited to the particular embodiments described herein , but is capable of various modifications , rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention . therefore , it is intended that the following claims cover all such modifications and changes as fall within the true spirit and scope of the invention .", "category": "Physics"}
|
{"patent": "the term \u201c integrated circuit \u201d ( ic ) is defined as an integrated circuit chip and a package or module containing the ic . the term \u201c integrated circuit chip \u201d is defined as the semiconductor ( e . g ., silicon ) die containing devices such as transistors , diodes , capacitors , resisters and inductors and the wiring layers built up on the die that interconnect the devices into circuits . the term micro - electronic device includes an ic or an integrated circuit chip . in magnetic resonance imaging ( also called nuclear magnetic resonance ( nmr ) imaging )) a sample is positioned in a static magnetic field and subjected to a pulsed radio frequency ( rf ) signal to place the sample in an excited state . the magnetic field may be turned off or adjusted and a rf return signal is produced by the sample returning to a normal from the excited state is then recorded . in order to allow spatial encoding , the static magnetic field is superimposed with a gradient magnetic field . the term validation means the process by which an unknown micro - electronic device is \u201c imaged \u201d by nmr and the resulting data is compared to data from a known good or trusted micro - electronic that was nmr \u201c imaged \u201d and that the unknown ic micro - electronic device should be essentially identical ( i . e ., of identical design and within fabrication specification limits ) to . known good micro - electronic devices include those fabricated under secure conditions , those from trusted sources and those thoroughly tested and physically inspected after an nmr signature has been obtained . fig1 is a diagram of an exemplary magnetic resonance inspection system according of the present invention . in fig1 , a magnetic resonance imaging ( mri ) system 100 includes a magnet unit 105 , a signal processing unit 110 and a computer 115 . magnet unit 105 of mri system 100 includes a gradient magnetic field coil 120 , a transmit coil 125 , a sample chamber 130 , a receive coil and a high gauss magnet 140 ( e . g ., a permanent magnet , coil magnet or super - conductive magnet ). signal processing unit 110 includes a driving circuit 145 , an rf power amplifier 150 , a preamplifier 155 , a sequence memory 160 , a modulation circuit 165 , an rf oscillator 170 , a phase detector 175 and an analog - to - digital ( a / d ) converter 180 . computer 115 includes a display 185 , an input ( e . g ., a keyboard , mouse , disk drive ) and an output ( e . g ., a display unit , a printer , a disk drive ). gradient magnetic field coil 120 , transmit coil 125 , receive coil 135 and high gauss magnet 140 are disposed so as to substantially surround sample chamber 130 . high gauss magnet 140 applies a static magnetic field having a constant strength to a sample in chamber 130 . gradient magnetic field coil 120 applies gradient magnetic fields selectively to mutually orthogonal x , y and z directions ( in imaging parlance , to a slice axis , phase axis and frequency axis ). transmit coil 125 supplies a pulsed rf signal for exciting spins of atomic nuclei within a micro - electronic device 200 in sample chamber 130 . receive coil 135 detects returned rf signals from the sample in chamber 130 generated when spins of atomic nuclei within micro - electronic device 200 return to a normal state from an exited state . gradient magnetic field coil 120 , transmit coil 125 , and receive coil 135 are operatively associated with driving circuit 145 , an rf power amplifier 150 , and a preamplifier 155 , respectively . sequence memory 160 operates driving circuit 145 based on a stored pulse sequence in response to instructions from computer 115 to thereby apply gradient fields from gradient magnetic field coil 120 in specific directions . sequence memory 160 also operates a modulation circuit 165 to modulate a carrier output signal from rf oscillator 170 into a pulsed rf signal of predefined timing and envelope shape . the pulsed rf signal is applied to rf power amplifier 150 and then the amplified pulsed rf signal is applied to transmit coil 125 . preamplifier 155 amplifies the return rf signal from micro - electronic device 200 in sample chamber 130 detected at receive coil 135 . preamplifier 155 amplifies the received rf signal and sends an amplified rf signal to phase detector 175 . phase detector 175 generates an analog phase - detect signal from the amplified rf signal using a carrier output signal from rf oscillator 170 as a reference signal , and supplies the phase - detected signal to a / d converter 180 . a / d converter 180 converts the phase - detected analog signal into a digital signal , which is supplied to the computer 115 . computer 115 reads and / or processes the data from a / d converter 1801 , and includes algorithms in the form of computer instructions which when executed perform various signal analyses and statistical analyses on the stored data as described infra . results of these analyses may be displayed on output unit 195 . computer 115 can also be responsible for overall control such as receiving information supplied from input 195 from an operator . mri system 100 includes an optional tester 205 , which is connected between a socket , or probe card ( not shown ) in sample chamber 130 and computer 185 . this allows voltage bias , analog signals , digital data patterns or combinations thereof to be applied to micro - electronic device 200 during the mri process . biasing , applying signals and test patterns serves two purposes . first it results in more complex return rf signals . second , if the biasing analog signals , digital data patterns are kept secure , it is very difficult for a an unauthorized party to place a masking circuit into an unauthorized micro electronic device , the purpose of the masking circuit being is to mask the presence of the unauthorized circuit and the masking circuit in the unauthorized micro - electronic device by altering the nmr image of the unauthorized micro - electronic device to mimic that of an authorized micro - electronic device . mri system 100 shown in fig1 is provided as an example , and it will be understood that embodiments of the invention are not limited to mri system 100 shown in fig1 . it will be understood that an mri system 100 according to aspects of the invention can include additional components to those shown in fig1 or may not include every component shown in fig1 . fig2 is a cross - section through an exemplary first type of integrated circuit . in fig2 , an ic 200 a includes an integrated circuit chip 210 physically and electrically connected to a module 215 by solder bumps 220 . wires 225 in module 225 connect solder bumps 220 to solder balls 230 . module 215 may be organic ( e . g ., fiberglass ) or ceramic and wires 25 may comprise one or more wiring layers . solder balls 230 are designed for surface mounting ic 200 a directly to a printed circuit board ( pcb ). solder balls 230 may be replaced by solder columns . solder balls 230 may be replaced with pins , which can be used to mount ics into sockets , which are mounted on a pcb or mounted directly to a pcb . ic 200 a is thus an example of an ic that uses flip - chip ( or c4 ) technology . fig3 is a cross - section through an exemplary second type of integrated circuit . in fig3 , an ic 200 b includes and integrated circuit chip contained within a plastic package 240 . wire bonds 245 electrically connect integrated circuit chip 235 to leads 250 . leads 250 are designed for surface mounting ic 200 b directly to a pcb . leads 250 may be replaced with pins , which can be used to mount ics into sockets , which are mounted on a pcb or mounted directly to a pcb . the leads on some plastic packages are designed to be mounted in sockets on a pcb . ic 200 b is thus an example of plastic packaging technology . fig4 is a top view of an exemplary integrated circuit chip for enhanced comparison according embodiments of the present invention . in fig4 , an exemplary integrated circuit chip 255 includes regions 260 a , 260 b , 260 c , 260 d , 260 e , 260 f , 260 g , 260 h , 2601 , 269 j and 260 k . there may be more or less regions than illustrated in fig4 . these regions often correspond to cores , which are pre - designed circuit functions . for example , a microprocessor may contain multiple processing cores , memory cores , arithmetic cores etc . formed , by way of example , in cores 260 b , 260 g , 260 i and 260 j are serpentine signal enhancing structures 265 which are not electrically connected to any wire or device ( e . g ., transistor , diode , resistor , capacitor or inductor ) of integrated circuit chip 255 . signal enhancing structures 265 may comprise an electrical conductor , a magnetic material , or an electrically conductive magnetic material . signal enhancing structures 265 are designed to interact with the magnetic fields from gradient magnetic field coil 120 and high gauss magnet 140 of mri system 100 ( see fig1 ) to generate a more complex return rf signal . fig5 is a top view of an exemplary integrated circuit chip for preventing or detecting comparison according embodiments of the present invention . in fig5 , an integrated circuit chip 255 a ( similar to integrated circuit chip 255 of fig4 ) includes a serpentine structure 265 a connected to a destruct circuit 270 . serpentine structure 265 a acts as an inductor which generates a current when subjected to a varying magnetic field . the current may be used by destruct circuit 270 to program fuses or activate transistors to render integrated circuit 155 a inoperable or to leave a signature that can later be read to indicate if an attempt at nmr imaging \u201d has been performed on integrated circuit chip 255 a . in one example , serpentine structure 265 a is designed to not be detectable by x - ray imaging . fig6 a is a cross - sectional view of an integrated circuit containing devices to flag an unauthorized comparison attempt according embodiments of the present invention . in fig6 a , an ic 200 c includes an integrated circuit chip 235 a in a plastic package body 240 a . a set ( two sets are shown in the example of fig6 a ) of three \u201c horseshoe \u201d magnets 275 a , 275 b and 275 c are placed in body 240 a . they are aligned so respective lines ( dashed lines ) passing through the poles of each magnet are mutually orthogonal . in fig6 a , the line passing through the poles of magnets 275 a and 275 c are in planes parallel to the top surface 272 of integrated circuit chip 235 a . other pole orientations are possible . in one example , only a single horseshoe magnet is used . horseshoe shaped magnets may be replaced by other shaped magnets such as bar magnets and disc magnets . fig6 b is a cross - sectional view of the integrated circuit of fig6 a after an unauthorized comparison attempt according embodiments of the present invention . when magnets 275 a , 275 b and 275 c are subjected to the intense magnetic field generated in an nmr machine , magnets 275 a , 275 b and 275 c are pulled / pushed by that field so strongly that cracks 280 are formed in body 240 a . fig7 is a flowchart of methods of comparison according to embodiments of the present invention . the steps of the flowchart of fig7 are performed first on a known or trusted micro - electronic device and then on an unknown ( or suspect ) micro - electronic device that and essentially identical to the known micro - electronic device ( i . e ., identically designed and within fabrication specification limits ). for the integrated circuit chip specification limits define allowable variations in material and structure and include , for example , the allowable differences in metal line widths and thickness differences in metal and insulating layers . for the package of the ic specification limits define allowable variations in material and structure and include , for example , package dimensions , size of solder bumps , positions and size of wire bonds , widths and thickness of land in modules . in step 300 an ic ( or integrated circuit is placed in the mri chamber . in step 305 the high gauss magnetic field is turned on . in one example , the high gauss magnetic field has a field strength of between about a 0 . 5 tesla and about 10 tesla and is applied in the z direction . the method can know follow one of two mutually exclusive paths . the first is the path ( 1 ) through steps 310 , 315 , 320 , 325 and 330 to step 335 . the second path ( 2 ) is through steps 315 a , 320 a and 325 a to step 335 . the unknown micro - electronic device advantageously follows the same path and is subjected to the same nmr conditions as the known micro - electronic device . in step 310 , a direction ( x , y or z ) is selected . in step 315 , a gradient magnetic field of , for example , 1 tesla is applied in the selected direction and in step 320 a rf pulse is directed to the ic . in step 325 the emitted ( returned ) rf signal is detected and information describing the return rf signal ( which is also a pulse ) is stored . the duration and time of reception of the return rf signal will vary . in step 330 , if another direction of the three possible directions ( x , y , z ) is selected and the method loops back to step 310 . this loop will repeat three times , once for each direction . in path ( 2 ) in step 315 a , gradient magnetic fields are applied in the x , y , and z directions simultaneously . each gradient field may have a same or a different field strength of between about 0 . 5 tesla and about 10 tesla . steps 320 a and 325 a are similar to steps 320 and 325 respectively . in steps 315 and 315 a , optional test conditions ( i . e ., voltage bias , analog signal , digital data pattern or combinations thereof ) may be applied to the micro - electronic device . in a first example , optional test conditions are applied only during step 315 ( or 315 a ). in a second example , optional test conditions are applied the only during steps 315 and 320 ( or 315 a and 320 a ). in a third example , the optional test conditions are applied the only during steps 305 , 310 , 315 and 320 ( or 305 , 315 a and 320 a ). in a fourth example , the optional test conditions are applied during steps 305 , 310 , 315 , 320 and 325 ( or 305 , 315 a , 320 a and 325 a ). in step 335 , the ic is removed from the mri chamber and it is determined whether the micro - electronic device was a known micro - electronic device or an unknown micro - electronic device . if the micro - electronic device was a known micro - electronic device then in step 340 analyses are performed and the analyses data is stored . if the micro - electronic device was an unknown micro - electronic device then in step 345 analyses are performed and the analyses is compared to analyses previously stored from a known micro - electronic device mri \u201c images \u201d under the same mri ( and bias / pattern ) conditions . if the compare is within predefined limits the unknown micro - electronic device is validated . it does not matter whether the known or unknown micro - electronic device is run first , but the comparison cannot be performed until both known or unknown micro - electronic devices have been run and the respective data analyzed . fig8 a and 8b illustrate a first method of data analysis for comparison according to embodiments of the present invention . fig8 a is a plot of the transverse component of the returned rf signal for a known micro - electronic device 350 and an unknown micro - electronic device 355 as relative rf strength versus time . direct comparison or statistical analysis of the difference between curves 350 and 355 is performed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig8 a , the difference between curve 355 relative to curve 350 is shown as a positive time shift . the shift may be negative . fig8 b is a plot of the longitudinal component of the returned rf signal for a known micro - electronic device 360 and an unknown micro - electronic device 365 as relative rf strength versus time . direct comparison or statistical analysis of the difference between curves 360 and 365 is performed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig8 b , the difference of curve 365 relative to curve 360 is shown as a negative time shift . the shift may be positive . fig9 a and 9b illustrate a second method of data analysis for comparison according to embodiments of the present invention . fig9 a is a plot of a fast fourier transform of the returned rf signal for a known ic ( or integrated circuit ) as rf amplitude versus rf frequency . element 370 is the returned signal and elements 370 a and 370 b are harmonic artifacts of element 370 . fig9 b is a plot of a fast fourier transform of the returned rf signal for an unknown ic ( or integrated circuit ) as rf amplitude versus rf frequency . element 375 is the returned signal and elements 375 a and 3750 b are harmonic artifacts of element 370 . element 370 and 375 are statistically compared in terms of amplitude of a given frequency range to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig9 a and 9b , the amplitude of the returned rf signal has been transferred from a time domain to a frequency domain . fig1 a , 10 b and 10 c illustrate a third method of data analysis for comparison according to embodiments of the present invention . fig1 a is a plot of the returned rf signal as relative rf signal strength versus time for a known micro - electronic device . rf signal 380 has a fixed amplitude between times a and b measured from when the rf pulse signal terminated ( or other convenient time reference ). an unknown micro - electronic device may exhibit a signal that is offset ion amplitude , phase , frequency or combinations thereof . two simple examples are given in fig1 b and 10c . fig1 b is a plot of the returned rf signal as relative rf signal strength versus time for an unknown micro - electronic device having only a frequency shift . rf signal 385 a has fixed amplitude between times a \u2032 and b \u2032 measured from the same reference as a and b of fig1 a . the shift between a and a \u2032 and b and b \u2032 is statistically to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . fig1 c is a plot of the returned rf signal as relative rf signal strength versus time for an unknown micro - electronic device having only an amplitude shift . rf signal 385 b has a fixed amplitude between times a and b measured from the same reference as a and b of fig1 a . the change in amplitude between curve 380 of fig1 a and curve 385 b of fig1 c is statistically to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . it will be appreciated that a complete statistical analysis would compare time , amplitude , frequency and phase differences of the curves described in fig1 a , 10 b and 10 c . fig1 a , 11 b and 11 c illustrate a fourth method of data analysis for comparison according to embodiments of the present invention . fig1 a is a plot of the returned rf signal xy space ( at a selected z ) as a function of frequency for a known micro - electronic device . as such fig1 a is closer to what is may be considered an \u201c image .\u201d fig1 b is a plot of the returned rf signal xy space ( at the selected z ) as a function of frequency for an unknown micro - electronic device . fig1 c is a plot of the delta between the plot of fig1 a and that of fig1 b . the amount of structure ( and optionally the position of the structures ) is statistically analyzed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . fig1 c is fig1 a \u201c subtracted \u201d from fig1 a . thus the embodiments of the present invention allow for relatively quick and inexpensive validation of integrated circuit chips . the description of the embodiments of the present invention is given above for the understanding of the present invention . it will be understood that the invention is not limited to the particular embodiments described herein , but is capable of various modifications , rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention . therefore , it is intended that the following claims cover all such modifications and changes as fall within the true spirit and scope of the invention .", "category": "Chemistry; Metallurgy"}
|
Is the category the most suitable category for the given patent?
| 0.25 |
d2a533e20b88cf12a66089384d35063da6ac708b8421ee9a63cae12490153828
| 0.055908 | 0.000006 | 0.099609 | 0.000261 | 0.419922 | 0.004913 |
null |
{"patent": "the term \u201c integrated circuit \u201d ( ic ) is defined as an integrated circuit chip and a package or module containing the ic . the term \u201c integrated circuit chip \u201d is defined as the semiconductor ( e . g ., silicon ) die containing devices such as transistors , diodes , capacitors , resisters and inductors and the wiring layers built up on the die that interconnect the devices into circuits . the term micro - electronic device includes an ic or an integrated circuit chip . in magnetic resonance imaging ( also called nuclear magnetic resonance ( nmr ) imaging )) a sample is positioned in a static magnetic field and subjected to a pulsed radio frequency ( rf ) signal to place the sample in an excited state . the magnetic field may be turned off or adjusted and a rf return signal is produced by the sample returning to a normal from the excited state is then recorded . in order to allow spatial encoding , the static magnetic field is superimposed with a gradient magnetic field . the term validation means the process by which an unknown micro - electronic device is \u201c imaged \u201d by nmr and the resulting data is compared to data from a known good or trusted micro - electronic that was nmr \u201c imaged \u201d and that the unknown ic micro - electronic device should be essentially identical ( i . e ., of identical design and within fabrication specification limits ) to . known good micro - electronic devices include those fabricated under secure conditions , those from trusted sources and those thoroughly tested and physically inspected after an nmr signature has been obtained . fig1 is a diagram of an exemplary magnetic resonance inspection system according of the present invention . in fig1 , a magnetic resonance imaging ( mri ) system 100 includes a magnet unit 105 , a signal processing unit 110 and a computer 115 . magnet unit 105 of mri system 100 includes a gradient magnetic field coil 120 , a transmit coil 125 , a sample chamber 130 , a receive coil and a high gauss magnet 140 ( e . g ., a permanent magnet , coil magnet or super - conductive magnet ). signal processing unit 110 includes a driving circuit 145 , an rf power amplifier 150 , a preamplifier 155 , a sequence memory 160 , a modulation circuit 165 , an rf oscillator 170 , a phase detector 175 and an analog - to - digital ( a / d ) converter 180 . computer 115 includes a display 185 , an input ( e . g ., a keyboard , mouse , disk drive ) and an output ( e . g ., a display unit , a printer , a disk drive ). gradient magnetic field coil 120 , transmit coil 125 , receive coil 135 and high gauss magnet 140 are disposed so as to substantially surround sample chamber 130 . high gauss magnet 140 applies a static magnetic field having a constant strength to a sample in chamber 130 . gradient magnetic field coil 120 applies gradient magnetic fields selectively to mutually orthogonal x , y and z directions ( in imaging parlance , to a slice axis , phase axis and frequency axis ). transmit coil 125 supplies a pulsed rf signal for exciting spins of atomic nuclei within a micro - electronic device 200 in sample chamber 130 . receive coil 135 detects returned rf signals from the sample in chamber 130 generated when spins of atomic nuclei within micro - electronic device 200 return to a normal state from an exited state . gradient magnetic field coil 120 , transmit coil 125 , and receive coil 135 are operatively associated with driving circuit 145 , an rf power amplifier 150 , and a preamplifier 155 , respectively . sequence memory 160 operates driving circuit 145 based on a stored pulse sequence in response to instructions from computer 115 to thereby apply gradient fields from gradient magnetic field coil 120 in specific directions . sequence memory 160 also operates a modulation circuit 165 to modulate a carrier output signal from rf oscillator 170 into a pulsed rf signal of predefined timing and envelope shape . the pulsed rf signal is applied to rf power amplifier 150 and then the amplified pulsed rf signal is applied to transmit coil 125 . preamplifier 155 amplifies the return rf signal from micro - electronic device 200 in sample chamber 130 detected at receive coil 135 . preamplifier 155 amplifies the received rf signal and sends an amplified rf signal to phase detector 175 . phase detector 175 generates an analog phase - detect signal from the amplified rf signal using a carrier output signal from rf oscillator 170 as a reference signal , and supplies the phase - detected signal to a / d converter 180 . a / d converter 180 converts the phase - detected analog signal into a digital signal , which is supplied to the computer 115 . computer 115 reads and / or processes the data from a / d converter 1801 , and includes algorithms in the form of computer instructions which when executed perform various signal analyses and statistical analyses on the stored data as described infra . results of these analyses may be displayed on output unit 195 . computer 115 can also be responsible for overall control such as receiving information supplied from input 195 from an operator . mri system 100 includes an optional tester 205 , which is connected between a socket , or probe card ( not shown ) in sample chamber 130 and computer 185 . this allows voltage bias , analog signals , digital data patterns or combinations thereof to be applied to micro - electronic device 200 during the mri process . biasing , applying signals and test patterns serves two purposes . first it results in more complex return rf signals . second , if the biasing analog signals , digital data patterns are kept secure , it is very difficult for a an unauthorized party to place a masking circuit into an unauthorized micro electronic device , the purpose of the masking circuit being is to mask the presence of the unauthorized circuit and the masking circuit in the unauthorized micro - electronic device by altering the nmr image of the unauthorized micro - electronic device to mimic that of an authorized micro - electronic device . mri system 100 shown in fig1 is provided as an example , and it will be understood that embodiments of the invention are not limited to mri system 100 shown in fig1 . it will be understood that an mri system 100 according to aspects of the invention can include additional components to those shown in fig1 or may not include every component shown in fig1 . fig2 is a cross - section through an exemplary first type of integrated circuit . in fig2 , an ic 200 a includes an integrated circuit chip 210 physically and electrically connected to a module 215 by solder bumps 220 . wires 225 in module 225 connect solder bumps 220 to solder balls 230 . module 215 may be organic ( e . g ., fiberglass ) or ceramic and wires 25 may comprise one or more wiring layers . solder balls 230 are designed for surface mounting ic 200 a directly to a printed circuit board ( pcb ). solder balls 230 may be replaced by solder columns . solder balls 230 may be replaced with pins , which can be used to mount ics into sockets , which are mounted on a pcb or mounted directly to a pcb . ic 200 a is thus an example of an ic that uses flip - chip ( or c4 ) technology . fig3 is a cross - section through an exemplary second type of integrated circuit . in fig3 , an ic 200 b includes and integrated circuit chip contained within a plastic package 240 . wire bonds 245 electrically connect integrated circuit chip 235 to leads 250 . leads 250 are designed for surface mounting ic 200 b directly to a pcb . leads 250 may be replaced with pins , which can be used to mount ics into sockets , which are mounted on a pcb or mounted directly to a pcb . the leads on some plastic packages are designed to be mounted in sockets on a pcb . ic 200 b is thus an example of plastic packaging technology . fig4 is a top view of an exemplary integrated circuit chip for enhanced comparison according embodiments of the present invention . in fig4 , an exemplary integrated circuit chip 255 includes regions 260 a , 260 b , 260 c , 260 d , 260 e , 260 f , 260 g , 260 h , 2601 , 269 j and 260 k . there may be more or less regions than illustrated in fig4 . these regions often correspond to cores , which are pre - designed circuit functions . for example , a microprocessor may contain multiple processing cores , memory cores , arithmetic cores etc . formed , by way of example , in cores 260 b , 260 g , 260 i and 260 j are serpentine signal enhancing structures 265 which are not electrically connected to any wire or device ( e . g ., transistor , diode , resistor , capacitor or inductor ) of integrated circuit chip 255 . signal enhancing structures 265 may comprise an electrical conductor , a magnetic material , or an electrically conductive magnetic material . signal enhancing structures 265 are designed to interact with the magnetic fields from gradient magnetic field coil 120 and high gauss magnet 140 of mri system 100 ( see fig1 ) to generate a more complex return rf signal . fig5 is a top view of an exemplary integrated circuit chip for preventing or detecting comparison according embodiments of the present invention . in fig5 , an integrated circuit chip 255 a ( similar to integrated circuit chip 255 of fig4 ) includes a serpentine structure 265 a connected to a destruct circuit 270 . serpentine structure 265 a acts as an inductor which generates a current when subjected to a varying magnetic field . the current may be used by destruct circuit 270 to program fuses or activate transistors to render integrated circuit 155 a inoperable or to leave a signature that can later be read to indicate if an attempt at nmr imaging \u201d has been performed on integrated circuit chip 255 a . in one example , serpentine structure 265 a is designed to not be detectable by x - ray imaging . fig6 a is a cross - sectional view of an integrated circuit containing devices to flag an unauthorized comparison attempt according embodiments of the present invention . in fig6 a , an ic 200 c includes an integrated circuit chip 235 a in a plastic package body 240 a . a set ( two sets are shown in the example of fig6 a ) of three \u201c horseshoe \u201d magnets 275 a , 275 b and 275 c are placed in body 240 a . they are aligned so respective lines ( dashed lines ) passing through the poles of each magnet are mutually orthogonal . in fig6 a , the line passing through the poles of magnets 275 a and 275 c are in planes parallel to the top surface 272 of integrated circuit chip 235 a . other pole orientations are possible . in one example , only a single horseshoe magnet is used . horseshoe shaped magnets may be replaced by other shaped magnets such as bar magnets and disc magnets . fig6 b is a cross - sectional view of the integrated circuit of fig6 a after an unauthorized comparison attempt according embodiments of the present invention . when magnets 275 a , 275 b and 275 c are subjected to the intense magnetic field generated in an nmr machine , magnets 275 a , 275 b and 275 c are pulled / pushed by that field so strongly that cracks 280 are formed in body 240 a . fig7 is a flowchart of methods of comparison according to embodiments of the present invention . the steps of the flowchart of fig7 are performed first on a known or trusted micro - electronic device and then on an unknown ( or suspect ) micro - electronic device that and essentially identical to the known micro - electronic device ( i . e ., identically designed and within fabrication specification limits ). for the integrated circuit chip specification limits define allowable variations in material and structure and include , for example , the allowable differences in metal line widths and thickness differences in metal and insulating layers . for the package of the ic specification limits define allowable variations in material and structure and include , for example , package dimensions , size of solder bumps , positions and size of wire bonds , widths and thickness of land in modules . in step 300 an ic ( or integrated circuit is placed in the mri chamber . in step 305 the high gauss magnetic field is turned on . in one example , the high gauss magnetic field has a field strength of between about a 0 . 5 tesla and about 10 tesla and is applied in the z direction . the method can know follow one of two mutually exclusive paths . the first is the path ( 1 ) through steps 310 , 315 , 320 , 325 and 330 to step 335 . the second path ( 2 ) is through steps 315 a , 320 a and 325 a to step 335 . the unknown micro - electronic device advantageously follows the same path and is subjected to the same nmr conditions as the known micro - electronic device . in step 310 , a direction ( x , y or z ) is selected . in step 315 , a gradient magnetic field of , for example , 1 tesla is applied in the selected direction and in step 320 a rf pulse is directed to the ic . in step 325 the emitted ( returned ) rf signal is detected and information describing the return rf signal ( which is also a pulse ) is stored . the duration and time of reception of the return rf signal will vary . in step 330 , if another direction of the three possible directions ( x , y , z ) is selected and the method loops back to step 310 . this loop will repeat three times , once for each direction . in path ( 2 ) in step 315 a , gradient magnetic fields are applied in the x , y , and z directions simultaneously . each gradient field may have a same or a different field strength of between about 0 . 5 tesla and about 10 tesla . steps 320 a and 325 a are similar to steps 320 and 325 respectively . in steps 315 and 315 a , optional test conditions ( i . e ., voltage bias , analog signal , digital data pattern or combinations thereof ) may be applied to the micro - electronic device . in a first example , optional test conditions are applied only during step 315 ( or 315 a ). in a second example , optional test conditions are applied the only during steps 315 and 320 ( or 315 a and 320 a ). in a third example , the optional test conditions are applied the only during steps 305 , 310 , 315 and 320 ( or 305 , 315 a and 320 a ). in a fourth example , the optional test conditions are applied during steps 305 , 310 , 315 , 320 and 325 ( or 305 , 315 a , 320 a and 325 a ). in step 335 , the ic is removed from the mri chamber and it is determined whether the micro - electronic device was a known micro - electronic device or an unknown micro - electronic device . if the micro - electronic device was a known micro - electronic device then in step 340 analyses are performed and the analyses data is stored . if the micro - electronic device was an unknown micro - electronic device then in step 345 analyses are performed and the analyses is compared to analyses previously stored from a known micro - electronic device mri \u201c images \u201d under the same mri ( and bias / pattern ) conditions . if the compare is within predefined limits the unknown micro - electronic device is validated . it does not matter whether the known or unknown micro - electronic device is run first , but the comparison cannot be performed until both known or unknown micro - electronic devices have been run and the respective data analyzed . fig8 a and 8b illustrate a first method of data analysis for comparison according to embodiments of the present invention . fig8 a is a plot of the transverse component of the returned rf signal for a known micro - electronic device 350 and an unknown micro - electronic device 355 as relative rf strength versus time . direct comparison or statistical analysis of the difference between curves 350 and 355 is performed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig8 a , the difference between curve 355 relative to curve 350 is shown as a positive time shift . the shift may be negative . fig8 b is a plot of the longitudinal component of the returned rf signal for a known micro - electronic device 360 and an unknown micro - electronic device 365 as relative rf strength versus time . direct comparison or statistical analysis of the difference between curves 360 and 365 is performed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig8 b , the difference of curve 365 relative to curve 360 is shown as a negative time shift . the shift may be positive . fig9 a and 9b illustrate a second method of data analysis for comparison according to embodiments of the present invention . fig9 a is a plot of a fast fourier transform of the returned rf signal for a known ic ( or integrated circuit ) as rf amplitude versus rf frequency . element 370 is the returned signal and elements 370 a and 370 b are harmonic artifacts of element 370 . fig9 b is a plot of a fast fourier transform of the returned rf signal for an unknown ic ( or integrated circuit ) as rf amplitude versus rf frequency . element 375 is the returned signal and elements 375 a and 3750 b are harmonic artifacts of element 370 . element 370 and 375 are statistically compared in terms of amplitude of a given frequency range to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig9 a and 9b , the amplitude of the returned rf signal has been transferred from a time domain to a frequency domain . fig1 a , 10 b and 10 c illustrate a third method of data analysis for comparison according to embodiments of the present invention . fig1 a is a plot of the returned rf signal as relative rf signal strength versus time for a known micro - electronic device . rf signal 380 has a fixed amplitude between times a and b measured from when the rf pulse signal terminated ( or other convenient time reference ). an unknown micro - electronic device may exhibit a signal that is offset ion amplitude , phase , frequency or combinations thereof . two simple examples are given in fig1 b and 10c . fig1 b is a plot of the returned rf signal as relative rf signal strength versus time for an unknown micro - electronic device having only a frequency shift . rf signal 385 a has fixed amplitude between times a \u2032 and b \u2032 measured from the same reference as a and b of fig1 a . the shift between a and a \u2032 and b and b \u2032 is statistically to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . fig1 c is a plot of the returned rf signal as relative rf signal strength versus time for an unknown micro - electronic device having only an amplitude shift . rf signal 385 b has a fixed amplitude between times a and b measured from the same reference as a and b of fig1 a . the change in amplitude between curve 380 of fig1 a and curve 385 b of fig1 c is statistically to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . it will be appreciated that a complete statistical analysis would compare time , amplitude , frequency and phase differences of the curves described in fig1 a , 10 b and 10 c . fig1 a , 11 b and 11 c illustrate a fourth method of data analysis for comparison according to embodiments of the present invention . fig1 a is a plot of the returned rf signal xy space ( at a selected z ) as a function of frequency for a known micro - electronic device . as such fig1 a is closer to what is may be considered an \u201c image .\u201d fig1 b is a plot of the returned rf signal xy space ( at the selected z ) as a function of frequency for an unknown micro - electronic device . fig1 c is a plot of the delta between the plot of fig1 a and that of fig1 b . the amount of structure ( and optionally the position of the structures ) is statistically analyzed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . fig1 c is fig1 a \u201c subtracted \u201d from fig1 a . thus the embodiments of the present invention allow for relatively quick and inexpensive validation of integrated circuit chips . the description of the embodiments of the present invention is given above for the understanding of the present invention . it will be understood that the invention is not limited to the particular embodiments described herein , but is capable of various modifications , rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention . therefore , it is intended that the following claims cover all such modifications and changes as fall within the true spirit and scope of the invention .", "category": "Physics"}
|
{"category": "Textiles; Paper", "patent": "the term \u201c integrated circuit \u201d ( ic ) is defined as an integrated circuit chip and a package or module containing the ic . the term \u201c integrated circuit chip \u201d is defined as the semiconductor ( e . g ., silicon ) die containing devices such as transistors , diodes , capacitors , resisters and inductors and the wiring layers built up on the die that interconnect the devices into circuits . the term micro - electronic device includes an ic or an integrated circuit chip . in magnetic resonance imaging ( also called nuclear magnetic resonance ( nmr ) imaging )) a sample is positioned in a static magnetic field and subjected to a pulsed radio frequency ( rf ) signal to place the sample in an excited state . the magnetic field may be turned off or adjusted and a rf return signal is produced by the sample returning to a normal from the excited state is then recorded . in order to allow spatial encoding , the static magnetic field is superimposed with a gradient magnetic field . the term validation means the process by which an unknown micro - electronic device is \u201c imaged \u201d by nmr and the resulting data is compared to data from a known good or trusted micro - electronic that was nmr \u201c imaged \u201d and that the unknown ic micro - electronic device should be essentially identical ( i . e ., of identical design and within fabrication specification limits ) to . known good micro - electronic devices include those fabricated under secure conditions , those from trusted sources and those thoroughly tested and physically inspected after an nmr signature has been obtained . fig1 is a diagram of an exemplary magnetic resonance inspection system according of the present invention . in fig1 , a magnetic resonance imaging ( mri ) system 100 includes a magnet unit 105 , a signal processing unit 110 and a computer 115 . magnet unit 105 of mri system 100 includes a gradient magnetic field coil 120 , a transmit coil 125 , a sample chamber 130 , a receive coil and a high gauss magnet 140 ( e . g ., a permanent magnet , coil magnet or super - conductive magnet ). signal processing unit 110 includes a driving circuit 145 , an rf power amplifier 150 , a preamplifier 155 , a sequence memory 160 , a modulation circuit 165 , an rf oscillator 170 , a phase detector 175 and an analog - to - digital ( a / d ) converter 180 . computer 115 includes a display 185 , an input ( e . g ., a keyboard , mouse , disk drive ) and an output ( e . g ., a display unit , a printer , a disk drive ). gradient magnetic field coil 120 , transmit coil 125 , receive coil 135 and high gauss magnet 140 are disposed so as to substantially surround sample chamber 130 . high gauss magnet 140 applies a static magnetic field having a constant strength to a sample in chamber 130 . gradient magnetic field coil 120 applies gradient magnetic fields selectively to mutually orthogonal x , y and z directions ( in imaging parlance , to a slice axis , phase axis and frequency axis ). transmit coil 125 supplies a pulsed rf signal for exciting spins of atomic nuclei within a micro - electronic device 200 in sample chamber 130 . receive coil 135 detects returned rf signals from the sample in chamber 130 generated when spins of atomic nuclei within micro - electronic device 200 return to a normal state from an exited state . gradient magnetic field coil 120 , transmit coil 125 , and receive coil 135 are operatively associated with driving circuit 145 , an rf power amplifier 150 , and a preamplifier 155 , respectively . sequence memory 160 operates driving circuit 145 based on a stored pulse sequence in response to instructions from computer 115 to thereby apply gradient fields from gradient magnetic field coil 120 in specific directions . sequence memory 160 also operates a modulation circuit 165 to modulate a carrier output signal from rf oscillator 170 into a pulsed rf signal of predefined timing and envelope shape . the pulsed rf signal is applied to rf power amplifier 150 and then the amplified pulsed rf signal is applied to transmit coil 125 . preamplifier 155 amplifies the return rf signal from micro - electronic device 200 in sample chamber 130 detected at receive coil 135 . preamplifier 155 amplifies the received rf signal and sends an amplified rf signal to phase detector 175 . phase detector 175 generates an analog phase - detect signal from the amplified rf signal using a carrier output signal from rf oscillator 170 as a reference signal , and supplies the phase - detected signal to a / d converter 180 . a / d converter 180 converts the phase - detected analog signal into a digital signal , which is supplied to the computer 115 . computer 115 reads and / or processes the data from a / d converter 1801 , and includes algorithms in the form of computer instructions which when executed perform various signal analyses and statistical analyses on the stored data as described infra . results of these analyses may be displayed on output unit 195 . computer 115 can also be responsible for overall control such as receiving information supplied from input 195 from an operator . mri system 100 includes an optional tester 205 , which is connected between a socket , or probe card ( not shown ) in sample chamber 130 and computer 185 . this allows voltage bias , analog signals , digital data patterns or combinations thereof to be applied to micro - electronic device 200 during the mri process . biasing , applying signals and test patterns serves two purposes . first it results in more complex return rf signals . second , if the biasing analog signals , digital data patterns are kept secure , it is very difficult for a an unauthorized party to place a masking circuit into an unauthorized micro electronic device , the purpose of the masking circuit being is to mask the presence of the unauthorized circuit and the masking circuit in the unauthorized micro - electronic device by altering the nmr image of the unauthorized micro - electronic device to mimic that of an authorized micro - electronic device . mri system 100 shown in fig1 is provided as an example , and it will be understood that embodiments of the invention are not limited to mri system 100 shown in fig1 . it will be understood that an mri system 100 according to aspects of the invention can include additional components to those shown in fig1 or may not include every component shown in fig1 . fig2 is a cross - section through an exemplary first type of integrated circuit . in fig2 , an ic 200 a includes an integrated circuit chip 210 physically and electrically connected to a module 215 by solder bumps 220 . wires 225 in module 225 connect solder bumps 220 to solder balls 230 . module 215 may be organic ( e . g ., fiberglass ) or ceramic and wires 25 may comprise one or more wiring layers . solder balls 230 are designed for surface mounting ic 200 a directly to a printed circuit board ( pcb ). solder balls 230 may be replaced by solder columns . solder balls 230 may be replaced with pins , which can be used to mount ics into sockets , which are mounted on a pcb or mounted directly to a pcb . ic 200 a is thus an example of an ic that uses flip - chip ( or c4 ) technology . fig3 is a cross - section through an exemplary second type of integrated circuit . in fig3 , an ic 200 b includes and integrated circuit chip contained within a plastic package 240 . wire bonds 245 electrically connect integrated circuit chip 235 to leads 250 . leads 250 are designed for surface mounting ic 200 b directly to a pcb . leads 250 may be replaced with pins , which can be used to mount ics into sockets , which are mounted on a pcb or mounted directly to a pcb . the leads on some plastic packages are designed to be mounted in sockets on a pcb . ic 200 b is thus an example of plastic packaging technology . fig4 is a top view of an exemplary integrated circuit chip for enhanced comparison according embodiments of the present invention . in fig4 , an exemplary integrated circuit chip 255 includes regions 260 a , 260 b , 260 c , 260 d , 260 e , 260 f , 260 g , 260 h , 2601 , 269 j and 260 k . there may be more or less regions than illustrated in fig4 . these regions often correspond to cores , which are pre - designed circuit functions . for example , a microprocessor may contain multiple processing cores , memory cores , arithmetic cores etc . formed , by way of example , in cores 260 b , 260 g , 260 i and 260 j are serpentine signal enhancing structures 265 which are not electrically connected to any wire or device ( e . g ., transistor , diode , resistor , capacitor or inductor ) of integrated circuit chip 255 . signal enhancing structures 265 may comprise an electrical conductor , a magnetic material , or an electrically conductive magnetic material . signal enhancing structures 265 are designed to interact with the magnetic fields from gradient magnetic field coil 120 and high gauss magnet 140 of mri system 100 ( see fig1 ) to generate a more complex return rf signal . fig5 is a top view of an exemplary integrated circuit chip for preventing or detecting comparison according embodiments of the present invention . in fig5 , an integrated circuit chip 255 a ( similar to integrated circuit chip 255 of fig4 ) includes a serpentine structure 265 a connected to a destruct circuit 270 . serpentine structure 265 a acts as an inductor which generates a current when subjected to a varying magnetic field . the current may be used by destruct circuit 270 to program fuses or activate transistors to render integrated circuit 155 a inoperable or to leave a signature that can later be read to indicate if an attempt at nmr imaging \u201d has been performed on integrated circuit chip 255 a . in one example , serpentine structure 265 a is designed to not be detectable by x - ray imaging . fig6 a is a cross - sectional view of an integrated circuit containing devices to flag an unauthorized comparison attempt according embodiments of the present invention . in fig6 a , an ic 200 c includes an integrated circuit chip 235 a in a plastic package body 240 a . a set ( two sets are shown in the example of fig6 a ) of three \u201c horseshoe \u201d magnets 275 a , 275 b and 275 c are placed in body 240 a . they are aligned so respective lines ( dashed lines ) passing through the poles of each magnet are mutually orthogonal . in fig6 a , the line passing through the poles of magnets 275 a and 275 c are in planes parallel to the top surface 272 of integrated circuit chip 235 a . other pole orientations are possible . in one example , only a single horseshoe magnet is used . horseshoe shaped magnets may be replaced by other shaped magnets such as bar magnets and disc magnets . fig6 b is a cross - sectional view of the integrated circuit of fig6 a after an unauthorized comparison attempt according embodiments of the present invention . when magnets 275 a , 275 b and 275 c are subjected to the intense magnetic field generated in an nmr machine , magnets 275 a , 275 b and 275 c are pulled / pushed by that field so strongly that cracks 280 are formed in body 240 a . fig7 is a flowchart of methods of comparison according to embodiments of the present invention . the steps of the flowchart of fig7 are performed first on a known or trusted micro - electronic device and then on an unknown ( or suspect ) micro - electronic device that and essentially identical to the known micro - electronic device ( i . e ., identically designed and within fabrication specification limits ). for the integrated circuit chip specification limits define allowable variations in material and structure and include , for example , the allowable differences in metal line widths and thickness differences in metal and insulating layers . for the package of the ic specification limits define allowable variations in material and structure and include , for example , package dimensions , size of solder bumps , positions and size of wire bonds , widths and thickness of land in modules . in step 300 an ic ( or integrated circuit is placed in the mri chamber . in step 305 the high gauss magnetic field is turned on . in one example , the high gauss magnetic field has a field strength of between about a 0 . 5 tesla and about 10 tesla and is applied in the z direction . the method can know follow one of two mutually exclusive paths . the first is the path ( 1 ) through steps 310 , 315 , 320 , 325 and 330 to step 335 . the second path ( 2 ) is through steps 315 a , 320 a and 325 a to step 335 . the unknown micro - electronic device advantageously follows the same path and is subjected to the same nmr conditions as the known micro - electronic device . in step 310 , a direction ( x , y or z ) is selected . in step 315 , a gradient magnetic field of , for example , 1 tesla is applied in the selected direction and in step 320 a rf pulse is directed to the ic . in step 325 the emitted ( returned ) rf signal is detected and information describing the return rf signal ( which is also a pulse ) is stored . the duration and time of reception of the return rf signal will vary . in step 330 , if another direction of the three possible directions ( x , y , z ) is selected and the method loops back to step 310 . this loop will repeat three times , once for each direction . in path ( 2 ) in step 315 a , gradient magnetic fields are applied in the x , y , and z directions simultaneously . each gradient field may have a same or a different field strength of between about 0 . 5 tesla and about 10 tesla . steps 320 a and 325 a are similar to steps 320 and 325 respectively . in steps 315 and 315 a , optional test conditions ( i . e ., voltage bias , analog signal , digital data pattern or combinations thereof ) may be applied to the micro - electronic device . in a first example , optional test conditions are applied only during step 315 ( or 315 a ). in a second example , optional test conditions are applied the only during steps 315 and 320 ( or 315 a and 320 a ). in a third example , the optional test conditions are applied the only during steps 305 , 310 , 315 and 320 ( or 305 , 315 a and 320 a ). in a fourth example , the optional test conditions are applied during steps 305 , 310 , 315 , 320 and 325 ( or 305 , 315 a , 320 a and 325 a ). in step 335 , the ic is removed from the mri chamber and it is determined whether the micro - electronic device was a known micro - electronic device or an unknown micro - electronic device . if the micro - electronic device was a known micro - electronic device then in step 340 analyses are performed and the analyses data is stored . if the micro - electronic device was an unknown micro - electronic device then in step 345 analyses are performed and the analyses is compared to analyses previously stored from a known micro - electronic device mri \u201c images \u201d under the same mri ( and bias / pattern ) conditions . if the compare is within predefined limits the unknown micro - electronic device is validated . it does not matter whether the known or unknown micro - electronic device is run first , but the comparison cannot be performed until both known or unknown micro - electronic devices have been run and the respective data analyzed . fig8 a and 8b illustrate a first method of data analysis for comparison according to embodiments of the present invention . fig8 a is a plot of the transverse component of the returned rf signal for a known micro - electronic device 350 and an unknown micro - electronic device 355 as relative rf strength versus time . direct comparison or statistical analysis of the difference between curves 350 and 355 is performed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig8 a , the difference between curve 355 relative to curve 350 is shown as a positive time shift . the shift may be negative . fig8 b is a plot of the longitudinal component of the returned rf signal for a known micro - electronic device 360 and an unknown micro - electronic device 365 as relative rf strength versus time . direct comparison or statistical analysis of the difference between curves 360 and 365 is performed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig8 b , the difference of curve 365 relative to curve 360 is shown as a negative time shift . the shift may be positive . fig9 a and 9b illustrate a second method of data analysis for comparison according to embodiments of the present invention . fig9 a is a plot of a fast fourier transform of the returned rf signal for a known ic ( or integrated circuit ) as rf amplitude versus rf frequency . element 370 is the returned signal and elements 370 a and 370 b are harmonic artifacts of element 370 . fig9 b is a plot of a fast fourier transform of the returned rf signal for an unknown ic ( or integrated circuit ) as rf amplitude versus rf frequency . element 375 is the returned signal and elements 375 a and 3750 b are harmonic artifacts of element 370 . element 370 and 375 are statistically compared in terms of amplitude of a given frequency range to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig9 a and 9b , the amplitude of the returned rf signal has been transferred from a time domain to a frequency domain . fig1 a , 10 b and 10 c illustrate a third method of data analysis for comparison according to embodiments of the present invention . fig1 a is a plot of the returned rf signal as relative rf signal strength versus time for a known micro - electronic device . rf signal 380 has a fixed amplitude between times a and b measured from when the rf pulse signal terminated ( or other convenient time reference ). an unknown micro - electronic device may exhibit a signal that is offset ion amplitude , phase , frequency or combinations thereof . two simple examples are given in fig1 b and 10c . fig1 b is a plot of the returned rf signal as relative rf signal strength versus time for an unknown micro - electronic device having only a frequency shift . rf signal 385 a has fixed amplitude between times a \u2032 and b \u2032 measured from the same reference as a and b of fig1 a . the shift between a and a \u2032 and b and b \u2032 is statistically to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . fig1 c is a plot of the returned rf signal as relative rf signal strength versus time for an unknown micro - electronic device having only an amplitude shift . rf signal 385 b has a fixed amplitude between times a and b measured from the same reference as a and b of fig1 a . the change in amplitude between curve 380 of fig1 a and curve 385 b of fig1 c is statistically to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . it will be appreciated that a complete statistical analysis would compare time , amplitude , frequency and phase differences of the curves described in fig1 a , 10 b and 10 c . fig1 a , 11 b and 11 c illustrate a fourth method of data analysis for comparison according to embodiments of the present invention . fig1 a is a plot of the returned rf signal xy space ( at a selected z ) as a function of frequency for a known micro - electronic device . as such fig1 a is closer to what is may be considered an \u201c image .\u201d fig1 b is a plot of the returned rf signal xy space ( at the selected z ) as a function of frequency for an unknown micro - electronic device . fig1 c is a plot of the delta between the plot of fig1 a and that of fig1 b . the amount of structure ( and optionally the position of the structures ) is statistically analyzed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . fig1 c is fig1 a \u201c subtracted \u201d from fig1 a . thus the embodiments of the present invention allow for relatively quick and inexpensive validation of integrated circuit chips . the description of the embodiments of the present invention is given above for the understanding of the present invention . it will be understood that the invention is not limited to the particular embodiments described herein , but is capable of various modifications , rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention . therefore , it is intended that the following claims cover all such modifications and changes as fall within the true spirit and scope of the invention ."}
|
Is the category the most suitable category for the given patent?
| 0.25 |
d2a533e20b88cf12a66089384d35063da6ac708b8421ee9a63cae12490153828
| 0.055908 | 0.004761 | 0.099609 | 0.000216 | 0.419922 | 0.00592 |
null |
{"category": "Physics", "patent": "the term \u201c integrated circuit \u201d ( ic ) is defined as an integrated circuit chip and a package or module containing the ic . the term \u201c integrated circuit chip \u201d is defined as the semiconductor ( e . g ., silicon ) die containing devices such as transistors , diodes , capacitors , resisters and inductors and the wiring layers built up on the die that interconnect the devices into circuits . the term micro - electronic device includes an ic or an integrated circuit chip . in magnetic resonance imaging ( also called nuclear magnetic resonance ( nmr ) imaging )) a sample is positioned in a static magnetic field and subjected to a pulsed radio frequency ( rf ) signal to place the sample in an excited state . the magnetic field may be turned off or adjusted and a rf return signal is produced by the sample returning to a normal from the excited state is then recorded . in order to allow spatial encoding , the static magnetic field is superimposed with a gradient magnetic field . the term validation means the process by which an unknown micro - electronic device is \u201c imaged \u201d by nmr and the resulting data is compared to data from a known good or trusted micro - electronic that was nmr \u201c imaged \u201d and that the unknown ic micro - electronic device should be essentially identical ( i . e ., of identical design and within fabrication specification limits ) to . known good micro - electronic devices include those fabricated under secure conditions , those from trusted sources and those thoroughly tested and physically inspected after an nmr signature has been obtained . fig1 is a diagram of an exemplary magnetic resonance inspection system according of the present invention . in fig1 , a magnetic resonance imaging ( mri ) system 100 includes a magnet unit 105 , a signal processing unit 110 and a computer 115 . magnet unit 105 of mri system 100 includes a gradient magnetic field coil 120 , a transmit coil 125 , a sample chamber 130 , a receive coil and a high gauss magnet 140 ( e . g ., a permanent magnet , coil magnet or super - conductive magnet ). signal processing unit 110 includes a driving circuit 145 , an rf power amplifier 150 , a preamplifier 155 , a sequence memory 160 , a modulation circuit 165 , an rf oscillator 170 , a phase detector 175 and an analog - to - digital ( a / d ) converter 180 . computer 115 includes a display 185 , an input ( e . g ., a keyboard , mouse , disk drive ) and an output ( e . g ., a display unit , a printer , a disk drive ). gradient magnetic field coil 120 , transmit coil 125 , receive coil 135 and high gauss magnet 140 are disposed so as to substantially surround sample chamber 130 . high gauss magnet 140 applies a static magnetic field having a constant strength to a sample in chamber 130 . gradient magnetic field coil 120 applies gradient magnetic fields selectively to mutually orthogonal x , y and z directions ( in imaging parlance , to a slice axis , phase axis and frequency axis ). transmit coil 125 supplies a pulsed rf signal for exciting spins of atomic nuclei within a micro - electronic device 200 in sample chamber 130 . receive coil 135 detects returned rf signals from the sample in chamber 130 generated when spins of atomic nuclei within micro - electronic device 200 return to a normal state from an exited state . gradient magnetic field coil 120 , transmit coil 125 , and receive coil 135 are operatively associated with driving circuit 145 , an rf power amplifier 150 , and a preamplifier 155 , respectively . sequence memory 160 operates driving circuit 145 based on a stored pulse sequence in response to instructions from computer 115 to thereby apply gradient fields from gradient magnetic field coil 120 in specific directions . sequence memory 160 also operates a modulation circuit 165 to modulate a carrier output signal from rf oscillator 170 into a pulsed rf signal of predefined timing and envelope shape . the pulsed rf signal is applied to rf power amplifier 150 and then the amplified pulsed rf signal is applied to transmit coil 125 . preamplifier 155 amplifies the return rf signal from micro - electronic device 200 in sample chamber 130 detected at receive coil 135 . preamplifier 155 amplifies the received rf signal and sends an amplified rf signal to phase detector 175 . phase detector 175 generates an analog phase - detect signal from the amplified rf signal using a carrier output signal from rf oscillator 170 as a reference signal , and supplies the phase - detected signal to a / d converter 180 . a / d converter 180 converts the phase - detected analog signal into a digital signal , which is supplied to the computer 115 . computer 115 reads and / or processes the data from a / d converter 1801 , and includes algorithms in the form of computer instructions which when executed perform various signal analyses and statistical analyses on the stored data as described infra . results of these analyses may be displayed on output unit 195 . computer 115 can also be responsible for overall control such as receiving information supplied from input 195 from an operator . mri system 100 includes an optional tester 205 , which is connected between a socket , or probe card ( not shown ) in sample chamber 130 and computer 185 . this allows voltage bias , analog signals , digital data patterns or combinations thereof to be applied to micro - electronic device 200 during the mri process . biasing , applying signals and test patterns serves two purposes . first it results in more complex return rf signals . second , if the biasing analog signals , digital data patterns are kept secure , it is very difficult for a an unauthorized party to place a masking circuit into an unauthorized micro electronic device , the purpose of the masking circuit being is to mask the presence of the unauthorized circuit and the masking circuit in the unauthorized micro - electronic device by altering the nmr image of the unauthorized micro - electronic device to mimic that of an authorized micro - electronic device . mri system 100 shown in fig1 is provided as an example , and it will be understood that embodiments of the invention are not limited to mri system 100 shown in fig1 . it will be understood that an mri system 100 according to aspects of the invention can include additional components to those shown in fig1 or may not include every component shown in fig1 . fig2 is a cross - section through an exemplary first type of integrated circuit . in fig2 , an ic 200 a includes an integrated circuit chip 210 physically and electrically connected to a module 215 by solder bumps 220 . wires 225 in module 225 connect solder bumps 220 to solder balls 230 . module 215 may be organic ( e . g ., fiberglass ) or ceramic and wires 25 may comprise one or more wiring layers . solder balls 230 are designed for surface mounting ic 200 a directly to a printed circuit board ( pcb ). solder balls 230 may be replaced by solder columns . solder balls 230 may be replaced with pins , which can be used to mount ics into sockets , which are mounted on a pcb or mounted directly to a pcb . ic 200 a is thus an example of an ic that uses flip - chip ( or c4 ) technology . fig3 is a cross - section through an exemplary second type of integrated circuit . in fig3 , an ic 200 b includes and integrated circuit chip contained within a plastic package 240 . wire bonds 245 electrically connect integrated circuit chip 235 to leads 250 . leads 250 are designed for surface mounting ic 200 b directly to a pcb . leads 250 may be replaced with pins , which can be used to mount ics into sockets , which are mounted on a pcb or mounted directly to a pcb . the leads on some plastic packages are designed to be mounted in sockets on a pcb . ic 200 b is thus an example of plastic packaging technology . fig4 is a top view of an exemplary integrated circuit chip for enhanced comparison according embodiments of the present invention . in fig4 , an exemplary integrated circuit chip 255 includes regions 260 a , 260 b , 260 c , 260 d , 260 e , 260 f , 260 g , 260 h , 2601 , 269 j and 260 k . there may be more or less regions than illustrated in fig4 . these regions often correspond to cores , which are pre - designed circuit functions . for example , a microprocessor may contain multiple processing cores , memory cores , arithmetic cores etc . formed , by way of example , in cores 260 b , 260 g , 260 i and 260 j are serpentine signal enhancing structures 265 which are not electrically connected to any wire or device ( e . g ., transistor , diode , resistor , capacitor or inductor ) of integrated circuit chip 255 . signal enhancing structures 265 may comprise an electrical conductor , a magnetic material , or an electrically conductive magnetic material . signal enhancing structures 265 are designed to interact with the magnetic fields from gradient magnetic field coil 120 and high gauss magnet 140 of mri system 100 ( see fig1 ) to generate a more complex return rf signal . fig5 is a top view of an exemplary integrated circuit chip for preventing or detecting comparison according embodiments of the present invention . in fig5 , an integrated circuit chip 255 a ( similar to integrated circuit chip 255 of fig4 ) includes a serpentine structure 265 a connected to a destruct circuit 270 . serpentine structure 265 a acts as an inductor which generates a current when subjected to a varying magnetic field . the current may be used by destruct circuit 270 to program fuses or activate transistors to render integrated circuit 155 a inoperable or to leave a signature that can later be read to indicate if an attempt at nmr imaging \u201d has been performed on integrated circuit chip 255 a . in one example , serpentine structure 265 a is designed to not be detectable by x - ray imaging . fig6 a is a cross - sectional view of an integrated circuit containing devices to flag an unauthorized comparison attempt according embodiments of the present invention . in fig6 a , an ic 200 c includes an integrated circuit chip 235 a in a plastic package body 240 a . a set ( two sets are shown in the example of fig6 a ) of three \u201c horseshoe \u201d magnets 275 a , 275 b and 275 c are placed in body 240 a . they are aligned so respective lines ( dashed lines ) passing through the poles of each magnet are mutually orthogonal . in fig6 a , the line passing through the poles of magnets 275 a and 275 c are in planes parallel to the top surface 272 of integrated circuit chip 235 a . other pole orientations are possible . in one example , only a single horseshoe magnet is used . horseshoe shaped magnets may be replaced by other shaped magnets such as bar magnets and disc magnets . fig6 b is a cross - sectional view of the integrated circuit of fig6 a after an unauthorized comparison attempt according embodiments of the present invention . when magnets 275 a , 275 b and 275 c are subjected to the intense magnetic field generated in an nmr machine , magnets 275 a , 275 b and 275 c are pulled / pushed by that field so strongly that cracks 280 are formed in body 240 a . fig7 is a flowchart of methods of comparison according to embodiments of the present invention . the steps of the flowchart of fig7 are performed first on a known or trusted micro - electronic device and then on an unknown ( or suspect ) micro - electronic device that and essentially identical to the known micro - electronic device ( i . e ., identically designed and within fabrication specification limits ). for the integrated circuit chip specification limits define allowable variations in material and structure and include , for example , the allowable differences in metal line widths and thickness differences in metal and insulating layers . for the package of the ic specification limits define allowable variations in material and structure and include , for example , package dimensions , size of solder bumps , positions and size of wire bonds , widths and thickness of land in modules . in step 300 an ic ( or integrated circuit is placed in the mri chamber . in step 305 the high gauss magnetic field is turned on . in one example , the high gauss magnetic field has a field strength of between about a 0 . 5 tesla and about 10 tesla and is applied in the z direction . the method can know follow one of two mutually exclusive paths . the first is the path ( 1 ) through steps 310 , 315 , 320 , 325 and 330 to step 335 . the second path ( 2 ) is through steps 315 a , 320 a and 325 a to step 335 . the unknown micro - electronic device advantageously follows the same path and is subjected to the same nmr conditions as the known micro - electronic device . in step 310 , a direction ( x , y or z ) is selected . in step 315 , a gradient magnetic field of , for example , 1 tesla is applied in the selected direction and in step 320 a rf pulse is directed to the ic . in step 325 the emitted ( returned ) rf signal is detected and information describing the return rf signal ( which is also a pulse ) is stored . the duration and time of reception of the return rf signal will vary . in step 330 , if another direction of the three possible directions ( x , y , z ) is selected and the method loops back to step 310 . this loop will repeat three times , once for each direction . in path ( 2 ) in step 315 a , gradient magnetic fields are applied in the x , y , and z directions simultaneously . each gradient field may have a same or a different field strength of between about 0 . 5 tesla and about 10 tesla . steps 320 a and 325 a are similar to steps 320 and 325 respectively . in steps 315 and 315 a , optional test conditions ( i . e ., voltage bias , analog signal , digital data pattern or combinations thereof ) may be applied to the micro - electronic device . in a first example , optional test conditions are applied only during step 315 ( or 315 a ). in a second example , optional test conditions are applied the only during steps 315 and 320 ( or 315 a and 320 a ). in a third example , the optional test conditions are applied the only during steps 305 , 310 , 315 and 320 ( or 305 , 315 a and 320 a ). in a fourth example , the optional test conditions are applied during steps 305 , 310 , 315 , 320 and 325 ( or 305 , 315 a , 320 a and 325 a ). in step 335 , the ic is removed from the mri chamber and it is determined whether the micro - electronic device was a known micro - electronic device or an unknown micro - electronic device . if the micro - electronic device was a known micro - electronic device then in step 340 analyses are performed and the analyses data is stored . if the micro - electronic device was an unknown micro - electronic device then in step 345 analyses are performed and the analyses is compared to analyses previously stored from a known micro - electronic device mri \u201c images \u201d under the same mri ( and bias / pattern ) conditions . if the compare is within predefined limits the unknown micro - electronic device is validated . it does not matter whether the known or unknown micro - electronic device is run first , but the comparison cannot be performed until both known or unknown micro - electronic devices have been run and the respective data analyzed . fig8 a and 8b illustrate a first method of data analysis for comparison according to embodiments of the present invention . fig8 a is a plot of the transverse component of the returned rf signal for a known micro - electronic device 350 and an unknown micro - electronic device 355 as relative rf strength versus time . direct comparison or statistical analysis of the difference between curves 350 and 355 is performed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig8 a , the difference between curve 355 relative to curve 350 is shown as a positive time shift . the shift may be negative . fig8 b is a plot of the longitudinal component of the returned rf signal for a known micro - electronic device 360 and an unknown micro - electronic device 365 as relative rf strength versus time . direct comparison or statistical analysis of the difference between curves 360 and 365 is performed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig8 b , the difference of curve 365 relative to curve 360 is shown as a negative time shift . the shift may be positive . fig9 a and 9b illustrate a second method of data analysis for comparison according to embodiments of the present invention . fig9 a is a plot of a fast fourier transform of the returned rf signal for a known ic ( or integrated circuit ) as rf amplitude versus rf frequency . element 370 is the returned signal and elements 370 a and 370 b are harmonic artifacts of element 370 . fig9 b is a plot of a fast fourier transform of the returned rf signal for an unknown ic ( or integrated circuit ) as rf amplitude versus rf frequency . element 375 is the returned signal and elements 375 a and 3750 b are harmonic artifacts of element 370 . element 370 and 375 are statistically compared in terms of amplitude of a given frequency range to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig9 a and 9b , the amplitude of the returned rf signal has been transferred from a time domain to a frequency domain . fig1 a , 10 b and 10 c illustrate a third method of data analysis for comparison according to embodiments of the present invention . fig1 a is a plot of the returned rf signal as relative rf signal strength versus time for a known micro - electronic device . rf signal 380 has a fixed amplitude between times a and b measured from when the rf pulse signal terminated ( or other convenient time reference ). an unknown micro - electronic device may exhibit a signal that is offset ion amplitude , phase , frequency or combinations thereof . two simple examples are given in fig1 b and 10c . fig1 b is a plot of the returned rf signal as relative rf signal strength versus time for an unknown micro - electronic device having only a frequency shift . rf signal 385 a has fixed amplitude between times a \u2032 and b \u2032 measured from the same reference as a and b of fig1 a . the shift between a and a \u2032 and b and b \u2032 is statistically to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . fig1 c is a plot of the returned rf signal as relative rf signal strength versus time for an unknown micro - electronic device having only an amplitude shift . rf signal 385 b has a fixed amplitude between times a and b measured from the same reference as a and b of fig1 a . the change in amplitude between curve 380 of fig1 a and curve 385 b of fig1 c is statistically to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . it will be appreciated that a complete statistical analysis would compare time , amplitude , frequency and phase differences of the curves described in fig1 a , 10 b and 10 c . fig1 a , 11 b and 11 c illustrate a fourth method of data analysis for comparison according to embodiments of the present invention . fig1 a is a plot of the returned rf signal xy space ( at a selected z ) as a function of frequency for a known micro - electronic device . as such fig1 a is closer to what is may be considered an \u201c image .\u201d fig1 b is a plot of the returned rf signal xy space ( at the selected z ) as a function of frequency for an unknown micro - electronic device . fig1 c is a plot of the delta between the plot of fig1 a and that of fig1 b . the amount of structure ( and optionally the position of the structures ) is statistically analyzed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . fig1 c is fig1 a \u201c subtracted \u201d from fig1 a . thus the embodiments of the present invention allow for relatively quick and inexpensive validation of integrated circuit chips . the description of the embodiments of the present invention is given above for the understanding of the present invention . it will be understood that the invention is not limited to the particular embodiments described herein , but is capable of various modifications , rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention . therefore , it is intended that the following claims cover all such modifications and changes as fall within the true spirit and scope of the invention ."}
|
{"patent": "the term \u201c integrated circuit \u201d ( ic ) is defined as an integrated circuit chip and a package or module containing the ic . the term \u201c integrated circuit chip \u201d is defined as the semiconductor ( e . g ., silicon ) die containing devices such as transistors , diodes , capacitors , resisters and inductors and the wiring layers built up on the die that interconnect the devices into circuits . the term micro - electronic device includes an ic or an integrated circuit chip . in magnetic resonance imaging ( also called nuclear magnetic resonance ( nmr ) imaging )) a sample is positioned in a static magnetic field and subjected to a pulsed radio frequency ( rf ) signal to place the sample in an excited state . the magnetic field may be turned off or adjusted and a rf return signal is produced by the sample returning to a normal from the excited state is then recorded . in order to allow spatial encoding , the static magnetic field is superimposed with a gradient magnetic field . the term validation means the process by which an unknown micro - electronic device is \u201c imaged \u201d by nmr and the resulting data is compared to data from a known good or trusted micro - electronic that was nmr \u201c imaged \u201d and that the unknown ic micro - electronic device should be essentially identical ( i . e ., of identical design and within fabrication specification limits ) to . known good micro - electronic devices include those fabricated under secure conditions , those from trusted sources and those thoroughly tested and physically inspected after an nmr signature has been obtained . fig1 is a diagram of an exemplary magnetic resonance inspection system according of the present invention . in fig1 , a magnetic resonance imaging ( mri ) system 100 includes a magnet unit 105 , a signal processing unit 110 and a computer 115 . magnet unit 105 of mri system 100 includes a gradient magnetic field coil 120 , a transmit coil 125 , a sample chamber 130 , a receive coil and a high gauss magnet 140 ( e . g ., a permanent magnet , coil magnet or super - conductive magnet ). signal processing unit 110 includes a driving circuit 145 , an rf power amplifier 150 , a preamplifier 155 , a sequence memory 160 , a modulation circuit 165 , an rf oscillator 170 , a phase detector 175 and an analog - to - digital ( a / d ) converter 180 . computer 115 includes a display 185 , an input ( e . g ., a keyboard , mouse , disk drive ) and an output ( e . g ., a display unit , a printer , a disk drive ). gradient magnetic field coil 120 , transmit coil 125 , receive coil 135 and high gauss magnet 140 are disposed so as to substantially surround sample chamber 130 . high gauss magnet 140 applies a static magnetic field having a constant strength to a sample in chamber 130 . gradient magnetic field coil 120 applies gradient magnetic fields selectively to mutually orthogonal x , y and z directions ( in imaging parlance , to a slice axis , phase axis and frequency axis ). transmit coil 125 supplies a pulsed rf signal for exciting spins of atomic nuclei within a micro - electronic device 200 in sample chamber 130 . receive coil 135 detects returned rf signals from the sample in chamber 130 generated when spins of atomic nuclei within micro - electronic device 200 return to a normal state from an exited state . gradient magnetic field coil 120 , transmit coil 125 , and receive coil 135 are operatively associated with driving circuit 145 , an rf power amplifier 150 , and a preamplifier 155 , respectively . sequence memory 160 operates driving circuit 145 based on a stored pulse sequence in response to instructions from computer 115 to thereby apply gradient fields from gradient magnetic field coil 120 in specific directions . sequence memory 160 also operates a modulation circuit 165 to modulate a carrier output signal from rf oscillator 170 into a pulsed rf signal of predefined timing and envelope shape . the pulsed rf signal is applied to rf power amplifier 150 and then the amplified pulsed rf signal is applied to transmit coil 125 . preamplifier 155 amplifies the return rf signal from micro - electronic device 200 in sample chamber 130 detected at receive coil 135 . preamplifier 155 amplifies the received rf signal and sends an amplified rf signal to phase detector 175 . phase detector 175 generates an analog phase - detect signal from the amplified rf signal using a carrier output signal from rf oscillator 170 as a reference signal , and supplies the phase - detected signal to a / d converter 180 . a / d converter 180 converts the phase - detected analog signal into a digital signal , which is supplied to the computer 115 . computer 115 reads and / or processes the data from a / d converter 1801 , and includes algorithms in the form of computer instructions which when executed perform various signal analyses and statistical analyses on the stored data as described infra . results of these analyses may be displayed on output unit 195 . computer 115 can also be responsible for overall control such as receiving information supplied from input 195 from an operator . mri system 100 includes an optional tester 205 , which is connected between a socket , or probe card ( not shown ) in sample chamber 130 and computer 185 . this allows voltage bias , analog signals , digital data patterns or combinations thereof to be applied to micro - electronic device 200 during the mri process . biasing , applying signals and test patterns serves two purposes . first it results in more complex return rf signals . second , if the biasing analog signals , digital data patterns are kept secure , it is very difficult for a an unauthorized party to place a masking circuit into an unauthorized micro electronic device , the purpose of the masking circuit being is to mask the presence of the unauthorized circuit and the masking circuit in the unauthorized micro - electronic device by altering the nmr image of the unauthorized micro - electronic device to mimic that of an authorized micro - electronic device . mri system 100 shown in fig1 is provided as an example , and it will be understood that embodiments of the invention are not limited to mri system 100 shown in fig1 . it will be understood that an mri system 100 according to aspects of the invention can include additional components to those shown in fig1 or may not include every component shown in fig1 . fig2 is a cross - section through an exemplary first type of integrated circuit . in fig2 , an ic 200 a includes an integrated circuit chip 210 physically and electrically connected to a module 215 by solder bumps 220 . wires 225 in module 225 connect solder bumps 220 to solder balls 230 . module 215 may be organic ( e . g ., fiberglass ) or ceramic and wires 25 may comprise one or more wiring layers . solder balls 230 are designed for surface mounting ic 200 a directly to a printed circuit board ( pcb ). solder balls 230 may be replaced by solder columns . solder balls 230 may be replaced with pins , which can be used to mount ics into sockets , which are mounted on a pcb or mounted directly to a pcb . ic 200 a is thus an example of an ic that uses flip - chip ( or c4 ) technology . fig3 is a cross - section through an exemplary second type of integrated circuit . in fig3 , an ic 200 b includes and integrated circuit chip contained within a plastic package 240 . wire bonds 245 electrically connect integrated circuit chip 235 to leads 250 . leads 250 are designed for surface mounting ic 200 b directly to a pcb . leads 250 may be replaced with pins , which can be used to mount ics into sockets , which are mounted on a pcb or mounted directly to a pcb . the leads on some plastic packages are designed to be mounted in sockets on a pcb . ic 200 b is thus an example of plastic packaging technology . fig4 is a top view of an exemplary integrated circuit chip for enhanced comparison according embodiments of the present invention . in fig4 , an exemplary integrated circuit chip 255 includes regions 260 a , 260 b , 260 c , 260 d , 260 e , 260 f , 260 g , 260 h , 2601 , 269 j and 260 k . there may be more or less regions than illustrated in fig4 . these regions often correspond to cores , which are pre - designed circuit functions . for example , a microprocessor may contain multiple processing cores , memory cores , arithmetic cores etc . formed , by way of example , in cores 260 b , 260 g , 260 i and 260 j are serpentine signal enhancing structures 265 which are not electrically connected to any wire or device ( e . g ., transistor , diode , resistor , capacitor or inductor ) of integrated circuit chip 255 . signal enhancing structures 265 may comprise an electrical conductor , a magnetic material , or an electrically conductive magnetic material . signal enhancing structures 265 are designed to interact with the magnetic fields from gradient magnetic field coil 120 and high gauss magnet 140 of mri system 100 ( see fig1 ) to generate a more complex return rf signal . fig5 is a top view of an exemplary integrated circuit chip for preventing or detecting comparison according embodiments of the present invention . in fig5 , an integrated circuit chip 255 a ( similar to integrated circuit chip 255 of fig4 ) includes a serpentine structure 265 a connected to a destruct circuit 270 . serpentine structure 265 a acts as an inductor which generates a current when subjected to a varying magnetic field . the current may be used by destruct circuit 270 to program fuses or activate transistors to render integrated circuit 155 a inoperable or to leave a signature that can later be read to indicate if an attempt at nmr imaging \u201d has been performed on integrated circuit chip 255 a . in one example , serpentine structure 265 a is designed to not be detectable by x - ray imaging . fig6 a is a cross - sectional view of an integrated circuit containing devices to flag an unauthorized comparison attempt according embodiments of the present invention . in fig6 a , an ic 200 c includes an integrated circuit chip 235 a in a plastic package body 240 a . a set ( two sets are shown in the example of fig6 a ) of three \u201c horseshoe \u201d magnets 275 a , 275 b and 275 c are placed in body 240 a . they are aligned so respective lines ( dashed lines ) passing through the poles of each magnet are mutually orthogonal . in fig6 a , the line passing through the poles of magnets 275 a and 275 c are in planes parallel to the top surface 272 of integrated circuit chip 235 a . other pole orientations are possible . in one example , only a single horseshoe magnet is used . horseshoe shaped magnets may be replaced by other shaped magnets such as bar magnets and disc magnets . fig6 b is a cross - sectional view of the integrated circuit of fig6 a after an unauthorized comparison attempt according embodiments of the present invention . when magnets 275 a , 275 b and 275 c are subjected to the intense magnetic field generated in an nmr machine , magnets 275 a , 275 b and 275 c are pulled / pushed by that field so strongly that cracks 280 are formed in body 240 a . fig7 is a flowchart of methods of comparison according to embodiments of the present invention . the steps of the flowchart of fig7 are performed first on a known or trusted micro - electronic device and then on an unknown ( or suspect ) micro - electronic device that and essentially identical to the known micro - electronic device ( i . e ., identically designed and within fabrication specification limits ). for the integrated circuit chip specification limits define allowable variations in material and structure and include , for example , the allowable differences in metal line widths and thickness differences in metal and insulating layers . for the package of the ic specification limits define allowable variations in material and structure and include , for example , package dimensions , size of solder bumps , positions and size of wire bonds , widths and thickness of land in modules . in step 300 an ic ( or integrated circuit is placed in the mri chamber . in step 305 the high gauss magnetic field is turned on . in one example , the high gauss magnetic field has a field strength of between about a 0 . 5 tesla and about 10 tesla and is applied in the z direction . the method can know follow one of two mutually exclusive paths . the first is the path ( 1 ) through steps 310 , 315 , 320 , 325 and 330 to step 335 . the second path ( 2 ) is through steps 315 a , 320 a and 325 a to step 335 . the unknown micro - electronic device advantageously follows the same path and is subjected to the same nmr conditions as the known micro - electronic device . in step 310 , a direction ( x , y or z ) is selected . in step 315 , a gradient magnetic field of , for example , 1 tesla is applied in the selected direction and in step 320 a rf pulse is directed to the ic . in step 325 the emitted ( returned ) rf signal is detected and information describing the return rf signal ( which is also a pulse ) is stored . the duration and time of reception of the return rf signal will vary . in step 330 , if another direction of the three possible directions ( x , y , z ) is selected and the method loops back to step 310 . this loop will repeat three times , once for each direction . in path ( 2 ) in step 315 a , gradient magnetic fields are applied in the x , y , and z directions simultaneously . each gradient field may have a same or a different field strength of between about 0 . 5 tesla and about 10 tesla . steps 320 a and 325 a are similar to steps 320 and 325 respectively . in steps 315 and 315 a , optional test conditions ( i . e ., voltage bias , analog signal , digital data pattern or combinations thereof ) may be applied to the micro - electronic device . in a first example , optional test conditions are applied only during step 315 ( or 315 a ). in a second example , optional test conditions are applied the only during steps 315 and 320 ( or 315 a and 320 a ). in a third example , the optional test conditions are applied the only during steps 305 , 310 , 315 and 320 ( or 305 , 315 a and 320 a ). in a fourth example , the optional test conditions are applied during steps 305 , 310 , 315 , 320 and 325 ( or 305 , 315 a , 320 a and 325 a ). in step 335 , the ic is removed from the mri chamber and it is determined whether the micro - electronic device was a known micro - electronic device or an unknown micro - electronic device . if the micro - electronic device was a known micro - electronic device then in step 340 analyses are performed and the analyses data is stored . if the micro - electronic device was an unknown micro - electronic device then in step 345 analyses are performed and the analyses is compared to analyses previously stored from a known micro - electronic device mri \u201c images \u201d under the same mri ( and bias / pattern ) conditions . if the compare is within predefined limits the unknown micro - electronic device is validated . it does not matter whether the known or unknown micro - electronic device is run first , but the comparison cannot be performed until both known or unknown micro - electronic devices have been run and the respective data analyzed . fig8 a and 8b illustrate a first method of data analysis for comparison according to embodiments of the present invention . fig8 a is a plot of the transverse component of the returned rf signal for a known micro - electronic device 350 and an unknown micro - electronic device 355 as relative rf strength versus time . direct comparison or statistical analysis of the difference between curves 350 and 355 is performed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig8 a , the difference between curve 355 relative to curve 350 is shown as a positive time shift . the shift may be negative . fig8 b is a plot of the longitudinal component of the returned rf signal for a known micro - electronic device 360 and an unknown micro - electronic device 365 as relative rf strength versus time . direct comparison or statistical analysis of the difference between curves 360 and 365 is performed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig8 b , the difference of curve 365 relative to curve 360 is shown as a negative time shift . the shift may be positive . fig9 a and 9b illustrate a second method of data analysis for comparison according to embodiments of the present invention . fig9 a is a plot of a fast fourier transform of the returned rf signal for a known ic ( or integrated circuit ) as rf amplitude versus rf frequency . element 370 is the returned signal and elements 370 a and 370 b are harmonic artifacts of element 370 . fig9 b is a plot of a fast fourier transform of the returned rf signal for an unknown ic ( or integrated circuit ) as rf amplitude versus rf frequency . element 375 is the returned signal and elements 375 a and 3750 b are harmonic artifacts of element 370 . element 370 and 375 are statistically compared in terms of amplitude of a given frequency range to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig9 a and 9b , the amplitude of the returned rf signal has been transferred from a time domain to a frequency domain . fig1 a , 10 b and 10 c illustrate a third method of data analysis for comparison according to embodiments of the present invention . fig1 a is a plot of the returned rf signal as relative rf signal strength versus time for a known micro - electronic device . rf signal 380 has a fixed amplitude between times a and b measured from when the rf pulse signal terminated ( or other convenient time reference ). an unknown micro - electronic device may exhibit a signal that is offset ion amplitude , phase , frequency or combinations thereof . two simple examples are given in fig1 b and 10c . fig1 b is a plot of the returned rf signal as relative rf signal strength versus time for an unknown micro - electronic device having only a frequency shift . rf signal 385 a has fixed amplitude between times a \u2032 and b \u2032 measured from the same reference as a and b of fig1 a . the shift between a and a \u2032 and b and b \u2032 is statistically to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . fig1 c is a plot of the returned rf signal as relative rf signal strength versus time for an unknown micro - electronic device having only an amplitude shift . rf signal 385 b has a fixed amplitude between times a and b measured from the same reference as a and b of fig1 a . the change in amplitude between curve 380 of fig1 a and curve 385 b of fig1 c is statistically to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . it will be appreciated that a complete statistical analysis would compare time , amplitude , frequency and phase differences of the curves described in fig1 a , 10 b and 10 c . fig1 a , 11 b and 11 c illustrate a fourth method of data analysis for comparison according to embodiments of the present invention . fig1 a is a plot of the returned rf signal xy space ( at a selected z ) as a function of frequency for a known micro - electronic device . as such fig1 a is closer to what is may be considered an \u201c image .\u201d fig1 b is a plot of the returned rf signal xy space ( at the selected z ) as a function of frequency for an unknown micro - electronic device . fig1 c is a plot of the delta between the plot of fig1 a and that of fig1 b . the amount of structure ( and optionally the position of the structures ) is statistically analyzed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . fig1 c is fig1 a \u201c subtracted \u201d from fig1 a . thus the embodiments of the present invention allow for relatively quick and inexpensive validation of integrated circuit chips . the description of the embodiments of the present invention is given above for the understanding of the present invention . it will be understood that the invention is not limited to the particular embodiments described herein , but is capable of various modifications , rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention . therefore , it is intended that the following claims cover all such modifications and changes as fall within the true spirit and scope of the invention .", "category": "Fixed Constructions"}
|
Is the patent correctly categorized?
| 0.25 |
d2a533e20b88cf12a66089384d35063da6ac708b8421ee9a63cae12490153828
| 0.1875 | 0.002975 | 0.255859 | 0.028442 | 0.652344 | 0.150391 |
null |
{"category": "Physics", "patent": "the term \u201c integrated circuit \u201d ( ic ) is defined as an integrated circuit chip and a package or module containing the ic . the term \u201c integrated circuit chip \u201d is defined as the semiconductor ( e . g ., silicon ) die containing devices such as transistors , diodes , capacitors , resisters and inductors and the wiring layers built up on the die that interconnect the devices into circuits . the term micro - electronic device includes an ic or an integrated circuit chip . in magnetic resonance imaging ( also called nuclear magnetic resonance ( nmr ) imaging )) a sample is positioned in a static magnetic field and subjected to a pulsed radio frequency ( rf ) signal to place the sample in an excited state . the magnetic field may be turned off or adjusted and a rf return signal is produced by the sample returning to a normal from the excited state is then recorded . in order to allow spatial encoding , the static magnetic field is superimposed with a gradient magnetic field . the term validation means the process by which an unknown micro - electronic device is \u201c imaged \u201d by nmr and the resulting data is compared to data from a known good or trusted micro - electronic that was nmr \u201c imaged \u201d and that the unknown ic micro - electronic device should be essentially identical ( i . e ., of identical design and within fabrication specification limits ) to . known good micro - electronic devices include those fabricated under secure conditions , those from trusted sources and those thoroughly tested and physically inspected after an nmr signature has been obtained . fig1 is a diagram of an exemplary magnetic resonance inspection system according of the present invention . in fig1 , a magnetic resonance imaging ( mri ) system 100 includes a magnet unit 105 , a signal processing unit 110 and a computer 115 . magnet unit 105 of mri system 100 includes a gradient magnetic field coil 120 , a transmit coil 125 , a sample chamber 130 , a receive coil and a high gauss magnet 140 ( e . g ., a permanent magnet , coil magnet or super - conductive magnet ). signal processing unit 110 includes a driving circuit 145 , an rf power amplifier 150 , a preamplifier 155 , a sequence memory 160 , a modulation circuit 165 , an rf oscillator 170 , a phase detector 175 and an analog - to - digital ( a / d ) converter 180 . computer 115 includes a display 185 , an input ( e . g ., a keyboard , mouse , disk drive ) and an output ( e . g ., a display unit , a printer , a disk drive ). gradient magnetic field coil 120 , transmit coil 125 , receive coil 135 and high gauss magnet 140 are disposed so as to substantially surround sample chamber 130 . high gauss magnet 140 applies a static magnetic field having a constant strength to a sample in chamber 130 . gradient magnetic field coil 120 applies gradient magnetic fields selectively to mutually orthogonal x , y and z directions ( in imaging parlance , to a slice axis , phase axis and frequency axis ). transmit coil 125 supplies a pulsed rf signal for exciting spins of atomic nuclei within a micro - electronic device 200 in sample chamber 130 . receive coil 135 detects returned rf signals from the sample in chamber 130 generated when spins of atomic nuclei within micro - electronic device 200 return to a normal state from an exited state . gradient magnetic field coil 120 , transmit coil 125 , and receive coil 135 are operatively associated with driving circuit 145 , an rf power amplifier 150 , and a preamplifier 155 , respectively . sequence memory 160 operates driving circuit 145 based on a stored pulse sequence in response to instructions from computer 115 to thereby apply gradient fields from gradient magnetic field coil 120 in specific directions . sequence memory 160 also operates a modulation circuit 165 to modulate a carrier output signal from rf oscillator 170 into a pulsed rf signal of predefined timing and envelope shape . the pulsed rf signal is applied to rf power amplifier 150 and then the amplified pulsed rf signal is applied to transmit coil 125 . preamplifier 155 amplifies the return rf signal from micro - electronic device 200 in sample chamber 130 detected at receive coil 135 . preamplifier 155 amplifies the received rf signal and sends an amplified rf signal to phase detector 175 . phase detector 175 generates an analog phase - detect signal from the amplified rf signal using a carrier output signal from rf oscillator 170 as a reference signal , and supplies the phase - detected signal to a / d converter 180 . a / d converter 180 converts the phase - detected analog signal into a digital signal , which is supplied to the computer 115 . computer 115 reads and / or processes the data from a / d converter 1801 , and includes algorithms in the form of computer instructions which when executed perform various signal analyses and statistical analyses on the stored data as described infra . results of these analyses may be displayed on output unit 195 . computer 115 can also be responsible for overall control such as receiving information supplied from input 195 from an operator . mri system 100 includes an optional tester 205 , which is connected between a socket , or probe card ( not shown ) in sample chamber 130 and computer 185 . this allows voltage bias , analog signals , digital data patterns or combinations thereof to be applied to micro - electronic device 200 during the mri process . biasing , applying signals and test patterns serves two purposes . first it results in more complex return rf signals . second , if the biasing analog signals , digital data patterns are kept secure , it is very difficult for a an unauthorized party to place a masking circuit into an unauthorized micro electronic device , the purpose of the masking circuit being is to mask the presence of the unauthorized circuit and the masking circuit in the unauthorized micro - electronic device by altering the nmr image of the unauthorized micro - electronic device to mimic that of an authorized micro - electronic device . mri system 100 shown in fig1 is provided as an example , and it will be understood that embodiments of the invention are not limited to mri system 100 shown in fig1 . it will be understood that an mri system 100 according to aspects of the invention can include additional components to those shown in fig1 or may not include every component shown in fig1 . fig2 is a cross - section through an exemplary first type of integrated circuit . in fig2 , an ic 200 a includes an integrated circuit chip 210 physically and electrically connected to a module 215 by solder bumps 220 . wires 225 in module 225 connect solder bumps 220 to solder balls 230 . module 215 may be organic ( e . g ., fiberglass ) or ceramic and wires 25 may comprise one or more wiring layers . solder balls 230 are designed for surface mounting ic 200 a directly to a printed circuit board ( pcb ). solder balls 230 may be replaced by solder columns . solder balls 230 may be replaced with pins , which can be used to mount ics into sockets , which are mounted on a pcb or mounted directly to a pcb . ic 200 a is thus an example of an ic that uses flip - chip ( or c4 ) technology . fig3 is a cross - section through an exemplary second type of integrated circuit . in fig3 , an ic 200 b includes and integrated circuit chip contained within a plastic package 240 . wire bonds 245 electrically connect integrated circuit chip 235 to leads 250 . leads 250 are designed for surface mounting ic 200 b directly to a pcb . leads 250 may be replaced with pins , which can be used to mount ics into sockets , which are mounted on a pcb or mounted directly to a pcb . the leads on some plastic packages are designed to be mounted in sockets on a pcb . ic 200 b is thus an example of plastic packaging technology . fig4 is a top view of an exemplary integrated circuit chip for enhanced comparison according embodiments of the present invention . in fig4 , an exemplary integrated circuit chip 255 includes regions 260 a , 260 b , 260 c , 260 d , 260 e , 260 f , 260 g , 260 h , 2601 , 269 j and 260 k . there may be more or less regions than illustrated in fig4 . these regions often correspond to cores , which are pre - designed circuit functions . for example , a microprocessor may contain multiple processing cores , memory cores , arithmetic cores etc . formed , by way of example , in cores 260 b , 260 g , 260 i and 260 j are serpentine signal enhancing structures 265 which are not electrically connected to any wire or device ( e . g ., transistor , diode , resistor , capacitor or inductor ) of integrated circuit chip 255 . signal enhancing structures 265 may comprise an electrical conductor , a magnetic material , or an electrically conductive magnetic material . signal enhancing structures 265 are designed to interact with the magnetic fields from gradient magnetic field coil 120 and high gauss magnet 140 of mri system 100 ( see fig1 ) to generate a more complex return rf signal . fig5 is a top view of an exemplary integrated circuit chip for preventing or detecting comparison according embodiments of the present invention . in fig5 , an integrated circuit chip 255 a ( similar to integrated circuit chip 255 of fig4 ) includes a serpentine structure 265 a connected to a destruct circuit 270 . serpentine structure 265 a acts as an inductor which generates a current when subjected to a varying magnetic field . the current may be used by destruct circuit 270 to program fuses or activate transistors to render integrated circuit 155 a inoperable or to leave a signature that can later be read to indicate if an attempt at nmr imaging \u201d has been performed on integrated circuit chip 255 a . in one example , serpentine structure 265 a is designed to not be detectable by x - ray imaging . fig6 a is a cross - sectional view of an integrated circuit containing devices to flag an unauthorized comparison attempt according embodiments of the present invention . in fig6 a , an ic 200 c includes an integrated circuit chip 235 a in a plastic package body 240 a . a set ( two sets are shown in the example of fig6 a ) of three \u201c horseshoe \u201d magnets 275 a , 275 b and 275 c are placed in body 240 a . they are aligned so respective lines ( dashed lines ) passing through the poles of each magnet are mutually orthogonal . in fig6 a , the line passing through the poles of magnets 275 a and 275 c are in planes parallel to the top surface 272 of integrated circuit chip 235 a . other pole orientations are possible . in one example , only a single horseshoe magnet is used . horseshoe shaped magnets may be replaced by other shaped magnets such as bar magnets and disc magnets . fig6 b is a cross - sectional view of the integrated circuit of fig6 a after an unauthorized comparison attempt according embodiments of the present invention . when magnets 275 a , 275 b and 275 c are subjected to the intense magnetic field generated in an nmr machine , magnets 275 a , 275 b and 275 c are pulled / pushed by that field so strongly that cracks 280 are formed in body 240 a . fig7 is a flowchart of methods of comparison according to embodiments of the present invention . the steps of the flowchart of fig7 are performed first on a known or trusted micro - electronic device and then on an unknown ( or suspect ) micro - electronic device that and essentially identical to the known micro - electronic device ( i . e ., identically designed and within fabrication specification limits ). for the integrated circuit chip specification limits define allowable variations in material and structure and include , for example , the allowable differences in metal line widths and thickness differences in metal and insulating layers . for the package of the ic specification limits define allowable variations in material and structure and include , for example , package dimensions , size of solder bumps , positions and size of wire bonds , widths and thickness of land in modules . in step 300 an ic ( or integrated circuit is placed in the mri chamber . in step 305 the high gauss magnetic field is turned on . in one example , the high gauss magnetic field has a field strength of between about a 0 . 5 tesla and about 10 tesla and is applied in the z direction . the method can know follow one of two mutually exclusive paths . the first is the path ( 1 ) through steps 310 , 315 , 320 , 325 and 330 to step 335 . the second path ( 2 ) is through steps 315 a , 320 a and 325 a to step 335 . the unknown micro - electronic device advantageously follows the same path and is subjected to the same nmr conditions as the known micro - electronic device . in step 310 , a direction ( x , y or z ) is selected . in step 315 , a gradient magnetic field of , for example , 1 tesla is applied in the selected direction and in step 320 a rf pulse is directed to the ic . in step 325 the emitted ( returned ) rf signal is detected and information describing the return rf signal ( which is also a pulse ) is stored . the duration and time of reception of the return rf signal will vary . in step 330 , if another direction of the three possible directions ( x , y , z ) is selected and the method loops back to step 310 . this loop will repeat three times , once for each direction . in path ( 2 ) in step 315 a , gradient magnetic fields are applied in the x , y , and z directions simultaneously . each gradient field may have a same or a different field strength of between about 0 . 5 tesla and about 10 tesla . steps 320 a and 325 a are similar to steps 320 and 325 respectively . in steps 315 and 315 a , optional test conditions ( i . e ., voltage bias , analog signal , digital data pattern or combinations thereof ) may be applied to the micro - electronic device . in a first example , optional test conditions are applied only during step 315 ( or 315 a ). in a second example , optional test conditions are applied the only during steps 315 and 320 ( or 315 a and 320 a ). in a third example , the optional test conditions are applied the only during steps 305 , 310 , 315 and 320 ( or 305 , 315 a and 320 a ). in a fourth example , the optional test conditions are applied during steps 305 , 310 , 315 , 320 and 325 ( or 305 , 315 a , 320 a and 325 a ). in step 335 , the ic is removed from the mri chamber and it is determined whether the micro - electronic device was a known micro - electronic device or an unknown micro - electronic device . if the micro - electronic device was a known micro - electronic device then in step 340 analyses are performed and the analyses data is stored . if the micro - electronic device was an unknown micro - electronic device then in step 345 analyses are performed and the analyses is compared to analyses previously stored from a known micro - electronic device mri \u201c images \u201d under the same mri ( and bias / pattern ) conditions . if the compare is within predefined limits the unknown micro - electronic device is validated . it does not matter whether the known or unknown micro - electronic device is run first , but the comparison cannot be performed until both known or unknown micro - electronic devices have been run and the respective data analyzed . fig8 a and 8b illustrate a first method of data analysis for comparison according to embodiments of the present invention . fig8 a is a plot of the transverse component of the returned rf signal for a known micro - electronic device 350 and an unknown micro - electronic device 355 as relative rf strength versus time . direct comparison or statistical analysis of the difference between curves 350 and 355 is performed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig8 a , the difference between curve 355 relative to curve 350 is shown as a positive time shift . the shift may be negative . fig8 b is a plot of the longitudinal component of the returned rf signal for a known micro - electronic device 360 and an unknown micro - electronic device 365 as relative rf strength versus time . direct comparison or statistical analysis of the difference between curves 360 and 365 is performed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig8 b , the difference of curve 365 relative to curve 360 is shown as a negative time shift . the shift may be positive . fig9 a and 9b illustrate a second method of data analysis for comparison according to embodiments of the present invention . fig9 a is a plot of a fast fourier transform of the returned rf signal for a known ic ( or integrated circuit ) as rf amplitude versus rf frequency . element 370 is the returned signal and elements 370 a and 370 b are harmonic artifacts of element 370 . fig9 b is a plot of a fast fourier transform of the returned rf signal for an unknown ic ( or integrated circuit ) as rf amplitude versus rf frequency . element 375 is the returned signal and elements 375 a and 3750 b are harmonic artifacts of element 370 . element 370 and 375 are statistically compared in terms of amplitude of a given frequency range to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig9 a and 9b , the amplitude of the returned rf signal has been transferred from a time domain to a frequency domain . fig1 a , 10 b and 10 c illustrate a third method of data analysis for comparison according to embodiments of the present invention . fig1 a is a plot of the returned rf signal as relative rf signal strength versus time for a known micro - electronic device . rf signal 380 has a fixed amplitude between times a and b measured from when the rf pulse signal terminated ( or other convenient time reference ). an unknown micro - electronic device may exhibit a signal that is offset ion amplitude , phase , frequency or combinations thereof . two simple examples are given in fig1 b and 10c . fig1 b is a plot of the returned rf signal as relative rf signal strength versus time for an unknown micro - electronic device having only a frequency shift . rf signal 385 a has fixed amplitude between times a \u2032 and b \u2032 measured from the same reference as a and b of fig1 a . the shift between a and a \u2032 and b and b \u2032 is statistically to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . fig1 c is a plot of the returned rf signal as relative rf signal strength versus time for an unknown micro - electronic device having only an amplitude shift . rf signal 385 b has a fixed amplitude between times a and b measured from the same reference as a and b of fig1 a . the change in amplitude between curve 380 of fig1 a and curve 385 b of fig1 c is statistically to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . it will be appreciated that a complete statistical analysis would compare time , amplitude , frequency and phase differences of the curves described in fig1 a , 10 b and 10 c . fig1 a , 11 b and 11 c illustrate a fourth method of data analysis for comparison according to embodiments of the present invention . fig1 a is a plot of the returned rf signal xy space ( at a selected z ) as a function of frequency for a known micro - electronic device . as such fig1 a is closer to what is may be considered an \u201c image .\u201d fig1 b is a plot of the returned rf signal xy space ( at the selected z ) as a function of frequency for an unknown micro - electronic device . fig1 c is a plot of the delta between the plot of fig1 a and that of fig1 b . the amount of structure ( and optionally the position of the structures ) is statistically analyzed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . fig1 c is fig1 a \u201c subtracted \u201d from fig1 a . thus the embodiments of the present invention allow for relatively quick and inexpensive validation of integrated circuit chips . the description of the embodiments of the present invention is given above for the understanding of the present invention . it will be understood that the invention is not limited to the particular embodiments described herein , but is capable of various modifications , rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention . therefore , it is intended that the following claims cover all such modifications and changes as fall within the true spirit and scope of the invention ."}
|
{"category": "Mechanical Engineering; Lightning; Heating; Weapons; Blasting", "patent": "the term \u201c integrated circuit \u201d ( ic ) is defined as an integrated circuit chip and a package or module containing the ic . the term \u201c integrated circuit chip \u201d is defined as the semiconductor ( e . g ., silicon ) die containing devices such as transistors , diodes , capacitors , resisters and inductors and the wiring layers built up on the die that interconnect the devices into circuits . the term micro - electronic device includes an ic or an integrated circuit chip . in magnetic resonance imaging ( also called nuclear magnetic resonance ( nmr ) imaging )) a sample is positioned in a static magnetic field and subjected to a pulsed radio frequency ( rf ) signal to place the sample in an excited state . the magnetic field may be turned off or adjusted and a rf return signal is produced by the sample returning to a normal from the excited state is then recorded . in order to allow spatial encoding , the static magnetic field is superimposed with a gradient magnetic field . the term validation means the process by which an unknown micro - electronic device is \u201c imaged \u201d by nmr and the resulting data is compared to data from a known good or trusted micro - electronic that was nmr \u201c imaged \u201d and that the unknown ic micro - electronic device should be essentially identical ( i . e ., of identical design and within fabrication specification limits ) to . known good micro - electronic devices include those fabricated under secure conditions , those from trusted sources and those thoroughly tested and physically inspected after an nmr signature has been obtained . fig1 is a diagram of an exemplary magnetic resonance inspection system according of the present invention . in fig1 , a magnetic resonance imaging ( mri ) system 100 includes a magnet unit 105 , a signal processing unit 110 and a computer 115 . magnet unit 105 of mri system 100 includes a gradient magnetic field coil 120 , a transmit coil 125 , a sample chamber 130 , a receive coil and a high gauss magnet 140 ( e . g ., a permanent magnet , coil magnet or super - conductive magnet ). signal processing unit 110 includes a driving circuit 145 , an rf power amplifier 150 , a preamplifier 155 , a sequence memory 160 , a modulation circuit 165 , an rf oscillator 170 , a phase detector 175 and an analog - to - digital ( a / d ) converter 180 . computer 115 includes a display 185 , an input ( e . g ., a keyboard , mouse , disk drive ) and an output ( e . g ., a display unit , a printer , a disk drive ). gradient magnetic field coil 120 , transmit coil 125 , receive coil 135 and high gauss magnet 140 are disposed so as to substantially surround sample chamber 130 . high gauss magnet 140 applies a static magnetic field having a constant strength to a sample in chamber 130 . gradient magnetic field coil 120 applies gradient magnetic fields selectively to mutually orthogonal x , y and z directions ( in imaging parlance , to a slice axis , phase axis and frequency axis ). transmit coil 125 supplies a pulsed rf signal for exciting spins of atomic nuclei within a micro - electronic device 200 in sample chamber 130 . receive coil 135 detects returned rf signals from the sample in chamber 130 generated when spins of atomic nuclei within micro - electronic device 200 return to a normal state from an exited state . gradient magnetic field coil 120 , transmit coil 125 , and receive coil 135 are operatively associated with driving circuit 145 , an rf power amplifier 150 , and a preamplifier 155 , respectively . sequence memory 160 operates driving circuit 145 based on a stored pulse sequence in response to instructions from computer 115 to thereby apply gradient fields from gradient magnetic field coil 120 in specific directions . sequence memory 160 also operates a modulation circuit 165 to modulate a carrier output signal from rf oscillator 170 into a pulsed rf signal of predefined timing and envelope shape . the pulsed rf signal is applied to rf power amplifier 150 and then the amplified pulsed rf signal is applied to transmit coil 125 . preamplifier 155 amplifies the return rf signal from micro - electronic device 200 in sample chamber 130 detected at receive coil 135 . preamplifier 155 amplifies the received rf signal and sends an amplified rf signal to phase detector 175 . phase detector 175 generates an analog phase - detect signal from the amplified rf signal using a carrier output signal from rf oscillator 170 as a reference signal , and supplies the phase - detected signal to a / d converter 180 . a / d converter 180 converts the phase - detected analog signal into a digital signal , which is supplied to the computer 115 . computer 115 reads and / or processes the data from a / d converter 1801 , and includes algorithms in the form of computer instructions which when executed perform various signal analyses and statistical analyses on the stored data as described infra . results of these analyses may be displayed on output unit 195 . computer 115 can also be responsible for overall control such as receiving information supplied from input 195 from an operator . mri system 100 includes an optional tester 205 , which is connected between a socket , or probe card ( not shown ) in sample chamber 130 and computer 185 . this allows voltage bias , analog signals , digital data patterns or combinations thereof to be applied to micro - electronic device 200 during the mri process . biasing , applying signals and test patterns serves two purposes . first it results in more complex return rf signals . second , if the biasing analog signals , digital data patterns are kept secure , it is very difficult for a an unauthorized party to place a masking circuit into an unauthorized micro electronic device , the purpose of the masking circuit being is to mask the presence of the unauthorized circuit and the masking circuit in the unauthorized micro - electronic device by altering the nmr image of the unauthorized micro - electronic device to mimic that of an authorized micro - electronic device . mri system 100 shown in fig1 is provided as an example , and it will be understood that embodiments of the invention are not limited to mri system 100 shown in fig1 . it will be understood that an mri system 100 according to aspects of the invention can include additional components to those shown in fig1 or may not include every component shown in fig1 . fig2 is a cross - section through an exemplary first type of integrated circuit . in fig2 , an ic 200 a includes an integrated circuit chip 210 physically and electrically connected to a module 215 by solder bumps 220 . wires 225 in module 225 connect solder bumps 220 to solder balls 230 . module 215 may be organic ( e . g ., fiberglass ) or ceramic and wires 25 may comprise one or more wiring layers . solder balls 230 are designed for surface mounting ic 200 a directly to a printed circuit board ( pcb ). solder balls 230 may be replaced by solder columns . solder balls 230 may be replaced with pins , which can be used to mount ics into sockets , which are mounted on a pcb or mounted directly to a pcb . ic 200 a is thus an example of an ic that uses flip - chip ( or c4 ) technology . fig3 is a cross - section through an exemplary second type of integrated circuit . in fig3 , an ic 200 b includes and integrated circuit chip contained within a plastic package 240 . wire bonds 245 electrically connect integrated circuit chip 235 to leads 250 . leads 250 are designed for surface mounting ic 200 b directly to a pcb . leads 250 may be replaced with pins , which can be used to mount ics into sockets , which are mounted on a pcb or mounted directly to a pcb . the leads on some plastic packages are designed to be mounted in sockets on a pcb . ic 200 b is thus an example of plastic packaging technology . fig4 is a top view of an exemplary integrated circuit chip for enhanced comparison according embodiments of the present invention . in fig4 , an exemplary integrated circuit chip 255 includes regions 260 a , 260 b , 260 c , 260 d , 260 e , 260 f , 260 g , 260 h , 2601 , 269 j and 260 k . there may be more or less regions than illustrated in fig4 . these regions often correspond to cores , which are pre - designed circuit functions . for example , a microprocessor may contain multiple processing cores , memory cores , arithmetic cores etc . formed , by way of example , in cores 260 b , 260 g , 260 i and 260 j are serpentine signal enhancing structures 265 which are not electrically connected to any wire or device ( e . g ., transistor , diode , resistor , capacitor or inductor ) of integrated circuit chip 255 . signal enhancing structures 265 may comprise an electrical conductor , a magnetic material , or an electrically conductive magnetic material . signal enhancing structures 265 are designed to interact with the magnetic fields from gradient magnetic field coil 120 and high gauss magnet 140 of mri system 100 ( see fig1 ) to generate a more complex return rf signal . fig5 is a top view of an exemplary integrated circuit chip for preventing or detecting comparison according embodiments of the present invention . in fig5 , an integrated circuit chip 255 a ( similar to integrated circuit chip 255 of fig4 ) includes a serpentine structure 265 a connected to a destruct circuit 270 . serpentine structure 265 a acts as an inductor which generates a current when subjected to a varying magnetic field . the current may be used by destruct circuit 270 to program fuses or activate transistors to render integrated circuit 155 a inoperable or to leave a signature that can later be read to indicate if an attempt at nmr imaging \u201d has been performed on integrated circuit chip 255 a . in one example , serpentine structure 265 a is designed to not be detectable by x - ray imaging . fig6 a is a cross - sectional view of an integrated circuit containing devices to flag an unauthorized comparison attempt according embodiments of the present invention . in fig6 a , an ic 200 c includes an integrated circuit chip 235 a in a plastic package body 240 a . a set ( two sets are shown in the example of fig6 a ) of three \u201c horseshoe \u201d magnets 275 a , 275 b and 275 c are placed in body 240 a . they are aligned so respective lines ( dashed lines ) passing through the poles of each magnet are mutually orthogonal . in fig6 a , the line passing through the poles of magnets 275 a and 275 c are in planes parallel to the top surface 272 of integrated circuit chip 235 a . other pole orientations are possible . in one example , only a single horseshoe magnet is used . horseshoe shaped magnets may be replaced by other shaped magnets such as bar magnets and disc magnets . fig6 b is a cross - sectional view of the integrated circuit of fig6 a after an unauthorized comparison attempt according embodiments of the present invention . when magnets 275 a , 275 b and 275 c are subjected to the intense magnetic field generated in an nmr machine , magnets 275 a , 275 b and 275 c are pulled / pushed by that field so strongly that cracks 280 are formed in body 240 a . fig7 is a flowchart of methods of comparison according to embodiments of the present invention . the steps of the flowchart of fig7 are performed first on a known or trusted micro - electronic device and then on an unknown ( or suspect ) micro - electronic device that and essentially identical to the known micro - electronic device ( i . e ., identically designed and within fabrication specification limits ). for the integrated circuit chip specification limits define allowable variations in material and structure and include , for example , the allowable differences in metal line widths and thickness differences in metal and insulating layers . for the package of the ic specification limits define allowable variations in material and structure and include , for example , package dimensions , size of solder bumps , positions and size of wire bonds , widths and thickness of land in modules . in step 300 an ic ( or integrated circuit is placed in the mri chamber . in step 305 the high gauss magnetic field is turned on . in one example , the high gauss magnetic field has a field strength of between about a 0 . 5 tesla and about 10 tesla and is applied in the z direction . the method can know follow one of two mutually exclusive paths . the first is the path ( 1 ) through steps 310 , 315 , 320 , 325 and 330 to step 335 . the second path ( 2 ) is through steps 315 a , 320 a and 325 a to step 335 . the unknown micro - electronic device advantageously follows the same path and is subjected to the same nmr conditions as the known micro - electronic device . in step 310 , a direction ( x , y or z ) is selected . in step 315 , a gradient magnetic field of , for example , 1 tesla is applied in the selected direction and in step 320 a rf pulse is directed to the ic . in step 325 the emitted ( returned ) rf signal is detected and information describing the return rf signal ( which is also a pulse ) is stored . the duration and time of reception of the return rf signal will vary . in step 330 , if another direction of the three possible directions ( x , y , z ) is selected and the method loops back to step 310 . this loop will repeat three times , once for each direction . in path ( 2 ) in step 315 a , gradient magnetic fields are applied in the x , y , and z directions simultaneously . each gradient field may have a same or a different field strength of between about 0 . 5 tesla and about 10 tesla . steps 320 a and 325 a are similar to steps 320 and 325 respectively . in steps 315 and 315 a , optional test conditions ( i . e ., voltage bias , analog signal , digital data pattern or combinations thereof ) may be applied to the micro - electronic device . in a first example , optional test conditions are applied only during step 315 ( or 315 a ). in a second example , optional test conditions are applied the only during steps 315 and 320 ( or 315 a and 320 a ). in a third example , the optional test conditions are applied the only during steps 305 , 310 , 315 and 320 ( or 305 , 315 a and 320 a ). in a fourth example , the optional test conditions are applied during steps 305 , 310 , 315 , 320 and 325 ( or 305 , 315 a , 320 a and 325 a ). in step 335 , the ic is removed from the mri chamber and it is determined whether the micro - electronic device was a known micro - electronic device or an unknown micro - electronic device . if the micro - electronic device was a known micro - electronic device then in step 340 analyses are performed and the analyses data is stored . if the micro - electronic device was an unknown micro - electronic device then in step 345 analyses are performed and the analyses is compared to analyses previously stored from a known micro - electronic device mri \u201c images \u201d under the same mri ( and bias / pattern ) conditions . if the compare is within predefined limits the unknown micro - electronic device is validated . it does not matter whether the known or unknown micro - electronic device is run first , but the comparison cannot be performed until both known or unknown micro - electronic devices have been run and the respective data analyzed . fig8 a and 8b illustrate a first method of data analysis for comparison according to embodiments of the present invention . fig8 a is a plot of the transverse component of the returned rf signal for a known micro - electronic device 350 and an unknown micro - electronic device 355 as relative rf strength versus time . direct comparison or statistical analysis of the difference between curves 350 and 355 is performed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig8 a , the difference between curve 355 relative to curve 350 is shown as a positive time shift . the shift may be negative . fig8 b is a plot of the longitudinal component of the returned rf signal for a known micro - electronic device 360 and an unknown micro - electronic device 365 as relative rf strength versus time . direct comparison or statistical analysis of the difference between curves 360 and 365 is performed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig8 b , the difference of curve 365 relative to curve 360 is shown as a negative time shift . the shift may be positive . fig9 a and 9b illustrate a second method of data analysis for comparison according to embodiments of the present invention . fig9 a is a plot of a fast fourier transform of the returned rf signal for a known ic ( or integrated circuit ) as rf amplitude versus rf frequency . element 370 is the returned signal and elements 370 a and 370 b are harmonic artifacts of element 370 . fig9 b is a plot of a fast fourier transform of the returned rf signal for an unknown ic ( or integrated circuit ) as rf amplitude versus rf frequency . element 375 is the returned signal and elements 375 a and 3750 b are harmonic artifacts of element 370 . element 370 and 375 are statistically compared in terms of amplitude of a given frequency range to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig9 a and 9b , the amplitude of the returned rf signal has been transferred from a time domain to a frequency domain . fig1 a , 10 b and 10 c illustrate a third method of data analysis for comparison according to embodiments of the present invention . fig1 a is a plot of the returned rf signal as relative rf signal strength versus time for a known micro - electronic device . rf signal 380 has a fixed amplitude between times a and b measured from when the rf pulse signal terminated ( or other convenient time reference ). an unknown micro - electronic device may exhibit a signal that is offset ion amplitude , phase , frequency or combinations thereof . two simple examples are given in fig1 b and 10c . fig1 b is a plot of the returned rf signal as relative rf signal strength versus time for an unknown micro - electronic device having only a frequency shift . rf signal 385 a has fixed amplitude between times a \u2032 and b \u2032 measured from the same reference as a and b of fig1 a . the shift between a and a \u2032 and b and b \u2032 is statistically to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . fig1 c is a plot of the returned rf signal as relative rf signal strength versus time for an unknown micro - electronic device having only an amplitude shift . rf signal 385 b has a fixed amplitude between times a and b measured from the same reference as a and b of fig1 a . the change in amplitude between curve 380 of fig1 a and curve 385 b of fig1 c is statistically to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . it will be appreciated that a complete statistical analysis would compare time , amplitude , frequency and phase differences of the curves described in fig1 a , 10 b and 10 c . fig1 a , 11 b and 11 c illustrate a fourth method of data analysis for comparison according to embodiments of the present invention . fig1 a is a plot of the returned rf signal xy space ( at a selected z ) as a function of frequency for a known micro - electronic device . as such fig1 a is closer to what is may be considered an \u201c image .\u201d fig1 b is a plot of the returned rf signal xy space ( at the selected z ) as a function of frequency for an unknown micro - electronic device . fig1 c is a plot of the delta between the plot of fig1 a and that of fig1 b . the amount of structure ( and optionally the position of the structures ) is statistically analyzed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . fig1 c is fig1 a \u201c subtracted \u201d from fig1 a . thus the embodiments of the present invention allow for relatively quick and inexpensive validation of integrated circuit chips . the description of the embodiments of the present invention is given above for the understanding of the present invention . it will be understood that the invention is not limited to the particular embodiments described herein , but is capable of various modifications , rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention . therefore , it is intended that the following claims cover all such modifications and changes as fall within the true spirit and scope of the invention ."}
|
Does the category match the content of the patent?
| 0.25 |
d2a533e20b88cf12a66089384d35063da6ac708b8421ee9a63cae12490153828
| 0.166016 | 0.001328 | 0.226563 | 0.000732 | 0.5625 | 0.012024 |
null |
{"category": "Physics", "patent": "the term \u201c integrated circuit \u201d ( ic ) is defined as an integrated circuit chip and a package or module containing the ic . the term \u201c integrated circuit chip \u201d is defined as the semiconductor ( e . g ., silicon ) die containing devices such as transistors , diodes , capacitors , resisters and inductors and the wiring layers built up on the die that interconnect the devices into circuits . the term micro - electronic device includes an ic or an integrated circuit chip . in magnetic resonance imaging ( also called nuclear magnetic resonance ( nmr ) imaging )) a sample is positioned in a static magnetic field and subjected to a pulsed radio frequency ( rf ) signal to place the sample in an excited state . the magnetic field may be turned off or adjusted and a rf return signal is produced by the sample returning to a normal from the excited state is then recorded . in order to allow spatial encoding , the static magnetic field is superimposed with a gradient magnetic field . the term validation means the process by which an unknown micro - electronic device is \u201c imaged \u201d by nmr and the resulting data is compared to data from a known good or trusted micro - electronic that was nmr \u201c imaged \u201d and that the unknown ic micro - electronic device should be essentially identical ( i . e ., of identical design and within fabrication specification limits ) to . known good micro - electronic devices include those fabricated under secure conditions , those from trusted sources and those thoroughly tested and physically inspected after an nmr signature has been obtained . fig1 is a diagram of an exemplary magnetic resonance inspection system according of the present invention . in fig1 , a magnetic resonance imaging ( mri ) system 100 includes a magnet unit 105 , a signal processing unit 110 and a computer 115 . magnet unit 105 of mri system 100 includes a gradient magnetic field coil 120 , a transmit coil 125 , a sample chamber 130 , a receive coil and a high gauss magnet 140 ( e . g ., a permanent magnet , coil magnet or super - conductive magnet ). signal processing unit 110 includes a driving circuit 145 , an rf power amplifier 150 , a preamplifier 155 , a sequence memory 160 , a modulation circuit 165 , an rf oscillator 170 , a phase detector 175 and an analog - to - digital ( a / d ) converter 180 . computer 115 includes a display 185 , an input ( e . g ., a keyboard , mouse , disk drive ) and an output ( e . g ., a display unit , a printer , a disk drive ). gradient magnetic field coil 120 , transmit coil 125 , receive coil 135 and high gauss magnet 140 are disposed so as to substantially surround sample chamber 130 . high gauss magnet 140 applies a static magnetic field having a constant strength to a sample in chamber 130 . gradient magnetic field coil 120 applies gradient magnetic fields selectively to mutually orthogonal x , y and z directions ( in imaging parlance , to a slice axis , phase axis and frequency axis ). transmit coil 125 supplies a pulsed rf signal for exciting spins of atomic nuclei within a micro - electronic device 200 in sample chamber 130 . receive coil 135 detects returned rf signals from the sample in chamber 130 generated when spins of atomic nuclei within micro - electronic device 200 return to a normal state from an exited state . gradient magnetic field coil 120 , transmit coil 125 , and receive coil 135 are operatively associated with driving circuit 145 , an rf power amplifier 150 , and a preamplifier 155 , respectively . sequence memory 160 operates driving circuit 145 based on a stored pulse sequence in response to instructions from computer 115 to thereby apply gradient fields from gradient magnetic field coil 120 in specific directions . sequence memory 160 also operates a modulation circuit 165 to modulate a carrier output signal from rf oscillator 170 into a pulsed rf signal of predefined timing and envelope shape . the pulsed rf signal is applied to rf power amplifier 150 and then the amplified pulsed rf signal is applied to transmit coil 125 . preamplifier 155 amplifies the return rf signal from micro - electronic device 200 in sample chamber 130 detected at receive coil 135 . preamplifier 155 amplifies the received rf signal and sends an amplified rf signal to phase detector 175 . phase detector 175 generates an analog phase - detect signal from the amplified rf signal using a carrier output signal from rf oscillator 170 as a reference signal , and supplies the phase - detected signal to a / d converter 180 . a / d converter 180 converts the phase - detected analog signal into a digital signal , which is supplied to the computer 115 . computer 115 reads and / or processes the data from a / d converter 1801 , and includes algorithms in the form of computer instructions which when executed perform various signal analyses and statistical analyses on the stored data as described infra . results of these analyses may be displayed on output unit 195 . computer 115 can also be responsible for overall control such as receiving information supplied from input 195 from an operator . mri system 100 includes an optional tester 205 , which is connected between a socket , or probe card ( not shown ) in sample chamber 130 and computer 185 . this allows voltage bias , analog signals , digital data patterns or combinations thereof to be applied to micro - electronic device 200 during the mri process . biasing , applying signals and test patterns serves two purposes . first it results in more complex return rf signals . second , if the biasing analog signals , digital data patterns are kept secure , it is very difficult for a an unauthorized party to place a masking circuit into an unauthorized micro electronic device , the purpose of the masking circuit being is to mask the presence of the unauthorized circuit and the masking circuit in the unauthorized micro - electronic device by altering the nmr image of the unauthorized micro - electronic device to mimic that of an authorized micro - electronic device . mri system 100 shown in fig1 is provided as an example , and it will be understood that embodiments of the invention are not limited to mri system 100 shown in fig1 . it will be understood that an mri system 100 according to aspects of the invention can include additional components to those shown in fig1 or may not include every component shown in fig1 . fig2 is a cross - section through an exemplary first type of integrated circuit . in fig2 , an ic 200 a includes an integrated circuit chip 210 physically and electrically connected to a module 215 by solder bumps 220 . wires 225 in module 225 connect solder bumps 220 to solder balls 230 . module 215 may be organic ( e . g ., fiberglass ) or ceramic and wires 25 may comprise one or more wiring layers . solder balls 230 are designed for surface mounting ic 200 a directly to a printed circuit board ( pcb ). solder balls 230 may be replaced by solder columns . solder balls 230 may be replaced with pins , which can be used to mount ics into sockets , which are mounted on a pcb or mounted directly to a pcb . ic 200 a is thus an example of an ic that uses flip - chip ( or c4 ) technology . fig3 is a cross - section through an exemplary second type of integrated circuit . in fig3 , an ic 200 b includes and integrated circuit chip contained within a plastic package 240 . wire bonds 245 electrically connect integrated circuit chip 235 to leads 250 . leads 250 are designed for surface mounting ic 200 b directly to a pcb . leads 250 may be replaced with pins , which can be used to mount ics into sockets , which are mounted on a pcb or mounted directly to a pcb . the leads on some plastic packages are designed to be mounted in sockets on a pcb . ic 200 b is thus an example of plastic packaging technology . fig4 is a top view of an exemplary integrated circuit chip for enhanced comparison according embodiments of the present invention . in fig4 , an exemplary integrated circuit chip 255 includes regions 260 a , 260 b , 260 c , 260 d , 260 e , 260 f , 260 g , 260 h , 2601 , 269 j and 260 k . there may be more or less regions than illustrated in fig4 . these regions often correspond to cores , which are pre - designed circuit functions . for example , a microprocessor may contain multiple processing cores , memory cores , arithmetic cores etc . formed , by way of example , in cores 260 b , 260 g , 260 i and 260 j are serpentine signal enhancing structures 265 which are not electrically connected to any wire or device ( e . g ., transistor , diode , resistor , capacitor or inductor ) of integrated circuit chip 255 . signal enhancing structures 265 may comprise an electrical conductor , a magnetic material , or an electrically conductive magnetic material . signal enhancing structures 265 are designed to interact with the magnetic fields from gradient magnetic field coil 120 and high gauss magnet 140 of mri system 100 ( see fig1 ) to generate a more complex return rf signal . fig5 is a top view of an exemplary integrated circuit chip for preventing or detecting comparison according embodiments of the present invention . in fig5 , an integrated circuit chip 255 a ( similar to integrated circuit chip 255 of fig4 ) includes a serpentine structure 265 a connected to a destruct circuit 270 . serpentine structure 265 a acts as an inductor which generates a current when subjected to a varying magnetic field . the current may be used by destruct circuit 270 to program fuses or activate transistors to render integrated circuit 155 a inoperable or to leave a signature that can later be read to indicate if an attempt at nmr imaging \u201d has been performed on integrated circuit chip 255 a . in one example , serpentine structure 265 a is designed to not be detectable by x - ray imaging . fig6 a is a cross - sectional view of an integrated circuit containing devices to flag an unauthorized comparison attempt according embodiments of the present invention . in fig6 a , an ic 200 c includes an integrated circuit chip 235 a in a plastic package body 240 a . a set ( two sets are shown in the example of fig6 a ) of three \u201c horseshoe \u201d magnets 275 a , 275 b and 275 c are placed in body 240 a . they are aligned so respective lines ( dashed lines ) passing through the poles of each magnet are mutually orthogonal . in fig6 a , the line passing through the poles of magnets 275 a and 275 c are in planes parallel to the top surface 272 of integrated circuit chip 235 a . other pole orientations are possible . in one example , only a single horseshoe magnet is used . horseshoe shaped magnets may be replaced by other shaped magnets such as bar magnets and disc magnets . fig6 b is a cross - sectional view of the integrated circuit of fig6 a after an unauthorized comparison attempt according embodiments of the present invention . when magnets 275 a , 275 b and 275 c are subjected to the intense magnetic field generated in an nmr machine , magnets 275 a , 275 b and 275 c are pulled / pushed by that field so strongly that cracks 280 are formed in body 240 a . fig7 is a flowchart of methods of comparison according to embodiments of the present invention . the steps of the flowchart of fig7 are performed first on a known or trusted micro - electronic device and then on an unknown ( or suspect ) micro - electronic device that and essentially identical to the known micro - electronic device ( i . e ., identically designed and within fabrication specification limits ). for the integrated circuit chip specification limits define allowable variations in material and structure and include , for example , the allowable differences in metal line widths and thickness differences in metal and insulating layers . for the package of the ic specification limits define allowable variations in material and structure and include , for example , package dimensions , size of solder bumps , positions and size of wire bonds , widths and thickness of land in modules . in step 300 an ic ( or integrated circuit is placed in the mri chamber . in step 305 the high gauss magnetic field is turned on . in one example , the high gauss magnetic field has a field strength of between about a 0 . 5 tesla and about 10 tesla and is applied in the z direction . the method can know follow one of two mutually exclusive paths . the first is the path ( 1 ) through steps 310 , 315 , 320 , 325 and 330 to step 335 . the second path ( 2 ) is through steps 315 a , 320 a and 325 a to step 335 . the unknown micro - electronic device advantageously follows the same path and is subjected to the same nmr conditions as the known micro - electronic device . in step 310 , a direction ( x , y or z ) is selected . in step 315 , a gradient magnetic field of , for example , 1 tesla is applied in the selected direction and in step 320 a rf pulse is directed to the ic . in step 325 the emitted ( returned ) rf signal is detected and information describing the return rf signal ( which is also a pulse ) is stored . the duration and time of reception of the return rf signal will vary . in step 330 , if another direction of the three possible directions ( x , y , z ) is selected and the method loops back to step 310 . this loop will repeat three times , once for each direction . in path ( 2 ) in step 315 a , gradient magnetic fields are applied in the x , y , and z directions simultaneously . each gradient field may have a same or a different field strength of between about 0 . 5 tesla and about 10 tesla . steps 320 a and 325 a are similar to steps 320 and 325 respectively . in steps 315 and 315 a , optional test conditions ( i . e ., voltage bias , analog signal , digital data pattern or combinations thereof ) may be applied to the micro - electronic device . in a first example , optional test conditions are applied only during step 315 ( or 315 a ). in a second example , optional test conditions are applied the only during steps 315 and 320 ( or 315 a and 320 a ). in a third example , the optional test conditions are applied the only during steps 305 , 310 , 315 and 320 ( or 305 , 315 a and 320 a ). in a fourth example , the optional test conditions are applied during steps 305 , 310 , 315 , 320 and 325 ( or 305 , 315 a , 320 a and 325 a ). in step 335 , the ic is removed from the mri chamber and it is determined whether the micro - electronic device was a known micro - electronic device or an unknown micro - electronic device . if the micro - electronic device was a known micro - electronic device then in step 340 analyses are performed and the analyses data is stored . if the micro - electronic device was an unknown micro - electronic device then in step 345 analyses are performed and the analyses is compared to analyses previously stored from a known micro - electronic device mri \u201c images \u201d under the same mri ( and bias / pattern ) conditions . if the compare is within predefined limits the unknown micro - electronic device is validated . it does not matter whether the known or unknown micro - electronic device is run first , but the comparison cannot be performed until both known or unknown micro - electronic devices have been run and the respective data analyzed . fig8 a and 8b illustrate a first method of data analysis for comparison according to embodiments of the present invention . fig8 a is a plot of the transverse component of the returned rf signal for a known micro - electronic device 350 and an unknown micro - electronic device 355 as relative rf strength versus time . direct comparison or statistical analysis of the difference between curves 350 and 355 is performed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig8 a , the difference between curve 355 relative to curve 350 is shown as a positive time shift . the shift may be negative . fig8 b is a plot of the longitudinal component of the returned rf signal for a known micro - electronic device 360 and an unknown micro - electronic device 365 as relative rf strength versus time . direct comparison or statistical analysis of the difference between curves 360 and 365 is performed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig8 b , the difference of curve 365 relative to curve 360 is shown as a negative time shift . the shift may be positive . fig9 a and 9b illustrate a second method of data analysis for comparison according to embodiments of the present invention . fig9 a is a plot of a fast fourier transform of the returned rf signal for a known ic ( or integrated circuit ) as rf amplitude versus rf frequency . element 370 is the returned signal and elements 370 a and 370 b are harmonic artifacts of element 370 . fig9 b is a plot of a fast fourier transform of the returned rf signal for an unknown ic ( or integrated circuit ) as rf amplitude versus rf frequency . element 375 is the returned signal and elements 375 a and 3750 b are harmonic artifacts of element 370 . element 370 and 375 are statistically compared in terms of amplitude of a given frequency range to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig9 a and 9b , the amplitude of the returned rf signal has been transferred from a time domain to a frequency domain . fig1 a , 10 b and 10 c illustrate a third method of data analysis for comparison according to embodiments of the present invention . fig1 a is a plot of the returned rf signal as relative rf signal strength versus time for a known micro - electronic device . rf signal 380 has a fixed amplitude between times a and b measured from when the rf pulse signal terminated ( or other convenient time reference ). an unknown micro - electronic device may exhibit a signal that is offset ion amplitude , phase , frequency or combinations thereof . two simple examples are given in fig1 b and 10c . fig1 b is a plot of the returned rf signal as relative rf signal strength versus time for an unknown micro - electronic device having only a frequency shift . rf signal 385 a has fixed amplitude between times a \u2032 and b \u2032 measured from the same reference as a and b of fig1 a . the shift between a and a \u2032 and b and b \u2032 is statistically to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . fig1 c is a plot of the returned rf signal as relative rf signal strength versus time for an unknown micro - electronic device having only an amplitude shift . rf signal 385 b has a fixed amplitude between times a and b measured from the same reference as a and b of fig1 a . the change in amplitude between curve 380 of fig1 a and curve 385 b of fig1 c is statistically to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . it will be appreciated that a complete statistical analysis would compare time , amplitude , frequency and phase differences of the curves described in fig1 a , 10 b and 10 c . fig1 a , 11 b and 11 c illustrate a fourth method of data analysis for comparison according to embodiments of the present invention . fig1 a is a plot of the returned rf signal xy space ( at a selected z ) as a function of frequency for a known micro - electronic device . as such fig1 a is closer to what is may be considered an \u201c image .\u201d fig1 b is a plot of the returned rf signal xy space ( at the selected z ) as a function of frequency for an unknown micro - electronic device . fig1 c is a plot of the delta between the plot of fig1 a and that of fig1 b . the amount of structure ( and optionally the position of the structures ) is statistically analyzed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . fig1 c is fig1 a \u201c subtracted \u201d from fig1 a . thus the embodiments of the present invention allow for relatively quick and inexpensive validation of integrated circuit chips . the description of the embodiments of the present invention is given above for the understanding of the present invention . it will be understood that the invention is not limited to the particular embodiments described herein , but is capable of various modifications , rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention . therefore , it is intended that the following claims cover all such modifications and changes as fall within the true spirit and scope of the invention ."}
|
{"category": "Electricity", "patent": "the term \u201c integrated circuit \u201d ( ic ) is defined as an integrated circuit chip and a package or module containing the ic . the term \u201c integrated circuit chip \u201d is defined as the semiconductor ( e . g ., silicon ) die containing devices such as transistors , diodes , capacitors , resisters and inductors and the wiring layers built up on the die that interconnect the devices into circuits . the term micro - electronic device includes an ic or an integrated circuit chip . in magnetic resonance imaging ( also called nuclear magnetic resonance ( nmr ) imaging )) a sample is positioned in a static magnetic field and subjected to a pulsed radio frequency ( rf ) signal to place the sample in an excited state . the magnetic field may be turned off or adjusted and a rf return signal is produced by the sample returning to a normal from the excited state is then recorded . in order to allow spatial encoding , the static magnetic field is superimposed with a gradient magnetic field . the term validation means the process by which an unknown micro - electronic device is \u201c imaged \u201d by nmr and the resulting data is compared to data from a known good or trusted micro - electronic that was nmr \u201c imaged \u201d and that the unknown ic micro - electronic device should be essentially identical ( i . e ., of identical design and within fabrication specification limits ) to . known good micro - electronic devices include those fabricated under secure conditions , those from trusted sources and those thoroughly tested and physically inspected after an nmr signature has been obtained . fig1 is a diagram of an exemplary magnetic resonance inspection system according of the present invention . in fig1 , a magnetic resonance imaging ( mri ) system 100 includes a magnet unit 105 , a signal processing unit 110 and a computer 115 . magnet unit 105 of mri system 100 includes a gradient magnetic field coil 120 , a transmit coil 125 , a sample chamber 130 , a receive coil and a high gauss magnet 140 ( e . g ., a permanent magnet , coil magnet or super - conductive magnet ). signal processing unit 110 includes a driving circuit 145 , an rf power amplifier 150 , a preamplifier 155 , a sequence memory 160 , a modulation circuit 165 , an rf oscillator 170 , a phase detector 175 and an analog - to - digital ( a / d ) converter 180 . computer 115 includes a display 185 , an input ( e . g ., a keyboard , mouse , disk drive ) and an output ( e . g ., a display unit , a printer , a disk drive ). gradient magnetic field coil 120 , transmit coil 125 , receive coil 135 and high gauss magnet 140 are disposed so as to substantially surround sample chamber 130 . high gauss magnet 140 applies a static magnetic field having a constant strength to a sample in chamber 130 . gradient magnetic field coil 120 applies gradient magnetic fields selectively to mutually orthogonal x , y and z directions ( in imaging parlance , to a slice axis , phase axis and frequency axis ). transmit coil 125 supplies a pulsed rf signal for exciting spins of atomic nuclei within a micro - electronic device 200 in sample chamber 130 . receive coil 135 detects returned rf signals from the sample in chamber 130 generated when spins of atomic nuclei within micro - electronic device 200 return to a normal state from an exited state . gradient magnetic field coil 120 , transmit coil 125 , and receive coil 135 are operatively associated with driving circuit 145 , an rf power amplifier 150 , and a preamplifier 155 , respectively . sequence memory 160 operates driving circuit 145 based on a stored pulse sequence in response to instructions from computer 115 to thereby apply gradient fields from gradient magnetic field coil 120 in specific directions . sequence memory 160 also operates a modulation circuit 165 to modulate a carrier output signal from rf oscillator 170 into a pulsed rf signal of predefined timing and envelope shape . the pulsed rf signal is applied to rf power amplifier 150 and then the amplified pulsed rf signal is applied to transmit coil 125 . preamplifier 155 amplifies the return rf signal from micro - electronic device 200 in sample chamber 130 detected at receive coil 135 . preamplifier 155 amplifies the received rf signal and sends an amplified rf signal to phase detector 175 . phase detector 175 generates an analog phase - detect signal from the amplified rf signal using a carrier output signal from rf oscillator 170 as a reference signal , and supplies the phase - detected signal to a / d converter 180 . a / d converter 180 converts the phase - detected analog signal into a digital signal , which is supplied to the computer 115 . computer 115 reads and / or processes the data from a / d converter 1801 , and includes algorithms in the form of computer instructions which when executed perform various signal analyses and statistical analyses on the stored data as described infra . results of these analyses may be displayed on output unit 195 . computer 115 can also be responsible for overall control such as receiving information supplied from input 195 from an operator . mri system 100 includes an optional tester 205 , which is connected between a socket , or probe card ( not shown ) in sample chamber 130 and computer 185 . this allows voltage bias , analog signals , digital data patterns or combinations thereof to be applied to micro - electronic device 200 during the mri process . biasing , applying signals and test patterns serves two purposes . first it results in more complex return rf signals . second , if the biasing analog signals , digital data patterns are kept secure , it is very difficult for a an unauthorized party to place a masking circuit into an unauthorized micro electronic device , the purpose of the masking circuit being is to mask the presence of the unauthorized circuit and the masking circuit in the unauthorized micro - electronic device by altering the nmr image of the unauthorized micro - electronic device to mimic that of an authorized micro - electronic device . mri system 100 shown in fig1 is provided as an example , and it will be understood that embodiments of the invention are not limited to mri system 100 shown in fig1 . it will be understood that an mri system 100 according to aspects of the invention can include additional components to those shown in fig1 or may not include every component shown in fig1 . fig2 is a cross - section through an exemplary first type of integrated circuit . in fig2 , an ic 200 a includes an integrated circuit chip 210 physically and electrically connected to a module 215 by solder bumps 220 . wires 225 in module 225 connect solder bumps 220 to solder balls 230 . module 215 may be organic ( e . g ., fiberglass ) or ceramic and wires 25 may comprise one or more wiring layers . solder balls 230 are designed for surface mounting ic 200 a directly to a printed circuit board ( pcb ). solder balls 230 may be replaced by solder columns . solder balls 230 may be replaced with pins , which can be used to mount ics into sockets , which are mounted on a pcb or mounted directly to a pcb . ic 200 a is thus an example of an ic that uses flip - chip ( or c4 ) technology . fig3 is a cross - section through an exemplary second type of integrated circuit . in fig3 , an ic 200 b includes and integrated circuit chip contained within a plastic package 240 . wire bonds 245 electrically connect integrated circuit chip 235 to leads 250 . leads 250 are designed for surface mounting ic 200 b directly to a pcb . leads 250 may be replaced with pins , which can be used to mount ics into sockets , which are mounted on a pcb or mounted directly to a pcb . the leads on some plastic packages are designed to be mounted in sockets on a pcb . ic 200 b is thus an example of plastic packaging technology . fig4 is a top view of an exemplary integrated circuit chip for enhanced comparison according embodiments of the present invention . in fig4 , an exemplary integrated circuit chip 255 includes regions 260 a , 260 b , 260 c , 260 d , 260 e , 260 f , 260 g , 260 h , 2601 , 269 j and 260 k . there may be more or less regions than illustrated in fig4 . these regions often correspond to cores , which are pre - designed circuit functions . for example , a microprocessor may contain multiple processing cores , memory cores , arithmetic cores etc . formed , by way of example , in cores 260 b , 260 g , 260 i and 260 j are serpentine signal enhancing structures 265 which are not electrically connected to any wire or device ( e . g ., transistor , diode , resistor , capacitor or inductor ) of integrated circuit chip 255 . signal enhancing structures 265 may comprise an electrical conductor , a magnetic material , or an electrically conductive magnetic material . signal enhancing structures 265 are designed to interact with the magnetic fields from gradient magnetic field coil 120 and high gauss magnet 140 of mri system 100 ( see fig1 ) to generate a more complex return rf signal . fig5 is a top view of an exemplary integrated circuit chip for preventing or detecting comparison according embodiments of the present invention . in fig5 , an integrated circuit chip 255 a ( similar to integrated circuit chip 255 of fig4 ) includes a serpentine structure 265 a connected to a destruct circuit 270 . serpentine structure 265 a acts as an inductor which generates a current when subjected to a varying magnetic field . the current may be used by destruct circuit 270 to program fuses or activate transistors to render integrated circuit 155 a inoperable or to leave a signature that can later be read to indicate if an attempt at nmr imaging \u201d has been performed on integrated circuit chip 255 a . in one example , serpentine structure 265 a is designed to not be detectable by x - ray imaging . fig6 a is a cross - sectional view of an integrated circuit containing devices to flag an unauthorized comparison attempt according embodiments of the present invention . in fig6 a , an ic 200 c includes an integrated circuit chip 235 a in a plastic package body 240 a . a set ( two sets are shown in the example of fig6 a ) of three \u201c horseshoe \u201d magnets 275 a , 275 b and 275 c are placed in body 240 a . they are aligned so respective lines ( dashed lines ) passing through the poles of each magnet are mutually orthogonal . in fig6 a , the line passing through the poles of magnets 275 a and 275 c are in planes parallel to the top surface 272 of integrated circuit chip 235 a . other pole orientations are possible . in one example , only a single horseshoe magnet is used . horseshoe shaped magnets may be replaced by other shaped magnets such as bar magnets and disc magnets . fig6 b is a cross - sectional view of the integrated circuit of fig6 a after an unauthorized comparison attempt according embodiments of the present invention . when magnets 275 a , 275 b and 275 c are subjected to the intense magnetic field generated in an nmr machine , magnets 275 a , 275 b and 275 c are pulled / pushed by that field so strongly that cracks 280 are formed in body 240 a . fig7 is a flowchart of methods of comparison according to embodiments of the present invention . the steps of the flowchart of fig7 are performed first on a known or trusted micro - electronic device and then on an unknown ( or suspect ) micro - electronic device that and essentially identical to the known micro - electronic device ( i . e ., identically designed and within fabrication specification limits ). for the integrated circuit chip specification limits define allowable variations in material and structure and include , for example , the allowable differences in metal line widths and thickness differences in metal and insulating layers . for the package of the ic specification limits define allowable variations in material and structure and include , for example , package dimensions , size of solder bumps , positions and size of wire bonds , widths and thickness of land in modules . in step 300 an ic ( or integrated circuit is placed in the mri chamber . in step 305 the high gauss magnetic field is turned on . in one example , the high gauss magnetic field has a field strength of between about a 0 . 5 tesla and about 10 tesla and is applied in the z direction . the method can know follow one of two mutually exclusive paths . the first is the path ( 1 ) through steps 310 , 315 , 320 , 325 and 330 to step 335 . the second path ( 2 ) is through steps 315 a , 320 a and 325 a to step 335 . the unknown micro - electronic device advantageously follows the same path and is subjected to the same nmr conditions as the known micro - electronic device . in step 310 , a direction ( x , y or z ) is selected . in step 315 , a gradient magnetic field of , for example , 1 tesla is applied in the selected direction and in step 320 a rf pulse is directed to the ic . in step 325 the emitted ( returned ) rf signal is detected and information describing the return rf signal ( which is also a pulse ) is stored . the duration and time of reception of the return rf signal will vary . in step 330 , if another direction of the three possible directions ( x , y , z ) is selected and the method loops back to step 310 . this loop will repeat three times , once for each direction . in path ( 2 ) in step 315 a , gradient magnetic fields are applied in the x , y , and z directions simultaneously . each gradient field may have a same or a different field strength of between about 0 . 5 tesla and about 10 tesla . steps 320 a and 325 a are similar to steps 320 and 325 respectively . in steps 315 and 315 a , optional test conditions ( i . e ., voltage bias , analog signal , digital data pattern or combinations thereof ) may be applied to the micro - electronic device . in a first example , optional test conditions are applied only during step 315 ( or 315 a ). in a second example , optional test conditions are applied the only during steps 315 and 320 ( or 315 a and 320 a ). in a third example , the optional test conditions are applied the only during steps 305 , 310 , 315 and 320 ( or 305 , 315 a and 320 a ). in a fourth example , the optional test conditions are applied during steps 305 , 310 , 315 , 320 and 325 ( or 305 , 315 a , 320 a and 325 a ). in step 335 , the ic is removed from the mri chamber and it is determined whether the micro - electronic device was a known micro - electronic device or an unknown micro - electronic device . if the micro - electronic device was a known micro - electronic device then in step 340 analyses are performed and the analyses data is stored . if the micro - electronic device was an unknown micro - electronic device then in step 345 analyses are performed and the analyses is compared to analyses previously stored from a known micro - electronic device mri \u201c images \u201d under the same mri ( and bias / pattern ) conditions . if the compare is within predefined limits the unknown micro - electronic device is validated . it does not matter whether the known or unknown micro - electronic device is run first , but the comparison cannot be performed until both known or unknown micro - electronic devices have been run and the respective data analyzed . fig8 a and 8b illustrate a first method of data analysis for comparison according to embodiments of the present invention . fig8 a is a plot of the transverse component of the returned rf signal for a known micro - electronic device 350 and an unknown micro - electronic device 355 as relative rf strength versus time . direct comparison or statistical analysis of the difference between curves 350 and 355 is performed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig8 a , the difference between curve 355 relative to curve 350 is shown as a positive time shift . the shift may be negative . fig8 b is a plot of the longitudinal component of the returned rf signal for a known micro - electronic device 360 and an unknown micro - electronic device 365 as relative rf strength versus time . direct comparison or statistical analysis of the difference between curves 360 and 365 is performed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig8 b , the difference of curve 365 relative to curve 360 is shown as a negative time shift . the shift may be positive . fig9 a and 9b illustrate a second method of data analysis for comparison according to embodiments of the present invention . fig9 a is a plot of a fast fourier transform of the returned rf signal for a known ic ( or integrated circuit ) as rf amplitude versus rf frequency . element 370 is the returned signal and elements 370 a and 370 b are harmonic artifacts of element 370 . fig9 b is a plot of a fast fourier transform of the returned rf signal for an unknown ic ( or integrated circuit ) as rf amplitude versus rf frequency . element 375 is the returned signal and elements 375 a and 3750 b are harmonic artifacts of element 370 . element 370 and 375 are statistically compared in terms of amplitude of a given frequency range to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig9 a and 9b , the amplitude of the returned rf signal has been transferred from a time domain to a frequency domain . fig1 a , 10 b and 10 c illustrate a third method of data analysis for comparison according to embodiments of the present invention . fig1 a is a plot of the returned rf signal as relative rf signal strength versus time for a known micro - electronic device . rf signal 380 has a fixed amplitude between times a and b measured from when the rf pulse signal terminated ( or other convenient time reference ). an unknown micro - electronic device may exhibit a signal that is offset ion amplitude , phase , frequency or combinations thereof . two simple examples are given in fig1 b and 10c . fig1 b is a plot of the returned rf signal as relative rf signal strength versus time for an unknown micro - electronic device having only a frequency shift . rf signal 385 a has fixed amplitude between times a \u2032 and b \u2032 measured from the same reference as a and b of fig1 a . the shift between a and a \u2032 and b and b \u2032 is statistically to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . fig1 c is a plot of the returned rf signal as relative rf signal strength versus time for an unknown micro - electronic device having only an amplitude shift . rf signal 385 b has a fixed amplitude between times a and b measured from the same reference as a and b of fig1 a . the change in amplitude between curve 380 of fig1 a and curve 385 b of fig1 c is statistically to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . it will be appreciated that a complete statistical analysis would compare time , amplitude , frequency and phase differences of the curves described in fig1 a , 10 b and 10 c . fig1 a , 11 b and 11 c illustrate a fourth method of data analysis for comparison according to embodiments of the present invention . fig1 a is a plot of the returned rf signal xy space ( at a selected z ) as a function of frequency for a known micro - electronic device . as such fig1 a is closer to what is may be considered an \u201c image .\u201d fig1 b is a plot of the returned rf signal xy space ( at the selected z ) as a function of frequency for an unknown micro - electronic device . fig1 c is a plot of the delta between the plot of fig1 a and that of fig1 b . the amount of structure ( and optionally the position of the structures ) is statistically analyzed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . fig1 c is fig1 a \u201c subtracted \u201d from fig1 a . thus the embodiments of the present invention allow for relatively quick and inexpensive validation of integrated circuit chips . the description of the embodiments of the present invention is given above for the understanding of the present invention . it will be understood that the invention is not limited to the particular embodiments described herein , but is capable of various modifications , rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention . therefore , it is intended that the following claims cover all such modifications and changes as fall within the true spirit and scope of the invention ."}
|
Is the category the most suitable category for the given patent?
| 0.25 |
d2a533e20b88cf12a66089384d35063da6ac708b8421ee9a63cae12490153828
| 0.041504 | 0.09668 | 0.013245 | 0.051758 | 0.375 | 0.131836 |
null |
{"patent": "the term \u201c integrated circuit \u201d ( ic ) is defined as an integrated circuit chip and a package or module containing the ic . the term \u201c integrated circuit chip \u201d is defined as the semiconductor ( e . g ., silicon ) die containing devices such as transistors , diodes , capacitors , resisters and inductors and the wiring layers built up on the die that interconnect the devices into circuits . the term micro - electronic device includes an ic or an integrated circuit chip . in magnetic resonance imaging ( also called nuclear magnetic resonance ( nmr ) imaging )) a sample is positioned in a static magnetic field and subjected to a pulsed radio frequency ( rf ) signal to place the sample in an excited state . the magnetic field may be turned off or adjusted and a rf return signal is produced by the sample returning to a normal from the excited state is then recorded . in order to allow spatial encoding , the static magnetic field is superimposed with a gradient magnetic field . the term validation means the process by which an unknown micro - electronic device is \u201c imaged \u201d by nmr and the resulting data is compared to data from a known good or trusted micro - electronic that was nmr \u201c imaged \u201d and that the unknown ic micro - electronic device should be essentially identical ( i . e ., of identical design and within fabrication specification limits ) to . known good micro - electronic devices include those fabricated under secure conditions , those from trusted sources and those thoroughly tested and physically inspected after an nmr signature has been obtained . fig1 is a diagram of an exemplary magnetic resonance inspection system according of the present invention . in fig1 , a magnetic resonance imaging ( mri ) system 100 includes a magnet unit 105 , a signal processing unit 110 and a computer 115 . magnet unit 105 of mri system 100 includes a gradient magnetic field coil 120 , a transmit coil 125 , a sample chamber 130 , a receive coil and a high gauss magnet 140 ( e . g ., a permanent magnet , coil magnet or super - conductive magnet ). signal processing unit 110 includes a driving circuit 145 , an rf power amplifier 150 , a preamplifier 155 , a sequence memory 160 , a modulation circuit 165 , an rf oscillator 170 , a phase detector 175 and an analog - to - digital ( a / d ) converter 180 . computer 115 includes a display 185 , an input ( e . g ., a keyboard , mouse , disk drive ) and an output ( e . g ., a display unit , a printer , a disk drive ). gradient magnetic field coil 120 , transmit coil 125 , receive coil 135 and high gauss magnet 140 are disposed so as to substantially surround sample chamber 130 . high gauss magnet 140 applies a static magnetic field having a constant strength to a sample in chamber 130 . gradient magnetic field coil 120 applies gradient magnetic fields selectively to mutually orthogonal x , y and z directions ( in imaging parlance , to a slice axis , phase axis and frequency axis ). transmit coil 125 supplies a pulsed rf signal for exciting spins of atomic nuclei within a micro - electronic device 200 in sample chamber 130 . receive coil 135 detects returned rf signals from the sample in chamber 130 generated when spins of atomic nuclei within micro - electronic device 200 return to a normal state from an exited state . gradient magnetic field coil 120 , transmit coil 125 , and receive coil 135 are operatively associated with driving circuit 145 , an rf power amplifier 150 , and a preamplifier 155 , respectively . sequence memory 160 operates driving circuit 145 based on a stored pulse sequence in response to instructions from computer 115 to thereby apply gradient fields from gradient magnetic field coil 120 in specific directions . sequence memory 160 also operates a modulation circuit 165 to modulate a carrier output signal from rf oscillator 170 into a pulsed rf signal of predefined timing and envelope shape . the pulsed rf signal is applied to rf power amplifier 150 and then the amplified pulsed rf signal is applied to transmit coil 125 . preamplifier 155 amplifies the return rf signal from micro - electronic device 200 in sample chamber 130 detected at receive coil 135 . preamplifier 155 amplifies the received rf signal and sends an amplified rf signal to phase detector 175 . phase detector 175 generates an analog phase - detect signal from the amplified rf signal using a carrier output signal from rf oscillator 170 as a reference signal , and supplies the phase - detected signal to a / d converter 180 . a / d converter 180 converts the phase - detected analog signal into a digital signal , which is supplied to the computer 115 . computer 115 reads and / or processes the data from a / d converter 1801 , and includes algorithms in the form of computer instructions which when executed perform various signal analyses and statistical analyses on the stored data as described infra . results of these analyses may be displayed on output unit 195 . computer 115 can also be responsible for overall control such as receiving information supplied from input 195 from an operator . mri system 100 includes an optional tester 205 , which is connected between a socket , or probe card ( not shown ) in sample chamber 130 and computer 185 . this allows voltage bias , analog signals , digital data patterns or combinations thereof to be applied to micro - electronic device 200 during the mri process . biasing , applying signals and test patterns serves two purposes . first it results in more complex return rf signals . second , if the biasing analog signals , digital data patterns are kept secure , it is very difficult for a an unauthorized party to place a masking circuit into an unauthorized micro electronic device , the purpose of the masking circuit being is to mask the presence of the unauthorized circuit and the masking circuit in the unauthorized micro - electronic device by altering the nmr image of the unauthorized micro - electronic device to mimic that of an authorized micro - electronic device . mri system 100 shown in fig1 is provided as an example , and it will be understood that embodiments of the invention are not limited to mri system 100 shown in fig1 . it will be understood that an mri system 100 according to aspects of the invention can include additional components to those shown in fig1 or may not include every component shown in fig1 . fig2 is a cross - section through an exemplary first type of integrated circuit . in fig2 , an ic 200 a includes an integrated circuit chip 210 physically and electrically connected to a module 215 by solder bumps 220 . wires 225 in module 225 connect solder bumps 220 to solder balls 230 . module 215 may be organic ( e . g ., fiberglass ) or ceramic and wires 25 may comprise one or more wiring layers . solder balls 230 are designed for surface mounting ic 200 a directly to a printed circuit board ( pcb ). solder balls 230 may be replaced by solder columns . solder balls 230 may be replaced with pins , which can be used to mount ics into sockets , which are mounted on a pcb or mounted directly to a pcb . ic 200 a is thus an example of an ic that uses flip - chip ( or c4 ) technology . fig3 is a cross - section through an exemplary second type of integrated circuit . in fig3 , an ic 200 b includes and integrated circuit chip contained within a plastic package 240 . wire bonds 245 electrically connect integrated circuit chip 235 to leads 250 . leads 250 are designed for surface mounting ic 200 b directly to a pcb . leads 250 may be replaced with pins , which can be used to mount ics into sockets , which are mounted on a pcb or mounted directly to a pcb . the leads on some plastic packages are designed to be mounted in sockets on a pcb . ic 200 b is thus an example of plastic packaging technology . fig4 is a top view of an exemplary integrated circuit chip for enhanced comparison according embodiments of the present invention . in fig4 , an exemplary integrated circuit chip 255 includes regions 260 a , 260 b , 260 c , 260 d , 260 e , 260 f , 260 g , 260 h , 2601 , 269 j and 260 k . there may be more or less regions than illustrated in fig4 . these regions often correspond to cores , which are pre - designed circuit functions . for example , a microprocessor may contain multiple processing cores , memory cores , arithmetic cores etc . formed , by way of example , in cores 260 b , 260 g , 260 i and 260 j are serpentine signal enhancing structures 265 which are not electrically connected to any wire or device ( e . g ., transistor , diode , resistor , capacitor or inductor ) of integrated circuit chip 255 . signal enhancing structures 265 may comprise an electrical conductor , a magnetic material , or an electrically conductive magnetic material . signal enhancing structures 265 are designed to interact with the magnetic fields from gradient magnetic field coil 120 and high gauss magnet 140 of mri system 100 ( see fig1 ) to generate a more complex return rf signal . fig5 is a top view of an exemplary integrated circuit chip for preventing or detecting comparison according embodiments of the present invention . in fig5 , an integrated circuit chip 255 a ( similar to integrated circuit chip 255 of fig4 ) includes a serpentine structure 265 a connected to a destruct circuit 270 . serpentine structure 265 a acts as an inductor which generates a current when subjected to a varying magnetic field . the current may be used by destruct circuit 270 to program fuses or activate transistors to render integrated circuit 155 a inoperable or to leave a signature that can later be read to indicate if an attempt at nmr imaging \u201d has been performed on integrated circuit chip 255 a . in one example , serpentine structure 265 a is designed to not be detectable by x - ray imaging . fig6 a is a cross - sectional view of an integrated circuit containing devices to flag an unauthorized comparison attempt according embodiments of the present invention . in fig6 a , an ic 200 c includes an integrated circuit chip 235 a in a plastic package body 240 a . a set ( two sets are shown in the example of fig6 a ) of three \u201c horseshoe \u201d magnets 275 a , 275 b and 275 c are placed in body 240 a . they are aligned so respective lines ( dashed lines ) passing through the poles of each magnet are mutually orthogonal . in fig6 a , the line passing through the poles of magnets 275 a and 275 c are in planes parallel to the top surface 272 of integrated circuit chip 235 a . other pole orientations are possible . in one example , only a single horseshoe magnet is used . horseshoe shaped magnets may be replaced by other shaped magnets such as bar magnets and disc magnets . fig6 b is a cross - sectional view of the integrated circuit of fig6 a after an unauthorized comparison attempt according embodiments of the present invention . when magnets 275 a , 275 b and 275 c are subjected to the intense magnetic field generated in an nmr machine , magnets 275 a , 275 b and 275 c are pulled / pushed by that field so strongly that cracks 280 are formed in body 240 a . fig7 is a flowchart of methods of comparison according to embodiments of the present invention . the steps of the flowchart of fig7 are performed first on a known or trusted micro - electronic device and then on an unknown ( or suspect ) micro - electronic device that and essentially identical to the known micro - electronic device ( i . e ., identically designed and within fabrication specification limits ). for the integrated circuit chip specification limits define allowable variations in material and structure and include , for example , the allowable differences in metal line widths and thickness differences in metal and insulating layers . for the package of the ic specification limits define allowable variations in material and structure and include , for example , package dimensions , size of solder bumps , positions and size of wire bonds , widths and thickness of land in modules . in step 300 an ic ( or integrated circuit is placed in the mri chamber . in step 305 the high gauss magnetic field is turned on . in one example , the high gauss magnetic field has a field strength of between about a 0 . 5 tesla and about 10 tesla and is applied in the z direction . the method can know follow one of two mutually exclusive paths . the first is the path ( 1 ) through steps 310 , 315 , 320 , 325 and 330 to step 335 . the second path ( 2 ) is through steps 315 a , 320 a and 325 a to step 335 . the unknown micro - electronic device advantageously follows the same path and is subjected to the same nmr conditions as the known micro - electronic device . in step 310 , a direction ( x , y or z ) is selected . in step 315 , a gradient magnetic field of , for example , 1 tesla is applied in the selected direction and in step 320 a rf pulse is directed to the ic . in step 325 the emitted ( returned ) rf signal is detected and information describing the return rf signal ( which is also a pulse ) is stored . the duration and time of reception of the return rf signal will vary . in step 330 , if another direction of the three possible directions ( x , y , z ) is selected and the method loops back to step 310 . this loop will repeat three times , once for each direction . in path ( 2 ) in step 315 a , gradient magnetic fields are applied in the x , y , and z directions simultaneously . each gradient field may have a same or a different field strength of between about 0 . 5 tesla and about 10 tesla . steps 320 a and 325 a are similar to steps 320 and 325 respectively . in steps 315 and 315 a , optional test conditions ( i . e ., voltage bias , analog signal , digital data pattern or combinations thereof ) may be applied to the micro - electronic device . in a first example , optional test conditions are applied only during step 315 ( or 315 a ). in a second example , optional test conditions are applied the only during steps 315 and 320 ( or 315 a and 320 a ). in a third example , the optional test conditions are applied the only during steps 305 , 310 , 315 and 320 ( or 305 , 315 a and 320 a ). in a fourth example , the optional test conditions are applied during steps 305 , 310 , 315 , 320 and 325 ( or 305 , 315 a , 320 a and 325 a ). in step 335 , the ic is removed from the mri chamber and it is determined whether the micro - electronic device was a known micro - electronic device or an unknown micro - electronic device . if the micro - electronic device was a known micro - electronic device then in step 340 analyses are performed and the analyses data is stored . if the micro - electronic device was an unknown micro - electronic device then in step 345 analyses are performed and the analyses is compared to analyses previously stored from a known micro - electronic device mri \u201c images \u201d under the same mri ( and bias / pattern ) conditions . if the compare is within predefined limits the unknown micro - electronic device is validated . it does not matter whether the known or unknown micro - electronic device is run first , but the comparison cannot be performed until both known or unknown micro - electronic devices have been run and the respective data analyzed . fig8 a and 8b illustrate a first method of data analysis for comparison according to embodiments of the present invention . fig8 a is a plot of the transverse component of the returned rf signal for a known micro - electronic device 350 and an unknown micro - electronic device 355 as relative rf strength versus time . direct comparison or statistical analysis of the difference between curves 350 and 355 is performed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig8 a , the difference between curve 355 relative to curve 350 is shown as a positive time shift . the shift may be negative . fig8 b is a plot of the longitudinal component of the returned rf signal for a known micro - electronic device 360 and an unknown micro - electronic device 365 as relative rf strength versus time . direct comparison or statistical analysis of the difference between curves 360 and 365 is performed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig8 b , the difference of curve 365 relative to curve 360 is shown as a negative time shift . the shift may be positive . fig9 a and 9b illustrate a second method of data analysis for comparison according to embodiments of the present invention . fig9 a is a plot of a fast fourier transform of the returned rf signal for a known ic ( or integrated circuit ) as rf amplitude versus rf frequency . element 370 is the returned signal and elements 370 a and 370 b are harmonic artifacts of element 370 . fig9 b is a plot of a fast fourier transform of the returned rf signal for an unknown ic ( or integrated circuit ) as rf amplitude versus rf frequency . element 375 is the returned signal and elements 375 a and 3750 b are harmonic artifacts of element 370 . element 370 and 375 are statistically compared in terms of amplitude of a given frequency range to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig9 a and 9b , the amplitude of the returned rf signal has been transferred from a time domain to a frequency domain . fig1 a , 10 b and 10 c illustrate a third method of data analysis for comparison according to embodiments of the present invention . fig1 a is a plot of the returned rf signal as relative rf signal strength versus time for a known micro - electronic device . rf signal 380 has a fixed amplitude between times a and b measured from when the rf pulse signal terminated ( or other convenient time reference ). an unknown micro - electronic device may exhibit a signal that is offset ion amplitude , phase , frequency or combinations thereof . two simple examples are given in fig1 b and 10c . fig1 b is a plot of the returned rf signal as relative rf signal strength versus time for an unknown micro - electronic device having only a frequency shift . rf signal 385 a has fixed amplitude between times a \u2032 and b \u2032 measured from the same reference as a and b of fig1 a . the shift between a and a \u2032 and b and b \u2032 is statistically to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . fig1 c is a plot of the returned rf signal as relative rf signal strength versus time for an unknown micro - electronic device having only an amplitude shift . rf signal 385 b has a fixed amplitude between times a and b measured from the same reference as a and b of fig1 a . the change in amplitude between curve 380 of fig1 a and curve 385 b of fig1 c is statistically to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . it will be appreciated that a complete statistical analysis would compare time , amplitude , frequency and phase differences of the curves described in fig1 a , 10 b and 10 c . fig1 a , 11 b and 11 c illustrate a fourth method of data analysis for comparison according to embodiments of the present invention . fig1 a is a plot of the returned rf signal xy space ( at a selected z ) as a function of frequency for a known micro - electronic device . as such fig1 a is closer to what is may be considered an \u201c image .\u201d fig1 b is a plot of the returned rf signal xy space ( at the selected z ) as a function of frequency for an unknown micro - electronic device . fig1 c is a plot of the delta between the plot of fig1 a and that of fig1 b . the amount of structure ( and optionally the position of the structures ) is statistically analyzed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . fig1 c is fig1 a \u201c subtracted \u201d from fig1 a . thus the embodiments of the present invention allow for relatively quick and inexpensive validation of integrated circuit chips . the description of the embodiments of the present invention is given above for the understanding of the present invention . it will be understood that the invention is not limited to the particular embodiments described herein , but is capable of various modifications , rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention . therefore , it is intended that the following claims cover all such modifications and changes as fall within the true spirit and scope of the invention .", "category": "Physics"}
|
{"category": "General tagging of new or cross-sectional technology", "patent": "the term \u201c integrated circuit \u201d ( ic ) is defined as an integrated circuit chip and a package or module containing the ic . the term \u201c integrated circuit chip \u201d is defined as the semiconductor ( e . g ., silicon ) die containing devices such as transistors , diodes , capacitors , resisters and inductors and the wiring layers built up on the die that interconnect the devices into circuits . the term micro - electronic device includes an ic or an integrated circuit chip . in magnetic resonance imaging ( also called nuclear magnetic resonance ( nmr ) imaging )) a sample is positioned in a static magnetic field and subjected to a pulsed radio frequency ( rf ) signal to place the sample in an excited state . the magnetic field may be turned off or adjusted and a rf return signal is produced by the sample returning to a normal from the excited state is then recorded . in order to allow spatial encoding , the static magnetic field is superimposed with a gradient magnetic field . the term validation means the process by which an unknown micro - electronic device is \u201c imaged \u201d by nmr and the resulting data is compared to data from a known good or trusted micro - electronic that was nmr \u201c imaged \u201d and that the unknown ic micro - electronic device should be essentially identical ( i . e ., of identical design and within fabrication specification limits ) to . known good micro - electronic devices include those fabricated under secure conditions , those from trusted sources and those thoroughly tested and physically inspected after an nmr signature has been obtained . fig1 is a diagram of an exemplary magnetic resonance inspection system according of the present invention . in fig1 , a magnetic resonance imaging ( mri ) system 100 includes a magnet unit 105 , a signal processing unit 110 and a computer 115 . magnet unit 105 of mri system 100 includes a gradient magnetic field coil 120 , a transmit coil 125 , a sample chamber 130 , a receive coil and a high gauss magnet 140 ( e . g ., a permanent magnet , coil magnet or super - conductive magnet ). signal processing unit 110 includes a driving circuit 145 , an rf power amplifier 150 , a preamplifier 155 , a sequence memory 160 , a modulation circuit 165 , an rf oscillator 170 , a phase detector 175 and an analog - to - digital ( a / d ) converter 180 . computer 115 includes a display 185 , an input ( e . g ., a keyboard , mouse , disk drive ) and an output ( e . g ., a display unit , a printer , a disk drive ). gradient magnetic field coil 120 , transmit coil 125 , receive coil 135 and high gauss magnet 140 are disposed so as to substantially surround sample chamber 130 . high gauss magnet 140 applies a static magnetic field having a constant strength to a sample in chamber 130 . gradient magnetic field coil 120 applies gradient magnetic fields selectively to mutually orthogonal x , y and z directions ( in imaging parlance , to a slice axis , phase axis and frequency axis ). transmit coil 125 supplies a pulsed rf signal for exciting spins of atomic nuclei within a micro - electronic device 200 in sample chamber 130 . receive coil 135 detects returned rf signals from the sample in chamber 130 generated when spins of atomic nuclei within micro - electronic device 200 return to a normal state from an exited state . gradient magnetic field coil 120 , transmit coil 125 , and receive coil 135 are operatively associated with driving circuit 145 , an rf power amplifier 150 , and a preamplifier 155 , respectively . sequence memory 160 operates driving circuit 145 based on a stored pulse sequence in response to instructions from computer 115 to thereby apply gradient fields from gradient magnetic field coil 120 in specific directions . sequence memory 160 also operates a modulation circuit 165 to modulate a carrier output signal from rf oscillator 170 into a pulsed rf signal of predefined timing and envelope shape . the pulsed rf signal is applied to rf power amplifier 150 and then the amplified pulsed rf signal is applied to transmit coil 125 . preamplifier 155 amplifies the return rf signal from micro - electronic device 200 in sample chamber 130 detected at receive coil 135 . preamplifier 155 amplifies the received rf signal and sends an amplified rf signal to phase detector 175 . phase detector 175 generates an analog phase - detect signal from the amplified rf signal using a carrier output signal from rf oscillator 170 as a reference signal , and supplies the phase - detected signal to a / d converter 180 . a / d converter 180 converts the phase - detected analog signal into a digital signal , which is supplied to the computer 115 . computer 115 reads and / or processes the data from a / d converter 1801 , and includes algorithms in the form of computer instructions which when executed perform various signal analyses and statistical analyses on the stored data as described infra . results of these analyses may be displayed on output unit 195 . computer 115 can also be responsible for overall control such as receiving information supplied from input 195 from an operator . mri system 100 includes an optional tester 205 , which is connected between a socket , or probe card ( not shown ) in sample chamber 130 and computer 185 . this allows voltage bias , analog signals , digital data patterns or combinations thereof to be applied to micro - electronic device 200 during the mri process . biasing , applying signals and test patterns serves two purposes . first it results in more complex return rf signals . second , if the biasing analog signals , digital data patterns are kept secure , it is very difficult for a an unauthorized party to place a masking circuit into an unauthorized micro electronic device , the purpose of the masking circuit being is to mask the presence of the unauthorized circuit and the masking circuit in the unauthorized micro - electronic device by altering the nmr image of the unauthorized micro - electronic device to mimic that of an authorized micro - electronic device . mri system 100 shown in fig1 is provided as an example , and it will be understood that embodiments of the invention are not limited to mri system 100 shown in fig1 . it will be understood that an mri system 100 according to aspects of the invention can include additional components to those shown in fig1 or may not include every component shown in fig1 . fig2 is a cross - section through an exemplary first type of integrated circuit . in fig2 , an ic 200 a includes an integrated circuit chip 210 physically and electrically connected to a module 215 by solder bumps 220 . wires 225 in module 225 connect solder bumps 220 to solder balls 230 . module 215 may be organic ( e . g ., fiberglass ) or ceramic and wires 25 may comprise one or more wiring layers . solder balls 230 are designed for surface mounting ic 200 a directly to a printed circuit board ( pcb ). solder balls 230 may be replaced by solder columns . solder balls 230 may be replaced with pins , which can be used to mount ics into sockets , which are mounted on a pcb or mounted directly to a pcb . ic 200 a is thus an example of an ic that uses flip - chip ( or c4 ) technology . fig3 is a cross - section through an exemplary second type of integrated circuit . in fig3 , an ic 200 b includes and integrated circuit chip contained within a plastic package 240 . wire bonds 245 electrically connect integrated circuit chip 235 to leads 250 . leads 250 are designed for surface mounting ic 200 b directly to a pcb . leads 250 may be replaced with pins , which can be used to mount ics into sockets , which are mounted on a pcb or mounted directly to a pcb . the leads on some plastic packages are designed to be mounted in sockets on a pcb . ic 200 b is thus an example of plastic packaging technology . fig4 is a top view of an exemplary integrated circuit chip for enhanced comparison according embodiments of the present invention . in fig4 , an exemplary integrated circuit chip 255 includes regions 260 a , 260 b , 260 c , 260 d , 260 e , 260 f , 260 g , 260 h , 2601 , 269 j and 260 k . there may be more or less regions than illustrated in fig4 . these regions often correspond to cores , which are pre - designed circuit functions . for example , a microprocessor may contain multiple processing cores , memory cores , arithmetic cores etc . formed , by way of example , in cores 260 b , 260 g , 260 i and 260 j are serpentine signal enhancing structures 265 which are not electrically connected to any wire or device ( e . g ., transistor , diode , resistor , capacitor or inductor ) of integrated circuit chip 255 . signal enhancing structures 265 may comprise an electrical conductor , a magnetic material , or an electrically conductive magnetic material . signal enhancing structures 265 are designed to interact with the magnetic fields from gradient magnetic field coil 120 and high gauss magnet 140 of mri system 100 ( see fig1 ) to generate a more complex return rf signal . fig5 is a top view of an exemplary integrated circuit chip for preventing or detecting comparison according embodiments of the present invention . in fig5 , an integrated circuit chip 255 a ( similar to integrated circuit chip 255 of fig4 ) includes a serpentine structure 265 a connected to a destruct circuit 270 . serpentine structure 265 a acts as an inductor which generates a current when subjected to a varying magnetic field . the current may be used by destruct circuit 270 to program fuses or activate transistors to render integrated circuit 155 a inoperable or to leave a signature that can later be read to indicate if an attempt at nmr imaging \u201d has been performed on integrated circuit chip 255 a . in one example , serpentine structure 265 a is designed to not be detectable by x - ray imaging . fig6 a is a cross - sectional view of an integrated circuit containing devices to flag an unauthorized comparison attempt according embodiments of the present invention . in fig6 a , an ic 200 c includes an integrated circuit chip 235 a in a plastic package body 240 a . a set ( two sets are shown in the example of fig6 a ) of three \u201c horseshoe \u201d magnets 275 a , 275 b and 275 c are placed in body 240 a . they are aligned so respective lines ( dashed lines ) passing through the poles of each magnet are mutually orthogonal . in fig6 a , the line passing through the poles of magnets 275 a and 275 c are in planes parallel to the top surface 272 of integrated circuit chip 235 a . other pole orientations are possible . in one example , only a single horseshoe magnet is used . horseshoe shaped magnets may be replaced by other shaped magnets such as bar magnets and disc magnets . fig6 b is a cross - sectional view of the integrated circuit of fig6 a after an unauthorized comparison attempt according embodiments of the present invention . when magnets 275 a , 275 b and 275 c are subjected to the intense magnetic field generated in an nmr machine , magnets 275 a , 275 b and 275 c are pulled / pushed by that field so strongly that cracks 280 are formed in body 240 a . fig7 is a flowchart of methods of comparison according to embodiments of the present invention . the steps of the flowchart of fig7 are performed first on a known or trusted micro - electronic device and then on an unknown ( or suspect ) micro - electronic device that and essentially identical to the known micro - electronic device ( i . e ., identically designed and within fabrication specification limits ). for the integrated circuit chip specification limits define allowable variations in material and structure and include , for example , the allowable differences in metal line widths and thickness differences in metal and insulating layers . for the package of the ic specification limits define allowable variations in material and structure and include , for example , package dimensions , size of solder bumps , positions and size of wire bonds , widths and thickness of land in modules . in step 300 an ic ( or integrated circuit is placed in the mri chamber . in step 305 the high gauss magnetic field is turned on . in one example , the high gauss magnetic field has a field strength of between about a 0 . 5 tesla and about 10 tesla and is applied in the z direction . the method can know follow one of two mutually exclusive paths . the first is the path ( 1 ) through steps 310 , 315 , 320 , 325 and 330 to step 335 . the second path ( 2 ) is through steps 315 a , 320 a and 325 a to step 335 . the unknown micro - electronic device advantageously follows the same path and is subjected to the same nmr conditions as the known micro - electronic device . in step 310 , a direction ( x , y or z ) is selected . in step 315 , a gradient magnetic field of , for example , 1 tesla is applied in the selected direction and in step 320 a rf pulse is directed to the ic . in step 325 the emitted ( returned ) rf signal is detected and information describing the return rf signal ( which is also a pulse ) is stored . the duration and time of reception of the return rf signal will vary . in step 330 , if another direction of the three possible directions ( x , y , z ) is selected and the method loops back to step 310 . this loop will repeat three times , once for each direction . in path ( 2 ) in step 315 a , gradient magnetic fields are applied in the x , y , and z directions simultaneously . each gradient field may have a same or a different field strength of between about 0 . 5 tesla and about 10 tesla . steps 320 a and 325 a are similar to steps 320 and 325 respectively . in steps 315 and 315 a , optional test conditions ( i . e ., voltage bias , analog signal , digital data pattern or combinations thereof ) may be applied to the micro - electronic device . in a first example , optional test conditions are applied only during step 315 ( or 315 a ). in a second example , optional test conditions are applied the only during steps 315 and 320 ( or 315 a and 320 a ). in a third example , the optional test conditions are applied the only during steps 305 , 310 , 315 and 320 ( or 305 , 315 a and 320 a ). in a fourth example , the optional test conditions are applied during steps 305 , 310 , 315 , 320 and 325 ( or 305 , 315 a , 320 a and 325 a ). in step 335 , the ic is removed from the mri chamber and it is determined whether the micro - electronic device was a known micro - electronic device or an unknown micro - electronic device . if the micro - electronic device was a known micro - electronic device then in step 340 analyses are performed and the analyses data is stored . if the micro - electronic device was an unknown micro - electronic device then in step 345 analyses are performed and the analyses is compared to analyses previously stored from a known micro - electronic device mri \u201c images \u201d under the same mri ( and bias / pattern ) conditions . if the compare is within predefined limits the unknown micro - electronic device is validated . it does not matter whether the known or unknown micro - electronic device is run first , but the comparison cannot be performed until both known or unknown micro - electronic devices have been run and the respective data analyzed . fig8 a and 8b illustrate a first method of data analysis for comparison according to embodiments of the present invention . fig8 a is a plot of the transverse component of the returned rf signal for a known micro - electronic device 350 and an unknown micro - electronic device 355 as relative rf strength versus time . direct comparison or statistical analysis of the difference between curves 350 and 355 is performed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig8 a , the difference between curve 355 relative to curve 350 is shown as a positive time shift . the shift may be negative . fig8 b is a plot of the longitudinal component of the returned rf signal for a known micro - electronic device 360 and an unknown micro - electronic device 365 as relative rf strength versus time . direct comparison or statistical analysis of the difference between curves 360 and 365 is performed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig8 b , the difference of curve 365 relative to curve 360 is shown as a negative time shift . the shift may be positive . fig9 a and 9b illustrate a second method of data analysis for comparison according to embodiments of the present invention . fig9 a is a plot of a fast fourier transform of the returned rf signal for a known ic ( or integrated circuit ) as rf amplitude versus rf frequency . element 370 is the returned signal and elements 370 a and 370 b are harmonic artifacts of element 370 . fig9 b is a plot of a fast fourier transform of the returned rf signal for an unknown ic ( or integrated circuit ) as rf amplitude versus rf frequency . element 375 is the returned signal and elements 375 a and 3750 b are harmonic artifacts of element 370 . element 370 and 375 are statistically compared in terms of amplitude of a given frequency range to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . in fig9 a and 9b , the amplitude of the returned rf signal has been transferred from a time domain to a frequency domain . fig1 a , 10 b and 10 c illustrate a third method of data analysis for comparison according to embodiments of the present invention . fig1 a is a plot of the returned rf signal as relative rf signal strength versus time for a known micro - electronic device . rf signal 380 has a fixed amplitude between times a and b measured from when the rf pulse signal terminated ( or other convenient time reference ). an unknown micro - electronic device may exhibit a signal that is offset ion amplitude , phase , frequency or combinations thereof . two simple examples are given in fig1 b and 10c . fig1 b is a plot of the returned rf signal as relative rf signal strength versus time for an unknown micro - electronic device having only a frequency shift . rf signal 385 a has fixed amplitude between times a \u2032 and b \u2032 measured from the same reference as a and b of fig1 a . the shift between a and a \u2032 and b and b \u2032 is statistically to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . fig1 c is a plot of the returned rf signal as relative rf signal strength versus time for an unknown micro - electronic device having only an amplitude shift . rf signal 385 b has a fixed amplitude between times a and b measured from the same reference as a and b of fig1 a . the change in amplitude between curve 380 of fig1 a and curve 385 b of fig1 c is statistically to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . it will be appreciated that a complete statistical analysis would compare time , amplitude , frequency and phase differences of the curves described in fig1 a , 10 b and 10 c . fig1 a , 11 b and 11 c illustrate a fourth method of data analysis for comparison according to embodiments of the present invention . fig1 a is a plot of the returned rf signal xy space ( at a selected z ) as a function of frequency for a known micro - electronic device . as such fig1 a is closer to what is may be considered an \u201c image .\u201d fig1 b is a plot of the returned rf signal xy space ( at the selected z ) as a function of frequency for an unknown micro - electronic device . fig1 c is a plot of the delta between the plot of fig1 a and that of fig1 b . the amount of structure ( and optionally the position of the structures ) is statistically analyzed to determine if the unknown micro - electronic device is significantly different from the unknown micro - electronic device . fig1 c is fig1 a \u201c subtracted \u201d from fig1 a . thus the embodiments of the present invention allow for relatively quick and inexpensive validation of integrated circuit chips . the description of the embodiments of the present invention is given above for the understanding of the present invention . it will be understood that the invention is not limited to the particular embodiments described herein , but is capable of various modifications , rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention . therefore , it is intended that the following claims cover all such modifications and changes as fall within the true spirit and scope of the invention ."}
|
Does the category match the content of the patent?
| 0.25 |
d2a533e20b88cf12a66089384d35063da6ac708b8421ee9a63cae12490153828
| 0.063477 | 0.417969 | 0.189453 | 0.212891 | 0.229492 | 0.19043 |
null |
{"patent": "with reference to the drawings , an embodiment of the present invention is described . the powder feeder comprises a final - stage powder container ( 1 ), a spreading feeder ( 2 ), and a multi - joint robot arm ( 3 ). the final - stage powder container ( 1 ) is equipped with a powder inlet line 13 and is mounted on the end of the revolving arm ( 12 ) held on the top of a column ( 11 ). final - stage powder container ( 1 ) is mounted at a position close to the base end of the tip arm ( 12a ). this revolving arm and tip arm are adequately driven by a driving device ( not shown ). here , the column ( 11 ) is taller than the height of an operator ( a ), thereby the revolving arm ( 12 ) and the final - stage powder container ( 1 ) are arranged above the working space of the mold . thus , the final - stage powder container ( 1 ) can be arranged on the operator side ( a ), for example , overhead of an operator . in fig2 two final - stage powder containers ( 1 )( 1 ) are arranged on the operator side . in the drawing , ( b ) indicates the side opposite an operator . a final - stage powder container ( 1 ) is provided with a meter such as a load cell platform scale ( 14 ) to weigh the spread quantity of casting powder . particularly , use of the &# 34 ; loss - in - weight &# 34 ; system permits recording of accurate spreading quantity and higher accuracy of control by a main computer of the continuous casting unit . the robot arm ( 3 ) comprises a base end arm ( 31 ), an intermediate arm ( 32 ), and a tip hand ( 33 ) movably connected by a first joint ( 34 ) and a second joint ( 35 ). the base end arm ( 31 ) is attached to the end of the tip arm ( 12a ) of the revolving arm . to the tip hand ( 33 ) is mounted a spreading feeder ( 2 ). the spreading feeder ( 2 ) rotates a spring in a tube by a motor ( 21 ) mounted at the tube &# 39 ; s base end to spread the casting powder from its tip on to the molten steel surface of the mold . in place of such mechanical means , other transfer means such as pneumatic transfer means can also be used . in the drawings , ( 5 ) represents a tundish car with a tundish ( 6 ) mounted thereon , a strand nozzle ( 6a ) ( 7 ) a ladle ( 7 ). to the tip on the intermediate arm ( 32 ) of the multi - joint robot arm , a sensor ( 9 ) for monitoring the spreading conditions of the casting powder is arranged . this sensor can be an infrared sensor or thermal sensor , and is used for detecting the exposed molten steel ( hot spot ) in the mold ( 4 ). based on the detection by the sensor , the multi - joint robot arm ( 3 ) is moved under automatic control of a computer so as to move the tip of the spreading casting powder feeder ( 2 ) to the hot spot for spreading . it is also possible to move the robot arm ( 3 ) according to a predetermined program for spreading , not using such a sensor . the base end of the spreading feeder ( 2 ) and the discharge port in the final - stage powder container ( 1 ) are connected with a flexible transfer path ( 8 ). the flexible transfer path comprises a transfer path having the flexibility to follow the movement of the robot arm , such as flexible pipe . therefore , non - flexible pipe may be used only in sections of the path where flexibility is not required . the flexible transfer path ( 8 ) is arranged from the discharge port of the powder container ( 1 ) above the tip arm ( 12a ) and along the robot arm ( 3 ) to the spreading feeder ( 2 ). however , to simplify the drawing , the illustration of the part along the robot arm ( 3 ) is omitted . it is possible to equip a feeding device in the flexible transfer path ( 8 ) along with the tip arm ( 12a ). the casting powder is transferred by said feeding device and dropped by gravity from the final - stage container ( 1 ) to the spreading feeder ( 2 ). this gravity drop is based on the energy saving concept using the height difference between the final - stage powder container ( 1 ) arranged in a high position and the spreading feeder ( 2 ) placed in low position , but forced transfer means can be added . ( i ) when the tundish car ( 5 ) stops at the position above the mold ( 4 ), the revolving arm ( 12 ) swings to move the feeder from the stand - by position ( i ) to the feed position ( ii ). ( ii ) the tip arm ( 12a ) of the revolving arm turns to face the powder feeder toward the mold . ( iii ) the robot arm ( 3 ) moves the tip of the spreading feeder ( 2 ) above a hot spot to spread the casting powder from its tip . in fig2 the area designated ( b ) shows the spreading area . ( iv ) in replacing the tundish , the powder feeder is returned to the stand - by position by the reverse operation of steps ( i ) and ( ii ) mentioned above . the flexibility of the multi - joint robot arm ( 3 ) removes the dead angle above the mold surface , and prevents contact between the spreading feeder ( 2 ) and the tundish strand nozzle ( 6a ), when the powder feeder is moved . the present invention is not limited to the above embodiment . if there is a high position such as a deck near the workshop , the final - stage container can be mounted on the deck and the column and revolving arm can be eliminated . the final - stage powder container may be also hung from a high position . since the final - stage powder container is arranged in a position higher than the working space on the powder feeder according to the invention , provision of this feeder in the operator side does not interfere with the work of an operator . since the operator side is free of the scattered dust and molten steel of the side opposite the operator , the following effects are obtained : ( i ) sharp decrease in trouble due to scattered dust and molten steel ,; ( iii ) improved work environment permitting use of precision instruments such as the robot arm ,; ( v ) the smaller distance to the control board or the operation board making the anti - nozzle provision for cpu wiring easier . in spreading the casting powder on the mold , the flexibility of the robot arm removes the dead angle on the spreading surface of the mold , and the powder is uniformly spreaded all over the mold surface . and the precision instruments and control equipment used in the robot arm can continue good operation in the good work environment as described in the item ( iii ) mentioned above . the provision of the spreading conditions monitor sensor on the robot arm , makes complete automation of hot spot detection and spreading possible by computer control of the spreading . this promotes labor saving , stabilizes the continuous casting and improves quality . this invention can be used in full automation of continuous casting . in further progress of continuous casting of highgrade steel , a powder feeder in continuous casting has been provided which can cope with feed automation of highgrade steel billet size casting powder use .", "category": "Performing Operations; Transporting"}
|
{"patent": "with reference to the drawings , an embodiment of the present invention is described . the powder feeder comprises a final - stage powder container ( 1 ), a spreading feeder ( 2 ), and a multi - joint robot arm ( 3 ). the final - stage powder container ( 1 ) is equipped with a powder inlet line 13 and is mounted on the end of the revolving arm ( 12 ) held on the top of a column ( 11 ). final - stage powder container ( 1 ) is mounted at a position close to the base end of the tip arm ( 12a ). this revolving arm and tip arm are adequately driven by a driving device ( not shown ). here , the column ( 11 ) is taller than the height of an operator ( a ), thereby the revolving arm ( 12 ) and the final - stage powder container ( 1 ) are arranged above the working space of the mold . thus , the final - stage powder container ( 1 ) can be arranged on the operator side ( a ), for example , overhead of an operator . in fig2 two final - stage powder containers ( 1 )( 1 ) are arranged on the operator side . in the drawing , ( b ) indicates the side opposite an operator . a final - stage powder container ( 1 ) is provided with a meter such as a load cell platform scale ( 14 ) to weigh the spread quantity of casting powder . particularly , use of the &# 34 ; loss - in - weight &# 34 ; system permits recording of accurate spreading quantity and higher accuracy of control by a main computer of the continuous casting unit . the robot arm ( 3 ) comprises a base end arm ( 31 ), an intermediate arm ( 32 ), and a tip hand ( 33 ) movably connected by a first joint ( 34 ) and a second joint ( 35 ). the base end arm ( 31 ) is attached to the end of the tip arm ( 12a ) of the revolving arm . to the tip hand ( 33 ) is mounted a spreading feeder ( 2 ). the spreading feeder ( 2 ) rotates a spring in a tube by a motor ( 21 ) mounted at the tube &# 39 ; s base end to spread the casting powder from its tip on to the molten steel surface of the mold . in place of such mechanical means , other transfer means such as pneumatic transfer means can also be used . in the drawings , ( 5 ) represents a tundish car with a tundish ( 6 ) mounted thereon , a strand nozzle ( 6a ) ( 7 ) a ladle ( 7 ). to the tip on the intermediate arm ( 32 ) of the multi - joint robot arm , a sensor ( 9 ) for monitoring the spreading conditions of the casting powder is arranged . this sensor can be an infrared sensor or thermal sensor , and is used for detecting the exposed molten steel ( hot spot ) in the mold ( 4 ). based on the detection by the sensor , the multi - joint robot arm ( 3 ) is moved under automatic control of a computer so as to move the tip of the spreading casting powder feeder ( 2 ) to the hot spot for spreading . it is also possible to move the robot arm ( 3 ) according to a predetermined program for spreading , not using such a sensor . the base end of the spreading feeder ( 2 ) and the discharge port in the final - stage powder container ( 1 ) are connected with a flexible transfer path ( 8 ). the flexible transfer path comprises a transfer path having the flexibility to follow the movement of the robot arm , such as flexible pipe . therefore , non - flexible pipe may be used only in sections of the path where flexibility is not required . the flexible transfer path ( 8 ) is arranged from the discharge port of the powder container ( 1 ) above the tip arm ( 12a ) and along the robot arm ( 3 ) to the spreading feeder ( 2 ). however , to simplify the drawing , the illustration of the part along the robot arm ( 3 ) is omitted . it is possible to equip a feeding device in the flexible transfer path ( 8 ) along with the tip arm ( 12a ). the casting powder is transferred by said feeding device and dropped by gravity from the final - stage container ( 1 ) to the spreading feeder ( 2 ). this gravity drop is based on the energy saving concept using the height difference between the final - stage powder container ( 1 ) arranged in a high position and the spreading feeder ( 2 ) placed in low position , but forced transfer means can be added . ( i ) when the tundish car ( 5 ) stops at the position above the mold ( 4 ), the revolving arm ( 12 ) swings to move the feeder from the stand - by position ( i ) to the feed position ( ii ). ( ii ) the tip arm ( 12a ) of the revolving arm turns to face the powder feeder toward the mold . ( iii ) the robot arm ( 3 ) moves the tip of the spreading feeder ( 2 ) above a hot spot to spread the casting powder from its tip . in fig2 the area designated ( b ) shows the spreading area . ( iv ) in replacing the tundish , the powder feeder is returned to the stand - by position by the reverse operation of steps ( i ) and ( ii ) mentioned above . the flexibility of the multi - joint robot arm ( 3 ) removes the dead angle above the mold surface , and prevents contact between the spreading feeder ( 2 ) and the tundish strand nozzle ( 6a ), when the powder feeder is moved . the present invention is not limited to the above embodiment . if there is a high position such as a deck near the workshop , the final - stage container can be mounted on the deck and the column and revolving arm can be eliminated . the final - stage powder container may be also hung from a high position . since the final - stage powder container is arranged in a position higher than the working space on the powder feeder according to the invention , provision of this feeder in the operator side does not interfere with the work of an operator . since the operator side is free of the scattered dust and molten steel of the side opposite the operator , the following effects are obtained : ( i ) sharp decrease in trouble due to scattered dust and molten steel ,; ( iii ) improved work environment permitting use of precision instruments such as the robot arm ,; ( v ) the smaller distance to the control board or the operation board making the anti - nozzle provision for cpu wiring easier . in spreading the casting powder on the mold , the flexibility of the robot arm removes the dead angle on the spreading surface of the mold , and the powder is uniformly spreaded all over the mold surface . and the precision instruments and control equipment used in the robot arm can continue good operation in the good work environment as described in the item ( iii ) mentioned above . the provision of the spreading conditions monitor sensor on the robot arm , makes complete automation of hot spot detection and spreading possible by computer control of the spreading . this promotes labor saving , stabilizes the continuous casting and improves quality . this invention can be used in full automation of continuous casting . in further progress of continuous casting of highgrade steel , a powder feeder in continuous casting has been provided which can cope with feed automation of highgrade steel billet size casting powder use .", "category": "Human Necessities"}
|
Is the patent correctly categorized?
| 0.25 |
49d27878671b44cf2401f85d19ab2921adb1331aa61116bc349d37d60692ad6f
| 0.016357 | 0.00383 | 0.235352 | 0.007355 | 0.204102 | 0.007355 |
null |
{"patent": "with reference to the drawings , an embodiment of the present invention is described . the powder feeder comprises a final - stage powder container ( 1 ), a spreading feeder ( 2 ), and a multi - joint robot arm ( 3 ). the final - stage powder container ( 1 ) is equipped with a powder inlet line 13 and is mounted on the end of the revolving arm ( 12 ) held on the top of a column ( 11 ). final - stage powder container ( 1 ) is mounted at a position close to the base end of the tip arm ( 12a ). this revolving arm and tip arm are adequately driven by a driving device ( not shown ). here , the column ( 11 ) is taller than the height of an operator ( a ), thereby the revolving arm ( 12 ) and the final - stage powder container ( 1 ) are arranged above the working space of the mold . thus , the final - stage powder container ( 1 ) can be arranged on the operator side ( a ), for example , overhead of an operator . in fig2 two final - stage powder containers ( 1 )( 1 ) are arranged on the operator side . in the drawing , ( b ) indicates the side opposite an operator . a final - stage powder container ( 1 ) is provided with a meter such as a load cell platform scale ( 14 ) to weigh the spread quantity of casting powder . particularly , use of the &# 34 ; loss - in - weight &# 34 ; system permits recording of accurate spreading quantity and higher accuracy of control by a main computer of the continuous casting unit . the robot arm ( 3 ) comprises a base end arm ( 31 ), an intermediate arm ( 32 ), and a tip hand ( 33 ) movably connected by a first joint ( 34 ) and a second joint ( 35 ). the base end arm ( 31 ) is attached to the end of the tip arm ( 12a ) of the revolving arm . to the tip hand ( 33 ) is mounted a spreading feeder ( 2 ). the spreading feeder ( 2 ) rotates a spring in a tube by a motor ( 21 ) mounted at the tube &# 39 ; s base end to spread the casting powder from its tip on to the molten steel surface of the mold . in place of such mechanical means , other transfer means such as pneumatic transfer means can also be used . in the drawings , ( 5 ) represents a tundish car with a tundish ( 6 ) mounted thereon , a strand nozzle ( 6a ) ( 7 ) a ladle ( 7 ). to the tip on the intermediate arm ( 32 ) of the multi - joint robot arm , a sensor ( 9 ) for monitoring the spreading conditions of the casting powder is arranged . this sensor can be an infrared sensor or thermal sensor , and is used for detecting the exposed molten steel ( hot spot ) in the mold ( 4 ). based on the detection by the sensor , the multi - joint robot arm ( 3 ) is moved under automatic control of a computer so as to move the tip of the spreading casting powder feeder ( 2 ) to the hot spot for spreading . it is also possible to move the robot arm ( 3 ) according to a predetermined program for spreading , not using such a sensor . the base end of the spreading feeder ( 2 ) and the discharge port in the final - stage powder container ( 1 ) are connected with a flexible transfer path ( 8 ). the flexible transfer path comprises a transfer path having the flexibility to follow the movement of the robot arm , such as flexible pipe . therefore , non - flexible pipe may be used only in sections of the path where flexibility is not required . the flexible transfer path ( 8 ) is arranged from the discharge port of the powder container ( 1 ) above the tip arm ( 12a ) and along the robot arm ( 3 ) to the spreading feeder ( 2 ). however , to simplify the drawing , the illustration of the part along the robot arm ( 3 ) is omitted . it is possible to equip a feeding device in the flexible transfer path ( 8 ) along with the tip arm ( 12a ). the casting powder is transferred by said feeding device and dropped by gravity from the final - stage container ( 1 ) to the spreading feeder ( 2 ). this gravity drop is based on the energy saving concept using the height difference between the final - stage powder container ( 1 ) arranged in a high position and the spreading feeder ( 2 ) placed in low position , but forced transfer means can be added . ( i ) when the tundish car ( 5 ) stops at the position above the mold ( 4 ), the revolving arm ( 12 ) swings to move the feeder from the stand - by position ( i ) to the feed position ( ii ). ( ii ) the tip arm ( 12a ) of the revolving arm turns to face the powder feeder toward the mold . ( iii ) the robot arm ( 3 ) moves the tip of the spreading feeder ( 2 ) above a hot spot to spread the casting powder from its tip . in fig2 the area designated ( b ) shows the spreading area . ( iv ) in replacing the tundish , the powder feeder is returned to the stand - by position by the reverse operation of steps ( i ) and ( ii ) mentioned above . the flexibility of the multi - joint robot arm ( 3 ) removes the dead angle above the mold surface , and prevents contact between the spreading feeder ( 2 ) and the tundish strand nozzle ( 6a ), when the powder feeder is moved . the present invention is not limited to the above embodiment . if there is a high position such as a deck near the workshop , the final - stage container can be mounted on the deck and the column and revolving arm can be eliminated . the final - stage powder container may be also hung from a high position . since the final - stage powder container is arranged in a position higher than the working space on the powder feeder according to the invention , provision of this feeder in the operator side does not interfere with the work of an operator . since the operator side is free of the scattered dust and molten steel of the side opposite the operator , the following effects are obtained : ( i ) sharp decrease in trouble due to scattered dust and molten steel ,; ( iii ) improved work environment permitting use of precision instruments such as the robot arm ,; ( v ) the smaller distance to the control board or the operation board making the anti - nozzle provision for cpu wiring easier . in spreading the casting powder on the mold , the flexibility of the robot arm removes the dead angle on the spreading surface of the mold , and the powder is uniformly spreaded all over the mold surface . and the precision instruments and control equipment used in the robot arm can continue good operation in the good work environment as described in the item ( iii ) mentioned above . the provision of the spreading conditions monitor sensor on the robot arm , makes complete automation of hot spot detection and spreading possible by computer control of the spreading . this promotes labor saving , stabilizes the continuous casting and improves quality . this invention can be used in full automation of continuous casting . in further progress of continuous casting of highgrade steel , a powder feeder in continuous casting has been provided which can cope with feed automation of highgrade steel billet size casting powder use .", "category": "Performing Operations; Transporting"}
|
{"patent": "with reference to the drawings , an embodiment of the present invention is described . the powder feeder comprises a final - stage powder container ( 1 ), a spreading feeder ( 2 ), and a multi - joint robot arm ( 3 ). the final - stage powder container ( 1 ) is equipped with a powder inlet line 13 and is mounted on the end of the revolving arm ( 12 ) held on the top of a column ( 11 ). final - stage powder container ( 1 ) is mounted at a position close to the base end of the tip arm ( 12a ). this revolving arm and tip arm are adequately driven by a driving device ( not shown ). here , the column ( 11 ) is taller than the height of an operator ( a ), thereby the revolving arm ( 12 ) and the final - stage powder container ( 1 ) are arranged above the working space of the mold . thus , the final - stage powder container ( 1 ) can be arranged on the operator side ( a ), for example , overhead of an operator . in fig2 two final - stage powder containers ( 1 )( 1 ) are arranged on the operator side . in the drawing , ( b ) indicates the side opposite an operator . a final - stage powder container ( 1 ) is provided with a meter such as a load cell platform scale ( 14 ) to weigh the spread quantity of casting powder . particularly , use of the &# 34 ; loss - in - weight &# 34 ; system permits recording of accurate spreading quantity and higher accuracy of control by a main computer of the continuous casting unit . the robot arm ( 3 ) comprises a base end arm ( 31 ), an intermediate arm ( 32 ), and a tip hand ( 33 ) movably connected by a first joint ( 34 ) and a second joint ( 35 ). the base end arm ( 31 ) is attached to the end of the tip arm ( 12a ) of the revolving arm . to the tip hand ( 33 ) is mounted a spreading feeder ( 2 ). the spreading feeder ( 2 ) rotates a spring in a tube by a motor ( 21 ) mounted at the tube &# 39 ; s base end to spread the casting powder from its tip on to the molten steel surface of the mold . in place of such mechanical means , other transfer means such as pneumatic transfer means can also be used . in the drawings , ( 5 ) represents a tundish car with a tundish ( 6 ) mounted thereon , a strand nozzle ( 6a ) ( 7 ) a ladle ( 7 ). to the tip on the intermediate arm ( 32 ) of the multi - joint robot arm , a sensor ( 9 ) for monitoring the spreading conditions of the casting powder is arranged . this sensor can be an infrared sensor or thermal sensor , and is used for detecting the exposed molten steel ( hot spot ) in the mold ( 4 ). based on the detection by the sensor , the multi - joint robot arm ( 3 ) is moved under automatic control of a computer so as to move the tip of the spreading casting powder feeder ( 2 ) to the hot spot for spreading . it is also possible to move the robot arm ( 3 ) according to a predetermined program for spreading , not using such a sensor . the base end of the spreading feeder ( 2 ) and the discharge port in the final - stage powder container ( 1 ) are connected with a flexible transfer path ( 8 ). the flexible transfer path comprises a transfer path having the flexibility to follow the movement of the robot arm , such as flexible pipe . therefore , non - flexible pipe may be used only in sections of the path where flexibility is not required . the flexible transfer path ( 8 ) is arranged from the discharge port of the powder container ( 1 ) above the tip arm ( 12a ) and along the robot arm ( 3 ) to the spreading feeder ( 2 ). however , to simplify the drawing , the illustration of the part along the robot arm ( 3 ) is omitted . it is possible to equip a feeding device in the flexible transfer path ( 8 ) along with the tip arm ( 12a ). the casting powder is transferred by said feeding device and dropped by gravity from the final - stage container ( 1 ) to the spreading feeder ( 2 ). this gravity drop is based on the energy saving concept using the height difference between the final - stage powder container ( 1 ) arranged in a high position and the spreading feeder ( 2 ) placed in low position , but forced transfer means can be added . ( i ) when the tundish car ( 5 ) stops at the position above the mold ( 4 ), the revolving arm ( 12 ) swings to move the feeder from the stand - by position ( i ) to the feed position ( ii ). ( ii ) the tip arm ( 12a ) of the revolving arm turns to face the powder feeder toward the mold . ( iii ) the robot arm ( 3 ) moves the tip of the spreading feeder ( 2 ) above a hot spot to spread the casting powder from its tip . in fig2 the area designated ( b ) shows the spreading area . ( iv ) in replacing the tundish , the powder feeder is returned to the stand - by position by the reverse operation of steps ( i ) and ( ii ) mentioned above . the flexibility of the multi - joint robot arm ( 3 ) removes the dead angle above the mold surface , and prevents contact between the spreading feeder ( 2 ) and the tundish strand nozzle ( 6a ), when the powder feeder is moved . the present invention is not limited to the above embodiment . if there is a high position such as a deck near the workshop , the final - stage container can be mounted on the deck and the column and revolving arm can be eliminated . the final - stage powder container may be also hung from a high position . since the final - stage powder container is arranged in a position higher than the working space on the powder feeder according to the invention , provision of this feeder in the operator side does not interfere with the work of an operator . since the operator side is free of the scattered dust and molten steel of the side opposite the operator , the following effects are obtained : ( i ) sharp decrease in trouble due to scattered dust and molten steel ,; ( iii ) improved work environment permitting use of precision instruments such as the robot arm ,; ( v ) the smaller distance to the control board or the operation board making the anti - nozzle provision for cpu wiring easier . in spreading the casting powder on the mold , the flexibility of the robot arm removes the dead angle on the spreading surface of the mold , and the powder is uniformly spreaded all over the mold surface . and the precision instruments and control equipment used in the robot arm can continue good operation in the good work environment as described in the item ( iii ) mentioned above . the provision of the spreading conditions monitor sensor on the robot arm , makes complete automation of hot spot detection and spreading possible by computer control of the spreading . this promotes labor saving , stabilizes the continuous casting and improves quality . this invention can be used in full automation of continuous casting . in further progress of continuous casting of highgrade steel , a powder feeder in continuous casting has been provided which can cope with feed automation of highgrade steel billet size casting powder use .", "category": "Chemistry; Metallurgy"}
|
Does the category match the content of the patent?
| 0.25 |
49d27878671b44cf2401f85d19ab2921adb1331aa61116bc349d37d60692ad6f
| 0.170898 | 0.043457 | 0.578125 | 0.108398 | 0.239258 | 0.126953 |
null |
{"category": "Performing Operations; Transporting", "patent": "with reference to the drawings , an embodiment of the present invention is described . the powder feeder comprises a final - stage powder container ( 1 ), a spreading feeder ( 2 ), and a multi - joint robot arm ( 3 ). the final - stage powder container ( 1 ) is equipped with a powder inlet line 13 and is mounted on the end of the revolving arm ( 12 ) held on the top of a column ( 11 ). final - stage powder container ( 1 ) is mounted at a position close to the base end of the tip arm ( 12a ). this revolving arm and tip arm are adequately driven by a driving device ( not shown ). here , the column ( 11 ) is taller than the height of an operator ( a ), thereby the revolving arm ( 12 ) and the final - stage powder container ( 1 ) are arranged above the working space of the mold . thus , the final - stage powder container ( 1 ) can be arranged on the operator side ( a ), for example , overhead of an operator . in fig2 two final - stage powder containers ( 1 )( 1 ) are arranged on the operator side . in the drawing , ( b ) indicates the side opposite an operator . a final - stage powder container ( 1 ) is provided with a meter such as a load cell platform scale ( 14 ) to weigh the spread quantity of casting powder . particularly , use of the &# 34 ; loss - in - weight &# 34 ; system permits recording of accurate spreading quantity and higher accuracy of control by a main computer of the continuous casting unit . the robot arm ( 3 ) comprises a base end arm ( 31 ), an intermediate arm ( 32 ), and a tip hand ( 33 ) movably connected by a first joint ( 34 ) and a second joint ( 35 ). the base end arm ( 31 ) is attached to the end of the tip arm ( 12a ) of the revolving arm . to the tip hand ( 33 ) is mounted a spreading feeder ( 2 ). the spreading feeder ( 2 ) rotates a spring in a tube by a motor ( 21 ) mounted at the tube &# 39 ; s base end to spread the casting powder from its tip on to the molten steel surface of the mold . in place of such mechanical means , other transfer means such as pneumatic transfer means can also be used . in the drawings , ( 5 ) represents a tundish car with a tundish ( 6 ) mounted thereon , a strand nozzle ( 6a ) ( 7 ) a ladle ( 7 ). to the tip on the intermediate arm ( 32 ) of the multi - joint robot arm , a sensor ( 9 ) for monitoring the spreading conditions of the casting powder is arranged . this sensor can be an infrared sensor or thermal sensor , and is used for detecting the exposed molten steel ( hot spot ) in the mold ( 4 ). based on the detection by the sensor , the multi - joint robot arm ( 3 ) is moved under automatic control of a computer so as to move the tip of the spreading casting powder feeder ( 2 ) to the hot spot for spreading . it is also possible to move the robot arm ( 3 ) according to a predetermined program for spreading , not using such a sensor . the base end of the spreading feeder ( 2 ) and the discharge port in the final - stage powder container ( 1 ) are connected with a flexible transfer path ( 8 ). the flexible transfer path comprises a transfer path having the flexibility to follow the movement of the robot arm , such as flexible pipe . therefore , non - flexible pipe may be used only in sections of the path where flexibility is not required . the flexible transfer path ( 8 ) is arranged from the discharge port of the powder container ( 1 ) above the tip arm ( 12a ) and along the robot arm ( 3 ) to the spreading feeder ( 2 ). however , to simplify the drawing , the illustration of the part along the robot arm ( 3 ) is omitted . it is possible to equip a feeding device in the flexible transfer path ( 8 ) along with the tip arm ( 12a ). the casting powder is transferred by said feeding device and dropped by gravity from the final - stage container ( 1 ) to the spreading feeder ( 2 ). this gravity drop is based on the energy saving concept using the height difference between the final - stage powder container ( 1 ) arranged in a high position and the spreading feeder ( 2 ) placed in low position , but forced transfer means can be added . ( i ) when the tundish car ( 5 ) stops at the position above the mold ( 4 ), the revolving arm ( 12 ) swings to move the feeder from the stand - by position ( i ) to the feed position ( ii ). ( ii ) the tip arm ( 12a ) of the revolving arm turns to face the powder feeder toward the mold . ( iii ) the robot arm ( 3 ) moves the tip of the spreading feeder ( 2 ) above a hot spot to spread the casting powder from its tip . in fig2 the area designated ( b ) shows the spreading area . ( iv ) in replacing the tundish , the powder feeder is returned to the stand - by position by the reverse operation of steps ( i ) and ( ii ) mentioned above . the flexibility of the multi - joint robot arm ( 3 ) removes the dead angle above the mold surface , and prevents contact between the spreading feeder ( 2 ) and the tundish strand nozzle ( 6a ), when the powder feeder is moved . the present invention is not limited to the above embodiment . if there is a high position such as a deck near the workshop , the final - stage container can be mounted on the deck and the column and revolving arm can be eliminated . the final - stage powder container may be also hung from a high position . since the final - stage powder container is arranged in a position higher than the working space on the powder feeder according to the invention , provision of this feeder in the operator side does not interfere with the work of an operator . since the operator side is free of the scattered dust and molten steel of the side opposite the operator , the following effects are obtained : ( i ) sharp decrease in trouble due to scattered dust and molten steel ,; ( iii ) improved work environment permitting use of precision instruments such as the robot arm ,; ( v ) the smaller distance to the control board or the operation board making the anti - nozzle provision for cpu wiring easier . in spreading the casting powder on the mold , the flexibility of the robot arm removes the dead angle on the spreading surface of the mold , and the powder is uniformly spreaded all over the mold surface . and the precision instruments and control equipment used in the robot arm can continue good operation in the good work environment as described in the item ( iii ) mentioned above . the provision of the spreading conditions monitor sensor on the robot arm , makes complete automation of hot spot detection and spreading possible by computer control of the spreading . this promotes labor saving , stabilizes the continuous casting and improves quality . this invention can be used in full automation of continuous casting . in further progress of continuous casting of highgrade steel , a powder feeder in continuous casting has been provided which can cope with feed automation of highgrade steel billet size casting powder use ."}
|
{"patent": "with reference to the drawings , an embodiment of the present invention is described . the powder feeder comprises a final - stage powder container ( 1 ), a spreading feeder ( 2 ), and a multi - joint robot arm ( 3 ). the final - stage powder container ( 1 ) is equipped with a powder inlet line 13 and is mounted on the end of the revolving arm ( 12 ) held on the top of a column ( 11 ). final - stage powder container ( 1 ) is mounted at a position close to the base end of the tip arm ( 12a ). this revolving arm and tip arm are adequately driven by a driving device ( not shown ). here , the column ( 11 ) is taller than the height of an operator ( a ), thereby the revolving arm ( 12 ) and the final - stage powder container ( 1 ) are arranged above the working space of the mold . thus , the final - stage powder container ( 1 ) can be arranged on the operator side ( a ), for example , overhead of an operator . in fig2 two final - stage powder containers ( 1 )( 1 ) are arranged on the operator side . in the drawing , ( b ) indicates the side opposite an operator . a final - stage powder container ( 1 ) is provided with a meter such as a load cell platform scale ( 14 ) to weigh the spread quantity of casting powder . particularly , use of the &# 34 ; loss - in - weight &# 34 ; system permits recording of accurate spreading quantity and higher accuracy of control by a main computer of the continuous casting unit . the robot arm ( 3 ) comprises a base end arm ( 31 ), an intermediate arm ( 32 ), and a tip hand ( 33 ) movably connected by a first joint ( 34 ) and a second joint ( 35 ). the base end arm ( 31 ) is attached to the end of the tip arm ( 12a ) of the revolving arm . to the tip hand ( 33 ) is mounted a spreading feeder ( 2 ). the spreading feeder ( 2 ) rotates a spring in a tube by a motor ( 21 ) mounted at the tube &# 39 ; s base end to spread the casting powder from its tip on to the molten steel surface of the mold . in place of such mechanical means , other transfer means such as pneumatic transfer means can also be used . in the drawings , ( 5 ) represents a tundish car with a tundish ( 6 ) mounted thereon , a strand nozzle ( 6a ) ( 7 ) a ladle ( 7 ). to the tip on the intermediate arm ( 32 ) of the multi - joint robot arm , a sensor ( 9 ) for monitoring the spreading conditions of the casting powder is arranged . this sensor can be an infrared sensor or thermal sensor , and is used for detecting the exposed molten steel ( hot spot ) in the mold ( 4 ). based on the detection by the sensor , the multi - joint robot arm ( 3 ) is moved under automatic control of a computer so as to move the tip of the spreading casting powder feeder ( 2 ) to the hot spot for spreading . it is also possible to move the robot arm ( 3 ) according to a predetermined program for spreading , not using such a sensor . the base end of the spreading feeder ( 2 ) and the discharge port in the final - stage powder container ( 1 ) are connected with a flexible transfer path ( 8 ). the flexible transfer path comprises a transfer path having the flexibility to follow the movement of the robot arm , such as flexible pipe . therefore , non - flexible pipe may be used only in sections of the path where flexibility is not required . the flexible transfer path ( 8 ) is arranged from the discharge port of the powder container ( 1 ) above the tip arm ( 12a ) and along the robot arm ( 3 ) to the spreading feeder ( 2 ). however , to simplify the drawing , the illustration of the part along the robot arm ( 3 ) is omitted . it is possible to equip a feeding device in the flexible transfer path ( 8 ) along with the tip arm ( 12a ). the casting powder is transferred by said feeding device and dropped by gravity from the final - stage container ( 1 ) to the spreading feeder ( 2 ). this gravity drop is based on the energy saving concept using the height difference between the final - stage powder container ( 1 ) arranged in a high position and the spreading feeder ( 2 ) placed in low position , but forced transfer means can be added . ( i ) when the tundish car ( 5 ) stops at the position above the mold ( 4 ), the revolving arm ( 12 ) swings to move the feeder from the stand - by position ( i ) to the feed position ( ii ). ( ii ) the tip arm ( 12a ) of the revolving arm turns to face the powder feeder toward the mold . ( iii ) the robot arm ( 3 ) moves the tip of the spreading feeder ( 2 ) above a hot spot to spread the casting powder from its tip . in fig2 the area designated ( b ) shows the spreading area . ( iv ) in replacing the tundish , the powder feeder is returned to the stand - by position by the reverse operation of steps ( i ) and ( ii ) mentioned above . the flexibility of the multi - joint robot arm ( 3 ) removes the dead angle above the mold surface , and prevents contact between the spreading feeder ( 2 ) and the tundish strand nozzle ( 6a ), when the powder feeder is moved . the present invention is not limited to the above embodiment . if there is a high position such as a deck near the workshop , the final - stage container can be mounted on the deck and the column and revolving arm can be eliminated . the final - stage powder container may be also hung from a high position . since the final - stage powder container is arranged in a position higher than the working space on the powder feeder according to the invention , provision of this feeder in the operator side does not interfere with the work of an operator . since the operator side is free of the scattered dust and molten steel of the side opposite the operator , the following effects are obtained : ( i ) sharp decrease in trouble due to scattered dust and molten steel ,; ( iii ) improved work environment permitting use of precision instruments such as the robot arm ,; ( v ) the smaller distance to the control board or the operation board making the anti - nozzle provision for cpu wiring easier . in spreading the casting powder on the mold , the flexibility of the robot arm removes the dead angle on the spreading surface of the mold , and the powder is uniformly spreaded all over the mold surface . and the precision instruments and control equipment used in the robot arm can continue good operation in the good work environment as described in the item ( iii ) mentioned above . the provision of the spreading conditions monitor sensor on the robot arm , makes complete automation of hot spot detection and spreading possible by computer control of the spreading . this promotes labor saving , stabilizes the continuous casting and improves quality . this invention can be used in full automation of continuous casting . in further progress of continuous casting of highgrade steel , a powder feeder in continuous casting has been provided which can cope with feed automation of highgrade steel billet size casting powder use .", "category": "Textiles; Paper"}
|
Is the patent correctly categorized?
| 0.25 |
49d27878671b44cf2401f85d19ab2921adb1331aa61116bc349d37d60692ad6f
| 0.088867 | 0.000999 | 0.119141 | 0.005737 | 0.667969 | 0.004059 |
null |
{"category": "Performing Operations; Transporting", "patent": "with reference to the drawings , an embodiment of the present invention is described . the powder feeder comprises a final - stage powder container ( 1 ), a spreading feeder ( 2 ), and a multi - joint robot arm ( 3 ). the final - stage powder container ( 1 ) is equipped with a powder inlet line 13 and is mounted on the end of the revolving arm ( 12 ) held on the top of a column ( 11 ). final - stage powder container ( 1 ) is mounted at a position close to the base end of the tip arm ( 12a ). this revolving arm and tip arm are adequately driven by a driving device ( not shown ). here , the column ( 11 ) is taller than the height of an operator ( a ), thereby the revolving arm ( 12 ) and the final - stage powder container ( 1 ) are arranged above the working space of the mold . thus , the final - stage powder container ( 1 ) can be arranged on the operator side ( a ), for example , overhead of an operator . in fig2 two final - stage powder containers ( 1 )( 1 ) are arranged on the operator side . in the drawing , ( b ) indicates the side opposite an operator . a final - stage powder container ( 1 ) is provided with a meter such as a load cell platform scale ( 14 ) to weigh the spread quantity of casting powder . particularly , use of the &# 34 ; loss - in - weight &# 34 ; system permits recording of accurate spreading quantity and higher accuracy of control by a main computer of the continuous casting unit . the robot arm ( 3 ) comprises a base end arm ( 31 ), an intermediate arm ( 32 ), and a tip hand ( 33 ) movably connected by a first joint ( 34 ) and a second joint ( 35 ). the base end arm ( 31 ) is attached to the end of the tip arm ( 12a ) of the revolving arm . to the tip hand ( 33 ) is mounted a spreading feeder ( 2 ). the spreading feeder ( 2 ) rotates a spring in a tube by a motor ( 21 ) mounted at the tube &# 39 ; s base end to spread the casting powder from its tip on to the molten steel surface of the mold . in place of such mechanical means , other transfer means such as pneumatic transfer means can also be used . in the drawings , ( 5 ) represents a tundish car with a tundish ( 6 ) mounted thereon , a strand nozzle ( 6a ) ( 7 ) a ladle ( 7 ). to the tip on the intermediate arm ( 32 ) of the multi - joint robot arm , a sensor ( 9 ) for monitoring the spreading conditions of the casting powder is arranged . this sensor can be an infrared sensor or thermal sensor , and is used for detecting the exposed molten steel ( hot spot ) in the mold ( 4 ). based on the detection by the sensor , the multi - joint robot arm ( 3 ) is moved under automatic control of a computer so as to move the tip of the spreading casting powder feeder ( 2 ) to the hot spot for spreading . it is also possible to move the robot arm ( 3 ) according to a predetermined program for spreading , not using such a sensor . the base end of the spreading feeder ( 2 ) and the discharge port in the final - stage powder container ( 1 ) are connected with a flexible transfer path ( 8 ). the flexible transfer path comprises a transfer path having the flexibility to follow the movement of the robot arm , such as flexible pipe . therefore , non - flexible pipe may be used only in sections of the path where flexibility is not required . the flexible transfer path ( 8 ) is arranged from the discharge port of the powder container ( 1 ) above the tip arm ( 12a ) and along the robot arm ( 3 ) to the spreading feeder ( 2 ). however , to simplify the drawing , the illustration of the part along the robot arm ( 3 ) is omitted . it is possible to equip a feeding device in the flexible transfer path ( 8 ) along with the tip arm ( 12a ). the casting powder is transferred by said feeding device and dropped by gravity from the final - stage container ( 1 ) to the spreading feeder ( 2 ). this gravity drop is based on the energy saving concept using the height difference between the final - stage powder container ( 1 ) arranged in a high position and the spreading feeder ( 2 ) placed in low position , but forced transfer means can be added . ( i ) when the tundish car ( 5 ) stops at the position above the mold ( 4 ), the revolving arm ( 12 ) swings to move the feeder from the stand - by position ( i ) to the feed position ( ii ). ( ii ) the tip arm ( 12a ) of the revolving arm turns to face the powder feeder toward the mold . ( iii ) the robot arm ( 3 ) moves the tip of the spreading feeder ( 2 ) above a hot spot to spread the casting powder from its tip . in fig2 the area designated ( b ) shows the spreading area . ( iv ) in replacing the tundish , the powder feeder is returned to the stand - by position by the reverse operation of steps ( i ) and ( ii ) mentioned above . the flexibility of the multi - joint robot arm ( 3 ) removes the dead angle above the mold surface , and prevents contact between the spreading feeder ( 2 ) and the tundish strand nozzle ( 6a ), when the powder feeder is moved . the present invention is not limited to the above embodiment . if there is a high position such as a deck near the workshop , the final - stage container can be mounted on the deck and the column and revolving arm can be eliminated . the final - stage powder container may be also hung from a high position . since the final - stage powder container is arranged in a position higher than the working space on the powder feeder according to the invention , provision of this feeder in the operator side does not interfere with the work of an operator . since the operator side is free of the scattered dust and molten steel of the side opposite the operator , the following effects are obtained : ( i ) sharp decrease in trouble due to scattered dust and molten steel ,; ( iii ) improved work environment permitting use of precision instruments such as the robot arm ,; ( v ) the smaller distance to the control board or the operation board making the anti - nozzle provision for cpu wiring easier . in spreading the casting powder on the mold , the flexibility of the robot arm removes the dead angle on the spreading surface of the mold , and the powder is uniformly spreaded all over the mold surface . and the precision instruments and control equipment used in the robot arm can continue good operation in the good work environment as described in the item ( iii ) mentioned above . the provision of the spreading conditions monitor sensor on the robot arm , makes complete automation of hot spot detection and spreading possible by computer control of the spreading . this promotes labor saving , stabilizes the continuous casting and improves quality . this invention can be used in full automation of continuous casting . in further progress of continuous casting of highgrade steel , a powder feeder in continuous casting has been provided which can cope with feed automation of highgrade steel billet size casting powder use ."}
|
{"category": "Fixed Constructions", "patent": "with reference to the drawings , an embodiment of the present invention is described . the powder feeder comprises a final - stage powder container ( 1 ), a spreading feeder ( 2 ), and a multi - joint robot arm ( 3 ). the final - stage powder container ( 1 ) is equipped with a powder inlet line 13 and is mounted on the end of the revolving arm ( 12 ) held on the top of a column ( 11 ). final - stage powder container ( 1 ) is mounted at a position close to the base end of the tip arm ( 12a ). this revolving arm and tip arm are adequately driven by a driving device ( not shown ). here , the column ( 11 ) is taller than the height of an operator ( a ), thereby the revolving arm ( 12 ) and the final - stage powder container ( 1 ) are arranged above the working space of the mold . thus , the final - stage powder container ( 1 ) can be arranged on the operator side ( a ), for example , overhead of an operator . in fig2 two final - stage powder containers ( 1 )( 1 ) are arranged on the operator side . in the drawing , ( b ) indicates the side opposite an operator . a final - stage powder container ( 1 ) is provided with a meter such as a load cell platform scale ( 14 ) to weigh the spread quantity of casting powder . particularly , use of the &# 34 ; loss - in - weight &# 34 ; system permits recording of accurate spreading quantity and higher accuracy of control by a main computer of the continuous casting unit . the robot arm ( 3 ) comprises a base end arm ( 31 ), an intermediate arm ( 32 ), and a tip hand ( 33 ) movably connected by a first joint ( 34 ) and a second joint ( 35 ). the base end arm ( 31 ) is attached to the end of the tip arm ( 12a ) of the revolving arm . to the tip hand ( 33 ) is mounted a spreading feeder ( 2 ). the spreading feeder ( 2 ) rotates a spring in a tube by a motor ( 21 ) mounted at the tube &# 39 ; s base end to spread the casting powder from its tip on to the molten steel surface of the mold . in place of such mechanical means , other transfer means such as pneumatic transfer means can also be used . in the drawings , ( 5 ) represents a tundish car with a tundish ( 6 ) mounted thereon , a strand nozzle ( 6a ) ( 7 ) a ladle ( 7 ). to the tip on the intermediate arm ( 32 ) of the multi - joint robot arm , a sensor ( 9 ) for monitoring the spreading conditions of the casting powder is arranged . this sensor can be an infrared sensor or thermal sensor , and is used for detecting the exposed molten steel ( hot spot ) in the mold ( 4 ). based on the detection by the sensor , the multi - joint robot arm ( 3 ) is moved under automatic control of a computer so as to move the tip of the spreading casting powder feeder ( 2 ) to the hot spot for spreading . it is also possible to move the robot arm ( 3 ) according to a predetermined program for spreading , not using such a sensor . the base end of the spreading feeder ( 2 ) and the discharge port in the final - stage powder container ( 1 ) are connected with a flexible transfer path ( 8 ). the flexible transfer path comprises a transfer path having the flexibility to follow the movement of the robot arm , such as flexible pipe . therefore , non - flexible pipe may be used only in sections of the path where flexibility is not required . the flexible transfer path ( 8 ) is arranged from the discharge port of the powder container ( 1 ) above the tip arm ( 12a ) and along the robot arm ( 3 ) to the spreading feeder ( 2 ). however , to simplify the drawing , the illustration of the part along the robot arm ( 3 ) is omitted . it is possible to equip a feeding device in the flexible transfer path ( 8 ) along with the tip arm ( 12a ). the casting powder is transferred by said feeding device and dropped by gravity from the final - stage container ( 1 ) to the spreading feeder ( 2 ). this gravity drop is based on the energy saving concept using the height difference between the final - stage powder container ( 1 ) arranged in a high position and the spreading feeder ( 2 ) placed in low position , but forced transfer means can be added . ( i ) when the tundish car ( 5 ) stops at the position above the mold ( 4 ), the revolving arm ( 12 ) swings to move the feeder from the stand - by position ( i ) to the feed position ( ii ). ( ii ) the tip arm ( 12a ) of the revolving arm turns to face the powder feeder toward the mold . ( iii ) the robot arm ( 3 ) moves the tip of the spreading feeder ( 2 ) above a hot spot to spread the casting powder from its tip . in fig2 the area designated ( b ) shows the spreading area . ( iv ) in replacing the tundish , the powder feeder is returned to the stand - by position by the reverse operation of steps ( i ) and ( ii ) mentioned above . the flexibility of the multi - joint robot arm ( 3 ) removes the dead angle above the mold surface , and prevents contact between the spreading feeder ( 2 ) and the tundish strand nozzle ( 6a ), when the powder feeder is moved . the present invention is not limited to the above embodiment . if there is a high position such as a deck near the workshop , the final - stage container can be mounted on the deck and the column and revolving arm can be eliminated . the final - stage powder container may be also hung from a high position . since the final - stage powder container is arranged in a position higher than the working space on the powder feeder according to the invention , provision of this feeder in the operator side does not interfere with the work of an operator . since the operator side is free of the scattered dust and molten steel of the side opposite the operator , the following effects are obtained : ( i ) sharp decrease in trouble due to scattered dust and molten steel ,; ( iii ) improved work environment permitting use of precision instruments such as the robot arm ,; ( v ) the smaller distance to the control board or the operation board making the anti - nozzle provision for cpu wiring easier . in spreading the casting powder on the mold , the flexibility of the robot arm removes the dead angle on the spreading surface of the mold , and the powder is uniformly spreaded all over the mold surface . and the precision instruments and control equipment used in the robot arm can continue good operation in the good work environment as described in the item ( iii ) mentioned above . the provision of the spreading conditions monitor sensor on the robot arm , makes complete automation of hot spot detection and spreading possible by computer control of the spreading . this promotes labor saving , stabilizes the continuous casting and improves quality . this invention can be used in full automation of continuous casting . in further progress of continuous casting of highgrade steel , a powder feeder in continuous casting has been provided which can cope with feed automation of highgrade steel billet size casting powder use ."}
|
Does the patent belong in this category?
| 0.25 |
49d27878671b44cf2401f85d19ab2921adb1331aa61116bc349d37d60692ad6f
| 0.101074 | 0.031738 | 0.117676 | 0.5625 | 0.726563 | 0.400391 |
null |
{"patent": "with reference to the drawings , an embodiment of the present invention is described . the powder feeder comprises a final - stage powder container ( 1 ), a spreading feeder ( 2 ), and a multi - joint robot arm ( 3 ). the final - stage powder container ( 1 ) is equipped with a powder inlet line 13 and is mounted on the end of the revolving arm ( 12 ) held on the top of a column ( 11 ). final - stage powder container ( 1 ) is mounted at a position close to the base end of the tip arm ( 12a ). this revolving arm and tip arm are adequately driven by a driving device ( not shown ). here , the column ( 11 ) is taller than the height of an operator ( a ), thereby the revolving arm ( 12 ) and the final - stage powder container ( 1 ) are arranged above the working space of the mold . thus , the final - stage powder container ( 1 ) can be arranged on the operator side ( a ), for example , overhead of an operator . in fig2 two final - stage powder containers ( 1 )( 1 ) are arranged on the operator side . in the drawing , ( b ) indicates the side opposite an operator . a final - stage powder container ( 1 ) is provided with a meter such as a load cell platform scale ( 14 ) to weigh the spread quantity of casting powder . particularly , use of the &# 34 ; loss - in - weight &# 34 ; system permits recording of accurate spreading quantity and higher accuracy of control by a main computer of the continuous casting unit . the robot arm ( 3 ) comprises a base end arm ( 31 ), an intermediate arm ( 32 ), and a tip hand ( 33 ) movably connected by a first joint ( 34 ) and a second joint ( 35 ). the base end arm ( 31 ) is attached to the end of the tip arm ( 12a ) of the revolving arm . to the tip hand ( 33 ) is mounted a spreading feeder ( 2 ). the spreading feeder ( 2 ) rotates a spring in a tube by a motor ( 21 ) mounted at the tube &# 39 ; s base end to spread the casting powder from its tip on to the molten steel surface of the mold . in place of such mechanical means , other transfer means such as pneumatic transfer means can also be used . in the drawings , ( 5 ) represents a tundish car with a tundish ( 6 ) mounted thereon , a strand nozzle ( 6a ) ( 7 ) a ladle ( 7 ). to the tip on the intermediate arm ( 32 ) of the multi - joint robot arm , a sensor ( 9 ) for monitoring the spreading conditions of the casting powder is arranged . this sensor can be an infrared sensor or thermal sensor , and is used for detecting the exposed molten steel ( hot spot ) in the mold ( 4 ). based on the detection by the sensor , the multi - joint robot arm ( 3 ) is moved under automatic control of a computer so as to move the tip of the spreading casting powder feeder ( 2 ) to the hot spot for spreading . it is also possible to move the robot arm ( 3 ) according to a predetermined program for spreading , not using such a sensor . the base end of the spreading feeder ( 2 ) and the discharge port in the final - stage powder container ( 1 ) are connected with a flexible transfer path ( 8 ). the flexible transfer path comprises a transfer path having the flexibility to follow the movement of the robot arm , such as flexible pipe . therefore , non - flexible pipe may be used only in sections of the path where flexibility is not required . the flexible transfer path ( 8 ) is arranged from the discharge port of the powder container ( 1 ) above the tip arm ( 12a ) and along the robot arm ( 3 ) to the spreading feeder ( 2 ). however , to simplify the drawing , the illustration of the part along the robot arm ( 3 ) is omitted . it is possible to equip a feeding device in the flexible transfer path ( 8 ) along with the tip arm ( 12a ). the casting powder is transferred by said feeding device and dropped by gravity from the final - stage container ( 1 ) to the spreading feeder ( 2 ). this gravity drop is based on the energy saving concept using the height difference between the final - stage powder container ( 1 ) arranged in a high position and the spreading feeder ( 2 ) placed in low position , but forced transfer means can be added . ( i ) when the tundish car ( 5 ) stops at the position above the mold ( 4 ), the revolving arm ( 12 ) swings to move the feeder from the stand - by position ( i ) to the feed position ( ii ). ( ii ) the tip arm ( 12a ) of the revolving arm turns to face the powder feeder toward the mold . ( iii ) the robot arm ( 3 ) moves the tip of the spreading feeder ( 2 ) above a hot spot to spread the casting powder from its tip . in fig2 the area designated ( b ) shows the spreading area . ( iv ) in replacing the tundish , the powder feeder is returned to the stand - by position by the reverse operation of steps ( i ) and ( ii ) mentioned above . the flexibility of the multi - joint robot arm ( 3 ) removes the dead angle above the mold surface , and prevents contact between the spreading feeder ( 2 ) and the tundish strand nozzle ( 6a ), when the powder feeder is moved . the present invention is not limited to the above embodiment . if there is a high position such as a deck near the workshop , the final - stage container can be mounted on the deck and the column and revolving arm can be eliminated . the final - stage powder container may be also hung from a high position . since the final - stage powder container is arranged in a position higher than the working space on the powder feeder according to the invention , provision of this feeder in the operator side does not interfere with the work of an operator . since the operator side is free of the scattered dust and molten steel of the side opposite the operator , the following effects are obtained : ( i ) sharp decrease in trouble due to scattered dust and molten steel ,; ( iii ) improved work environment permitting use of precision instruments such as the robot arm ,; ( v ) the smaller distance to the control board or the operation board making the anti - nozzle provision for cpu wiring easier . in spreading the casting powder on the mold , the flexibility of the robot arm removes the dead angle on the spreading surface of the mold , and the powder is uniformly spreaded all over the mold surface . and the precision instruments and control equipment used in the robot arm can continue good operation in the good work environment as described in the item ( iii ) mentioned above . the provision of the spreading conditions monitor sensor on the robot arm , makes complete automation of hot spot detection and spreading possible by computer control of the spreading . this promotes labor saving , stabilizes the continuous casting and improves quality . this invention can be used in full automation of continuous casting . in further progress of continuous casting of highgrade steel , a powder feeder in continuous casting has been provided which can cope with feed automation of highgrade steel billet size casting powder use .", "category": "Performing Operations; Transporting"}
|
{"patent": "with reference to the drawings , an embodiment of the present invention is described . the powder feeder comprises a final - stage powder container ( 1 ), a spreading feeder ( 2 ), and a multi - joint robot arm ( 3 ). the final - stage powder container ( 1 ) is equipped with a powder inlet line 13 and is mounted on the end of the revolving arm ( 12 ) held on the top of a column ( 11 ). final - stage powder container ( 1 ) is mounted at a position close to the base end of the tip arm ( 12a ). this revolving arm and tip arm are adequately driven by a driving device ( not shown ). here , the column ( 11 ) is taller than the height of an operator ( a ), thereby the revolving arm ( 12 ) and the final - stage powder container ( 1 ) are arranged above the working space of the mold . thus , the final - stage powder container ( 1 ) can be arranged on the operator side ( a ), for example , overhead of an operator . in fig2 two final - stage powder containers ( 1 )( 1 ) are arranged on the operator side . in the drawing , ( b ) indicates the side opposite an operator . a final - stage powder container ( 1 ) is provided with a meter such as a load cell platform scale ( 14 ) to weigh the spread quantity of casting powder . particularly , use of the &# 34 ; loss - in - weight &# 34 ; system permits recording of accurate spreading quantity and higher accuracy of control by a main computer of the continuous casting unit . the robot arm ( 3 ) comprises a base end arm ( 31 ), an intermediate arm ( 32 ), and a tip hand ( 33 ) movably connected by a first joint ( 34 ) and a second joint ( 35 ). the base end arm ( 31 ) is attached to the end of the tip arm ( 12a ) of the revolving arm . to the tip hand ( 33 ) is mounted a spreading feeder ( 2 ). the spreading feeder ( 2 ) rotates a spring in a tube by a motor ( 21 ) mounted at the tube &# 39 ; s base end to spread the casting powder from its tip on to the molten steel surface of the mold . in place of such mechanical means , other transfer means such as pneumatic transfer means can also be used . in the drawings , ( 5 ) represents a tundish car with a tundish ( 6 ) mounted thereon , a strand nozzle ( 6a ) ( 7 ) a ladle ( 7 ). to the tip on the intermediate arm ( 32 ) of the multi - joint robot arm , a sensor ( 9 ) for monitoring the spreading conditions of the casting powder is arranged . this sensor can be an infrared sensor or thermal sensor , and is used for detecting the exposed molten steel ( hot spot ) in the mold ( 4 ). based on the detection by the sensor , the multi - joint robot arm ( 3 ) is moved under automatic control of a computer so as to move the tip of the spreading casting powder feeder ( 2 ) to the hot spot for spreading . it is also possible to move the robot arm ( 3 ) according to a predetermined program for spreading , not using such a sensor . the base end of the spreading feeder ( 2 ) and the discharge port in the final - stage powder container ( 1 ) are connected with a flexible transfer path ( 8 ). the flexible transfer path comprises a transfer path having the flexibility to follow the movement of the robot arm , such as flexible pipe . therefore , non - flexible pipe may be used only in sections of the path where flexibility is not required . the flexible transfer path ( 8 ) is arranged from the discharge port of the powder container ( 1 ) above the tip arm ( 12a ) and along the robot arm ( 3 ) to the spreading feeder ( 2 ). however , to simplify the drawing , the illustration of the part along the robot arm ( 3 ) is omitted . it is possible to equip a feeding device in the flexible transfer path ( 8 ) along with the tip arm ( 12a ). the casting powder is transferred by said feeding device and dropped by gravity from the final - stage container ( 1 ) to the spreading feeder ( 2 ). this gravity drop is based on the energy saving concept using the height difference between the final - stage powder container ( 1 ) arranged in a high position and the spreading feeder ( 2 ) placed in low position , but forced transfer means can be added . ( i ) when the tundish car ( 5 ) stops at the position above the mold ( 4 ), the revolving arm ( 12 ) swings to move the feeder from the stand - by position ( i ) to the feed position ( ii ). ( ii ) the tip arm ( 12a ) of the revolving arm turns to face the powder feeder toward the mold . ( iii ) the robot arm ( 3 ) moves the tip of the spreading feeder ( 2 ) above a hot spot to spread the casting powder from its tip . in fig2 the area designated ( b ) shows the spreading area . ( iv ) in replacing the tundish , the powder feeder is returned to the stand - by position by the reverse operation of steps ( i ) and ( ii ) mentioned above . the flexibility of the multi - joint robot arm ( 3 ) removes the dead angle above the mold surface , and prevents contact between the spreading feeder ( 2 ) and the tundish strand nozzle ( 6a ), when the powder feeder is moved . the present invention is not limited to the above embodiment . if there is a high position such as a deck near the workshop , the final - stage container can be mounted on the deck and the column and revolving arm can be eliminated . the final - stage powder container may be also hung from a high position . since the final - stage powder container is arranged in a position higher than the working space on the powder feeder according to the invention , provision of this feeder in the operator side does not interfere with the work of an operator . since the operator side is free of the scattered dust and molten steel of the side opposite the operator , the following effects are obtained : ( i ) sharp decrease in trouble due to scattered dust and molten steel ,; ( iii ) improved work environment permitting use of precision instruments such as the robot arm ,; ( v ) the smaller distance to the control board or the operation board making the anti - nozzle provision for cpu wiring easier . in spreading the casting powder on the mold , the flexibility of the robot arm removes the dead angle on the spreading surface of the mold , and the powder is uniformly spreaded all over the mold surface . and the precision instruments and control equipment used in the robot arm can continue good operation in the good work environment as described in the item ( iii ) mentioned above . the provision of the spreading conditions monitor sensor on the robot arm , makes complete automation of hot spot detection and spreading possible by computer control of the spreading . this promotes labor saving , stabilizes the continuous casting and improves quality . this invention can be used in full automation of continuous casting . in further progress of continuous casting of highgrade steel , a powder feeder in continuous casting has been provided which can cope with feed automation of highgrade steel billet size casting powder use .", "category": "Mechanical Engineering; Lightning; Heating; Weapons; Blasting"}
|
Does the patent belong in this category?
| 0.25 |
49d27878671b44cf2401f85d19ab2921adb1331aa61116bc349d37d60692ad6f
| 0.051025 | 0.006104 | 0.359375 | 0.036865 | 0.251953 | 0.022949 |
null |
{"patent": "with reference to the drawings , an embodiment of the present invention is described . the powder feeder comprises a final - stage powder container ( 1 ), a spreading feeder ( 2 ), and a multi - joint robot arm ( 3 ). the final - stage powder container ( 1 ) is equipped with a powder inlet line 13 and is mounted on the end of the revolving arm ( 12 ) held on the top of a column ( 11 ). final - stage powder container ( 1 ) is mounted at a position close to the base end of the tip arm ( 12a ). this revolving arm and tip arm are adequately driven by a driving device ( not shown ). here , the column ( 11 ) is taller than the height of an operator ( a ), thereby the revolving arm ( 12 ) and the final - stage powder container ( 1 ) are arranged above the working space of the mold . thus , the final - stage powder container ( 1 ) can be arranged on the operator side ( a ), for example , overhead of an operator . in fig2 two final - stage powder containers ( 1 )( 1 ) are arranged on the operator side . in the drawing , ( b ) indicates the side opposite an operator . a final - stage powder container ( 1 ) is provided with a meter such as a load cell platform scale ( 14 ) to weigh the spread quantity of casting powder . particularly , use of the &# 34 ; loss - in - weight &# 34 ; system permits recording of accurate spreading quantity and higher accuracy of control by a main computer of the continuous casting unit . the robot arm ( 3 ) comprises a base end arm ( 31 ), an intermediate arm ( 32 ), and a tip hand ( 33 ) movably connected by a first joint ( 34 ) and a second joint ( 35 ). the base end arm ( 31 ) is attached to the end of the tip arm ( 12a ) of the revolving arm . to the tip hand ( 33 ) is mounted a spreading feeder ( 2 ). the spreading feeder ( 2 ) rotates a spring in a tube by a motor ( 21 ) mounted at the tube &# 39 ; s base end to spread the casting powder from its tip on to the molten steel surface of the mold . in place of such mechanical means , other transfer means such as pneumatic transfer means can also be used . in the drawings , ( 5 ) represents a tundish car with a tundish ( 6 ) mounted thereon , a strand nozzle ( 6a ) ( 7 ) a ladle ( 7 ). to the tip on the intermediate arm ( 32 ) of the multi - joint robot arm , a sensor ( 9 ) for monitoring the spreading conditions of the casting powder is arranged . this sensor can be an infrared sensor or thermal sensor , and is used for detecting the exposed molten steel ( hot spot ) in the mold ( 4 ). based on the detection by the sensor , the multi - joint robot arm ( 3 ) is moved under automatic control of a computer so as to move the tip of the spreading casting powder feeder ( 2 ) to the hot spot for spreading . it is also possible to move the robot arm ( 3 ) according to a predetermined program for spreading , not using such a sensor . the base end of the spreading feeder ( 2 ) and the discharge port in the final - stage powder container ( 1 ) are connected with a flexible transfer path ( 8 ). the flexible transfer path comprises a transfer path having the flexibility to follow the movement of the robot arm , such as flexible pipe . therefore , non - flexible pipe may be used only in sections of the path where flexibility is not required . the flexible transfer path ( 8 ) is arranged from the discharge port of the powder container ( 1 ) above the tip arm ( 12a ) and along the robot arm ( 3 ) to the spreading feeder ( 2 ). however , to simplify the drawing , the illustration of the part along the robot arm ( 3 ) is omitted . it is possible to equip a feeding device in the flexible transfer path ( 8 ) along with the tip arm ( 12a ). the casting powder is transferred by said feeding device and dropped by gravity from the final - stage container ( 1 ) to the spreading feeder ( 2 ). this gravity drop is based on the energy saving concept using the height difference between the final - stage powder container ( 1 ) arranged in a high position and the spreading feeder ( 2 ) placed in low position , but forced transfer means can be added . ( i ) when the tundish car ( 5 ) stops at the position above the mold ( 4 ), the revolving arm ( 12 ) swings to move the feeder from the stand - by position ( i ) to the feed position ( ii ). ( ii ) the tip arm ( 12a ) of the revolving arm turns to face the powder feeder toward the mold . ( iii ) the robot arm ( 3 ) moves the tip of the spreading feeder ( 2 ) above a hot spot to spread the casting powder from its tip . in fig2 the area designated ( b ) shows the spreading area . ( iv ) in replacing the tundish , the powder feeder is returned to the stand - by position by the reverse operation of steps ( i ) and ( ii ) mentioned above . the flexibility of the multi - joint robot arm ( 3 ) removes the dead angle above the mold surface , and prevents contact between the spreading feeder ( 2 ) and the tundish strand nozzle ( 6a ), when the powder feeder is moved . the present invention is not limited to the above embodiment . if there is a high position such as a deck near the workshop , the final - stage container can be mounted on the deck and the column and revolving arm can be eliminated . the final - stage powder container may be also hung from a high position . since the final - stage powder container is arranged in a position higher than the working space on the powder feeder according to the invention , provision of this feeder in the operator side does not interfere with the work of an operator . since the operator side is free of the scattered dust and molten steel of the side opposite the operator , the following effects are obtained : ( i ) sharp decrease in trouble due to scattered dust and molten steel ,; ( iii ) improved work environment permitting use of precision instruments such as the robot arm ,; ( v ) the smaller distance to the control board or the operation board making the anti - nozzle provision for cpu wiring easier . in spreading the casting powder on the mold , the flexibility of the robot arm removes the dead angle on the spreading surface of the mold , and the powder is uniformly spreaded all over the mold surface . and the precision instruments and control equipment used in the robot arm can continue good operation in the good work environment as described in the item ( iii ) mentioned above . the provision of the spreading conditions monitor sensor on the robot arm , makes complete automation of hot spot detection and spreading possible by computer control of the spreading . this promotes labor saving , stabilizes the continuous casting and improves quality . this invention can be used in full automation of continuous casting . in further progress of continuous casting of highgrade steel , a powder feeder in continuous casting has been provided which can cope with feed automation of highgrade steel billet size casting powder use .", "category": "Performing Operations; Transporting"}
|
{"patent": "with reference to the drawings , an embodiment of the present invention is described . the powder feeder comprises a final - stage powder container ( 1 ), a spreading feeder ( 2 ), and a multi - joint robot arm ( 3 ). the final - stage powder container ( 1 ) is equipped with a powder inlet line 13 and is mounted on the end of the revolving arm ( 12 ) held on the top of a column ( 11 ). final - stage powder container ( 1 ) is mounted at a position close to the base end of the tip arm ( 12a ). this revolving arm and tip arm are adequately driven by a driving device ( not shown ). here , the column ( 11 ) is taller than the height of an operator ( a ), thereby the revolving arm ( 12 ) and the final - stage powder container ( 1 ) are arranged above the working space of the mold . thus , the final - stage powder container ( 1 ) can be arranged on the operator side ( a ), for example , overhead of an operator . in fig2 two final - stage powder containers ( 1 )( 1 ) are arranged on the operator side . in the drawing , ( b ) indicates the side opposite an operator . a final - stage powder container ( 1 ) is provided with a meter such as a load cell platform scale ( 14 ) to weigh the spread quantity of casting powder . particularly , use of the &# 34 ; loss - in - weight &# 34 ; system permits recording of accurate spreading quantity and higher accuracy of control by a main computer of the continuous casting unit . the robot arm ( 3 ) comprises a base end arm ( 31 ), an intermediate arm ( 32 ), and a tip hand ( 33 ) movably connected by a first joint ( 34 ) and a second joint ( 35 ). the base end arm ( 31 ) is attached to the end of the tip arm ( 12a ) of the revolving arm . to the tip hand ( 33 ) is mounted a spreading feeder ( 2 ). the spreading feeder ( 2 ) rotates a spring in a tube by a motor ( 21 ) mounted at the tube &# 39 ; s base end to spread the casting powder from its tip on to the molten steel surface of the mold . in place of such mechanical means , other transfer means such as pneumatic transfer means can also be used . in the drawings , ( 5 ) represents a tundish car with a tundish ( 6 ) mounted thereon , a strand nozzle ( 6a ) ( 7 ) a ladle ( 7 ). to the tip on the intermediate arm ( 32 ) of the multi - joint robot arm , a sensor ( 9 ) for monitoring the spreading conditions of the casting powder is arranged . this sensor can be an infrared sensor or thermal sensor , and is used for detecting the exposed molten steel ( hot spot ) in the mold ( 4 ). based on the detection by the sensor , the multi - joint robot arm ( 3 ) is moved under automatic control of a computer so as to move the tip of the spreading casting powder feeder ( 2 ) to the hot spot for spreading . it is also possible to move the robot arm ( 3 ) according to a predetermined program for spreading , not using such a sensor . the base end of the spreading feeder ( 2 ) and the discharge port in the final - stage powder container ( 1 ) are connected with a flexible transfer path ( 8 ). the flexible transfer path comprises a transfer path having the flexibility to follow the movement of the robot arm , such as flexible pipe . therefore , non - flexible pipe may be used only in sections of the path where flexibility is not required . the flexible transfer path ( 8 ) is arranged from the discharge port of the powder container ( 1 ) above the tip arm ( 12a ) and along the robot arm ( 3 ) to the spreading feeder ( 2 ). however , to simplify the drawing , the illustration of the part along the robot arm ( 3 ) is omitted . it is possible to equip a feeding device in the flexible transfer path ( 8 ) along with the tip arm ( 12a ). the casting powder is transferred by said feeding device and dropped by gravity from the final - stage container ( 1 ) to the spreading feeder ( 2 ). this gravity drop is based on the energy saving concept using the height difference between the final - stage powder container ( 1 ) arranged in a high position and the spreading feeder ( 2 ) placed in low position , but forced transfer means can be added . ( i ) when the tundish car ( 5 ) stops at the position above the mold ( 4 ), the revolving arm ( 12 ) swings to move the feeder from the stand - by position ( i ) to the feed position ( ii ). ( ii ) the tip arm ( 12a ) of the revolving arm turns to face the powder feeder toward the mold . ( iii ) the robot arm ( 3 ) moves the tip of the spreading feeder ( 2 ) above a hot spot to spread the casting powder from its tip . in fig2 the area designated ( b ) shows the spreading area . ( iv ) in replacing the tundish , the powder feeder is returned to the stand - by position by the reverse operation of steps ( i ) and ( ii ) mentioned above . the flexibility of the multi - joint robot arm ( 3 ) removes the dead angle above the mold surface , and prevents contact between the spreading feeder ( 2 ) and the tundish strand nozzle ( 6a ), when the powder feeder is moved . the present invention is not limited to the above embodiment . if there is a high position such as a deck near the workshop , the final - stage container can be mounted on the deck and the column and revolving arm can be eliminated . the final - stage powder container may be also hung from a high position . since the final - stage powder container is arranged in a position higher than the working space on the powder feeder according to the invention , provision of this feeder in the operator side does not interfere with the work of an operator . since the operator side is free of the scattered dust and molten steel of the side opposite the operator , the following effects are obtained : ( i ) sharp decrease in trouble due to scattered dust and molten steel ,; ( iii ) improved work environment permitting use of precision instruments such as the robot arm ,; ( v ) the smaller distance to the control board or the operation board making the anti - nozzle provision for cpu wiring easier . in spreading the casting powder on the mold , the flexibility of the robot arm removes the dead angle on the spreading surface of the mold , and the powder is uniformly spreaded all over the mold surface . and the precision instruments and control equipment used in the robot arm can continue good operation in the good work environment as described in the item ( iii ) mentioned above . the provision of the spreading conditions monitor sensor on the robot arm , makes complete automation of hot spot detection and spreading possible by computer control of the spreading . this promotes labor saving , stabilizes the continuous casting and improves quality . this invention can be used in full automation of continuous casting . in further progress of continuous casting of highgrade steel , a powder feeder in continuous casting has been provided which can cope with feed automation of highgrade steel billet size casting powder use .", "category": "Physics"}
|
Is the patent correctly categorized?
| 0.25 |
49d27878671b44cf2401f85d19ab2921adb1331aa61116bc349d37d60692ad6f
| 0.016357 | 0.003479 | 0.235352 | 0.006683 | 0.204102 | 0.011353 |
null |
{"category": "Performing Operations; Transporting", "patent": "with reference to the drawings , an embodiment of the present invention is described . the powder feeder comprises a final - stage powder container ( 1 ), a spreading feeder ( 2 ), and a multi - joint robot arm ( 3 ). the final - stage powder container ( 1 ) is equipped with a powder inlet line 13 and is mounted on the end of the revolving arm ( 12 ) held on the top of a column ( 11 ). final - stage powder container ( 1 ) is mounted at a position close to the base end of the tip arm ( 12a ). this revolving arm and tip arm are adequately driven by a driving device ( not shown ). here , the column ( 11 ) is taller than the height of an operator ( a ), thereby the revolving arm ( 12 ) and the final - stage powder container ( 1 ) are arranged above the working space of the mold . thus , the final - stage powder container ( 1 ) can be arranged on the operator side ( a ), for example , overhead of an operator . in fig2 two final - stage powder containers ( 1 )( 1 ) are arranged on the operator side . in the drawing , ( b ) indicates the side opposite an operator . a final - stage powder container ( 1 ) is provided with a meter such as a load cell platform scale ( 14 ) to weigh the spread quantity of casting powder . particularly , use of the &# 34 ; loss - in - weight &# 34 ; system permits recording of accurate spreading quantity and higher accuracy of control by a main computer of the continuous casting unit . the robot arm ( 3 ) comprises a base end arm ( 31 ), an intermediate arm ( 32 ), and a tip hand ( 33 ) movably connected by a first joint ( 34 ) and a second joint ( 35 ). the base end arm ( 31 ) is attached to the end of the tip arm ( 12a ) of the revolving arm . to the tip hand ( 33 ) is mounted a spreading feeder ( 2 ). the spreading feeder ( 2 ) rotates a spring in a tube by a motor ( 21 ) mounted at the tube &# 39 ; s base end to spread the casting powder from its tip on to the molten steel surface of the mold . in place of such mechanical means , other transfer means such as pneumatic transfer means can also be used . in the drawings , ( 5 ) represents a tundish car with a tundish ( 6 ) mounted thereon , a strand nozzle ( 6a ) ( 7 ) a ladle ( 7 ). to the tip on the intermediate arm ( 32 ) of the multi - joint robot arm , a sensor ( 9 ) for monitoring the spreading conditions of the casting powder is arranged . this sensor can be an infrared sensor or thermal sensor , and is used for detecting the exposed molten steel ( hot spot ) in the mold ( 4 ). based on the detection by the sensor , the multi - joint robot arm ( 3 ) is moved under automatic control of a computer so as to move the tip of the spreading casting powder feeder ( 2 ) to the hot spot for spreading . it is also possible to move the robot arm ( 3 ) according to a predetermined program for spreading , not using such a sensor . the base end of the spreading feeder ( 2 ) and the discharge port in the final - stage powder container ( 1 ) are connected with a flexible transfer path ( 8 ). the flexible transfer path comprises a transfer path having the flexibility to follow the movement of the robot arm , such as flexible pipe . therefore , non - flexible pipe may be used only in sections of the path where flexibility is not required . the flexible transfer path ( 8 ) is arranged from the discharge port of the powder container ( 1 ) above the tip arm ( 12a ) and along the robot arm ( 3 ) to the spreading feeder ( 2 ). however , to simplify the drawing , the illustration of the part along the robot arm ( 3 ) is omitted . it is possible to equip a feeding device in the flexible transfer path ( 8 ) along with the tip arm ( 12a ). the casting powder is transferred by said feeding device and dropped by gravity from the final - stage container ( 1 ) to the spreading feeder ( 2 ). this gravity drop is based on the energy saving concept using the height difference between the final - stage powder container ( 1 ) arranged in a high position and the spreading feeder ( 2 ) placed in low position , but forced transfer means can be added . ( i ) when the tundish car ( 5 ) stops at the position above the mold ( 4 ), the revolving arm ( 12 ) swings to move the feeder from the stand - by position ( i ) to the feed position ( ii ). ( ii ) the tip arm ( 12a ) of the revolving arm turns to face the powder feeder toward the mold . ( iii ) the robot arm ( 3 ) moves the tip of the spreading feeder ( 2 ) above a hot spot to spread the casting powder from its tip . in fig2 the area designated ( b ) shows the spreading area . ( iv ) in replacing the tundish , the powder feeder is returned to the stand - by position by the reverse operation of steps ( i ) and ( ii ) mentioned above . the flexibility of the multi - joint robot arm ( 3 ) removes the dead angle above the mold surface , and prevents contact between the spreading feeder ( 2 ) and the tundish strand nozzle ( 6a ), when the powder feeder is moved . the present invention is not limited to the above embodiment . if there is a high position such as a deck near the workshop , the final - stage container can be mounted on the deck and the column and revolving arm can be eliminated . the final - stage powder container may be also hung from a high position . since the final - stage powder container is arranged in a position higher than the working space on the powder feeder according to the invention , provision of this feeder in the operator side does not interfere with the work of an operator . since the operator side is free of the scattered dust and molten steel of the side opposite the operator , the following effects are obtained : ( i ) sharp decrease in trouble due to scattered dust and molten steel ,; ( iii ) improved work environment permitting use of precision instruments such as the robot arm ,; ( v ) the smaller distance to the control board or the operation board making the anti - nozzle provision for cpu wiring easier . in spreading the casting powder on the mold , the flexibility of the robot arm removes the dead angle on the spreading surface of the mold , and the powder is uniformly spreaded all over the mold surface . and the precision instruments and control equipment used in the robot arm can continue good operation in the good work environment as described in the item ( iii ) mentioned above . the provision of the spreading conditions monitor sensor on the robot arm , makes complete automation of hot spot detection and spreading possible by computer control of the spreading . this promotes labor saving , stabilizes the continuous casting and improves quality . this invention can be used in full automation of continuous casting . in further progress of continuous casting of highgrade steel , a powder feeder in continuous casting has been provided which can cope with feed automation of highgrade steel billet size casting powder use ."}
|
{"patent": "with reference to the drawings , an embodiment of the present invention is described . the powder feeder comprises a final - stage powder container ( 1 ), a spreading feeder ( 2 ), and a multi - joint robot arm ( 3 ). the final - stage powder container ( 1 ) is equipped with a powder inlet line 13 and is mounted on the end of the revolving arm ( 12 ) held on the top of a column ( 11 ). final - stage powder container ( 1 ) is mounted at a position close to the base end of the tip arm ( 12a ). this revolving arm and tip arm are adequately driven by a driving device ( not shown ). here , the column ( 11 ) is taller than the height of an operator ( a ), thereby the revolving arm ( 12 ) and the final - stage powder container ( 1 ) are arranged above the working space of the mold . thus , the final - stage powder container ( 1 ) can be arranged on the operator side ( a ), for example , overhead of an operator . in fig2 two final - stage powder containers ( 1 )( 1 ) are arranged on the operator side . in the drawing , ( b ) indicates the side opposite an operator . a final - stage powder container ( 1 ) is provided with a meter such as a load cell platform scale ( 14 ) to weigh the spread quantity of casting powder . particularly , use of the &# 34 ; loss - in - weight &# 34 ; system permits recording of accurate spreading quantity and higher accuracy of control by a main computer of the continuous casting unit . the robot arm ( 3 ) comprises a base end arm ( 31 ), an intermediate arm ( 32 ), and a tip hand ( 33 ) movably connected by a first joint ( 34 ) and a second joint ( 35 ). the base end arm ( 31 ) is attached to the end of the tip arm ( 12a ) of the revolving arm . to the tip hand ( 33 ) is mounted a spreading feeder ( 2 ). the spreading feeder ( 2 ) rotates a spring in a tube by a motor ( 21 ) mounted at the tube &# 39 ; s base end to spread the casting powder from its tip on to the molten steel surface of the mold . in place of such mechanical means , other transfer means such as pneumatic transfer means can also be used . in the drawings , ( 5 ) represents a tundish car with a tundish ( 6 ) mounted thereon , a strand nozzle ( 6a ) ( 7 ) a ladle ( 7 ). to the tip on the intermediate arm ( 32 ) of the multi - joint robot arm , a sensor ( 9 ) for monitoring the spreading conditions of the casting powder is arranged . this sensor can be an infrared sensor or thermal sensor , and is used for detecting the exposed molten steel ( hot spot ) in the mold ( 4 ). based on the detection by the sensor , the multi - joint robot arm ( 3 ) is moved under automatic control of a computer so as to move the tip of the spreading casting powder feeder ( 2 ) to the hot spot for spreading . it is also possible to move the robot arm ( 3 ) according to a predetermined program for spreading , not using such a sensor . the base end of the spreading feeder ( 2 ) and the discharge port in the final - stage powder container ( 1 ) are connected with a flexible transfer path ( 8 ). the flexible transfer path comprises a transfer path having the flexibility to follow the movement of the robot arm , such as flexible pipe . therefore , non - flexible pipe may be used only in sections of the path where flexibility is not required . the flexible transfer path ( 8 ) is arranged from the discharge port of the powder container ( 1 ) above the tip arm ( 12a ) and along the robot arm ( 3 ) to the spreading feeder ( 2 ). however , to simplify the drawing , the illustration of the part along the robot arm ( 3 ) is omitted . it is possible to equip a feeding device in the flexible transfer path ( 8 ) along with the tip arm ( 12a ). the casting powder is transferred by said feeding device and dropped by gravity from the final - stage container ( 1 ) to the spreading feeder ( 2 ). this gravity drop is based on the energy saving concept using the height difference between the final - stage powder container ( 1 ) arranged in a high position and the spreading feeder ( 2 ) placed in low position , but forced transfer means can be added . ( i ) when the tundish car ( 5 ) stops at the position above the mold ( 4 ), the revolving arm ( 12 ) swings to move the feeder from the stand - by position ( i ) to the feed position ( ii ). ( ii ) the tip arm ( 12a ) of the revolving arm turns to face the powder feeder toward the mold . ( iii ) the robot arm ( 3 ) moves the tip of the spreading feeder ( 2 ) above a hot spot to spread the casting powder from its tip . in fig2 the area designated ( b ) shows the spreading area . ( iv ) in replacing the tundish , the powder feeder is returned to the stand - by position by the reverse operation of steps ( i ) and ( ii ) mentioned above . the flexibility of the multi - joint robot arm ( 3 ) removes the dead angle above the mold surface , and prevents contact between the spreading feeder ( 2 ) and the tundish strand nozzle ( 6a ), when the powder feeder is moved . the present invention is not limited to the above embodiment . if there is a high position such as a deck near the workshop , the final - stage container can be mounted on the deck and the column and revolving arm can be eliminated . the final - stage powder container may be also hung from a high position . since the final - stage powder container is arranged in a position higher than the working space on the powder feeder according to the invention , provision of this feeder in the operator side does not interfere with the work of an operator . since the operator side is free of the scattered dust and molten steel of the side opposite the operator , the following effects are obtained : ( i ) sharp decrease in trouble due to scattered dust and molten steel ,; ( iii ) improved work environment permitting use of precision instruments such as the robot arm ,; ( v ) the smaller distance to the control board or the operation board making the anti - nozzle provision for cpu wiring easier . in spreading the casting powder on the mold , the flexibility of the robot arm removes the dead angle on the spreading surface of the mold , and the powder is uniformly spreaded all over the mold surface . and the precision instruments and control equipment used in the robot arm can continue good operation in the good work environment as described in the item ( iii ) mentioned above . the provision of the spreading conditions monitor sensor on the robot arm , makes complete automation of hot spot detection and spreading possible by computer control of the spreading . this promotes labor saving , stabilizes the continuous casting and improves quality . this invention can be used in full automation of continuous casting . in further progress of continuous casting of highgrade steel , a powder feeder in continuous casting has been provided which can cope with feed automation of highgrade steel billet size casting powder use .", "category": "Electricity"}
|
Does the category match the content of the patent?
| 0.25 |
49d27878671b44cf2401f85d19ab2921adb1331aa61116bc349d37d60692ad6f
| 0.088867 | 0.006683 | 0.077148 | 0.008057 | 0.574219 | 0.011658 |
null |
{"category": "Performing Operations; Transporting", "patent": "with reference to the drawings , an embodiment of the present invention is described . the powder feeder comprises a final - stage powder container ( 1 ), a spreading feeder ( 2 ), and a multi - joint robot arm ( 3 ). the final - stage powder container ( 1 ) is equipped with a powder inlet line 13 and is mounted on the end of the revolving arm ( 12 ) held on the top of a column ( 11 ). final - stage powder container ( 1 ) is mounted at a position close to the base end of the tip arm ( 12a ). this revolving arm and tip arm are adequately driven by a driving device ( not shown ). here , the column ( 11 ) is taller than the height of an operator ( a ), thereby the revolving arm ( 12 ) and the final - stage powder container ( 1 ) are arranged above the working space of the mold . thus , the final - stage powder container ( 1 ) can be arranged on the operator side ( a ), for example , overhead of an operator . in fig2 two final - stage powder containers ( 1 )( 1 ) are arranged on the operator side . in the drawing , ( b ) indicates the side opposite an operator . a final - stage powder container ( 1 ) is provided with a meter such as a load cell platform scale ( 14 ) to weigh the spread quantity of casting powder . particularly , use of the &# 34 ; loss - in - weight &# 34 ; system permits recording of accurate spreading quantity and higher accuracy of control by a main computer of the continuous casting unit . the robot arm ( 3 ) comprises a base end arm ( 31 ), an intermediate arm ( 32 ), and a tip hand ( 33 ) movably connected by a first joint ( 34 ) and a second joint ( 35 ). the base end arm ( 31 ) is attached to the end of the tip arm ( 12a ) of the revolving arm . to the tip hand ( 33 ) is mounted a spreading feeder ( 2 ). the spreading feeder ( 2 ) rotates a spring in a tube by a motor ( 21 ) mounted at the tube &# 39 ; s base end to spread the casting powder from its tip on to the molten steel surface of the mold . in place of such mechanical means , other transfer means such as pneumatic transfer means can also be used . in the drawings , ( 5 ) represents a tundish car with a tundish ( 6 ) mounted thereon , a strand nozzle ( 6a ) ( 7 ) a ladle ( 7 ). to the tip on the intermediate arm ( 32 ) of the multi - joint robot arm , a sensor ( 9 ) for monitoring the spreading conditions of the casting powder is arranged . this sensor can be an infrared sensor or thermal sensor , and is used for detecting the exposed molten steel ( hot spot ) in the mold ( 4 ). based on the detection by the sensor , the multi - joint robot arm ( 3 ) is moved under automatic control of a computer so as to move the tip of the spreading casting powder feeder ( 2 ) to the hot spot for spreading . it is also possible to move the robot arm ( 3 ) according to a predetermined program for spreading , not using such a sensor . the base end of the spreading feeder ( 2 ) and the discharge port in the final - stage powder container ( 1 ) are connected with a flexible transfer path ( 8 ). the flexible transfer path comprises a transfer path having the flexibility to follow the movement of the robot arm , such as flexible pipe . therefore , non - flexible pipe may be used only in sections of the path where flexibility is not required . the flexible transfer path ( 8 ) is arranged from the discharge port of the powder container ( 1 ) above the tip arm ( 12a ) and along the robot arm ( 3 ) to the spreading feeder ( 2 ). however , to simplify the drawing , the illustration of the part along the robot arm ( 3 ) is omitted . it is possible to equip a feeding device in the flexible transfer path ( 8 ) along with the tip arm ( 12a ). the casting powder is transferred by said feeding device and dropped by gravity from the final - stage container ( 1 ) to the spreading feeder ( 2 ). this gravity drop is based on the energy saving concept using the height difference between the final - stage powder container ( 1 ) arranged in a high position and the spreading feeder ( 2 ) placed in low position , but forced transfer means can be added . ( i ) when the tundish car ( 5 ) stops at the position above the mold ( 4 ), the revolving arm ( 12 ) swings to move the feeder from the stand - by position ( i ) to the feed position ( ii ). ( ii ) the tip arm ( 12a ) of the revolving arm turns to face the powder feeder toward the mold . ( iii ) the robot arm ( 3 ) moves the tip of the spreading feeder ( 2 ) above a hot spot to spread the casting powder from its tip . in fig2 the area designated ( b ) shows the spreading area . ( iv ) in replacing the tundish , the powder feeder is returned to the stand - by position by the reverse operation of steps ( i ) and ( ii ) mentioned above . the flexibility of the multi - joint robot arm ( 3 ) removes the dead angle above the mold surface , and prevents contact between the spreading feeder ( 2 ) and the tundish strand nozzle ( 6a ), when the powder feeder is moved . the present invention is not limited to the above embodiment . if there is a high position such as a deck near the workshop , the final - stage container can be mounted on the deck and the column and revolving arm can be eliminated . the final - stage powder container may be also hung from a high position . since the final - stage powder container is arranged in a position higher than the working space on the powder feeder according to the invention , provision of this feeder in the operator side does not interfere with the work of an operator . since the operator side is free of the scattered dust and molten steel of the side opposite the operator , the following effects are obtained : ( i ) sharp decrease in trouble due to scattered dust and molten steel ,; ( iii ) improved work environment permitting use of precision instruments such as the robot arm ,; ( v ) the smaller distance to the control board or the operation board making the anti - nozzle provision for cpu wiring easier . in spreading the casting powder on the mold , the flexibility of the robot arm removes the dead angle on the spreading surface of the mold , and the powder is uniformly spreaded all over the mold surface . and the precision instruments and control equipment used in the robot arm can continue good operation in the good work environment as described in the item ( iii ) mentioned above . the provision of the spreading conditions monitor sensor on the robot arm , makes complete automation of hot spot detection and spreading possible by computer control of the spreading . this promotes labor saving , stabilizes the continuous casting and improves quality . this invention can be used in full automation of continuous casting . in further progress of continuous casting of highgrade steel , a powder feeder in continuous casting has been provided which can cope with feed automation of highgrade steel billet size casting powder use ."}
|
{"category": "General tagging of new or cross-sectional technology", "patent": "with reference to the drawings , an embodiment of the present invention is described . the powder feeder comprises a final - stage powder container ( 1 ), a spreading feeder ( 2 ), and a multi - joint robot arm ( 3 ). the final - stage powder container ( 1 ) is equipped with a powder inlet line 13 and is mounted on the end of the revolving arm ( 12 ) held on the top of a column ( 11 ). final - stage powder container ( 1 ) is mounted at a position close to the base end of the tip arm ( 12a ). this revolving arm and tip arm are adequately driven by a driving device ( not shown ). here , the column ( 11 ) is taller than the height of an operator ( a ), thereby the revolving arm ( 12 ) and the final - stage powder container ( 1 ) are arranged above the working space of the mold . thus , the final - stage powder container ( 1 ) can be arranged on the operator side ( a ), for example , overhead of an operator . in fig2 two final - stage powder containers ( 1 )( 1 ) are arranged on the operator side . in the drawing , ( b ) indicates the side opposite an operator . a final - stage powder container ( 1 ) is provided with a meter such as a load cell platform scale ( 14 ) to weigh the spread quantity of casting powder . particularly , use of the &# 34 ; loss - in - weight &# 34 ; system permits recording of accurate spreading quantity and higher accuracy of control by a main computer of the continuous casting unit . the robot arm ( 3 ) comprises a base end arm ( 31 ), an intermediate arm ( 32 ), and a tip hand ( 33 ) movably connected by a first joint ( 34 ) and a second joint ( 35 ). the base end arm ( 31 ) is attached to the end of the tip arm ( 12a ) of the revolving arm . to the tip hand ( 33 ) is mounted a spreading feeder ( 2 ). the spreading feeder ( 2 ) rotates a spring in a tube by a motor ( 21 ) mounted at the tube &# 39 ; s base end to spread the casting powder from its tip on to the molten steel surface of the mold . in place of such mechanical means , other transfer means such as pneumatic transfer means can also be used . in the drawings , ( 5 ) represents a tundish car with a tundish ( 6 ) mounted thereon , a strand nozzle ( 6a ) ( 7 ) a ladle ( 7 ). to the tip on the intermediate arm ( 32 ) of the multi - joint robot arm , a sensor ( 9 ) for monitoring the spreading conditions of the casting powder is arranged . this sensor can be an infrared sensor or thermal sensor , and is used for detecting the exposed molten steel ( hot spot ) in the mold ( 4 ). based on the detection by the sensor , the multi - joint robot arm ( 3 ) is moved under automatic control of a computer so as to move the tip of the spreading casting powder feeder ( 2 ) to the hot spot for spreading . it is also possible to move the robot arm ( 3 ) according to a predetermined program for spreading , not using such a sensor . the base end of the spreading feeder ( 2 ) and the discharge port in the final - stage powder container ( 1 ) are connected with a flexible transfer path ( 8 ). the flexible transfer path comprises a transfer path having the flexibility to follow the movement of the robot arm , such as flexible pipe . therefore , non - flexible pipe may be used only in sections of the path where flexibility is not required . the flexible transfer path ( 8 ) is arranged from the discharge port of the powder container ( 1 ) above the tip arm ( 12a ) and along the robot arm ( 3 ) to the spreading feeder ( 2 ). however , to simplify the drawing , the illustration of the part along the robot arm ( 3 ) is omitted . it is possible to equip a feeding device in the flexible transfer path ( 8 ) along with the tip arm ( 12a ). the casting powder is transferred by said feeding device and dropped by gravity from the final - stage container ( 1 ) to the spreading feeder ( 2 ). this gravity drop is based on the energy saving concept using the height difference between the final - stage powder container ( 1 ) arranged in a high position and the spreading feeder ( 2 ) placed in low position , but forced transfer means can be added . ( i ) when the tundish car ( 5 ) stops at the position above the mold ( 4 ), the revolving arm ( 12 ) swings to move the feeder from the stand - by position ( i ) to the feed position ( ii ). ( ii ) the tip arm ( 12a ) of the revolving arm turns to face the powder feeder toward the mold . ( iii ) the robot arm ( 3 ) moves the tip of the spreading feeder ( 2 ) above a hot spot to spread the casting powder from its tip . in fig2 the area designated ( b ) shows the spreading area . ( iv ) in replacing the tundish , the powder feeder is returned to the stand - by position by the reverse operation of steps ( i ) and ( ii ) mentioned above . the flexibility of the multi - joint robot arm ( 3 ) removes the dead angle above the mold surface , and prevents contact between the spreading feeder ( 2 ) and the tundish strand nozzle ( 6a ), when the powder feeder is moved . the present invention is not limited to the above embodiment . if there is a high position such as a deck near the workshop , the final - stage container can be mounted on the deck and the column and revolving arm can be eliminated . the final - stage powder container may be also hung from a high position . since the final - stage powder container is arranged in a position higher than the working space on the powder feeder according to the invention , provision of this feeder in the operator side does not interfere with the work of an operator . since the operator side is free of the scattered dust and molten steel of the side opposite the operator , the following effects are obtained : ( i ) sharp decrease in trouble due to scattered dust and molten steel ,; ( iii ) improved work environment permitting use of precision instruments such as the robot arm ,; ( v ) the smaller distance to the control board or the operation board making the anti - nozzle provision for cpu wiring easier . in spreading the casting powder on the mold , the flexibility of the robot arm removes the dead angle on the spreading surface of the mold , and the powder is uniformly spreaded all over the mold surface . and the precision instruments and control equipment used in the robot arm can continue good operation in the good work environment as described in the item ( iii ) mentioned above . the provision of the spreading conditions monitor sensor on the robot arm , makes complete automation of hot spot detection and spreading possible by computer control of the spreading . this promotes labor saving , stabilizes the continuous casting and improves quality . this invention can be used in full automation of continuous casting . in further progress of continuous casting of highgrade steel , a powder feeder in continuous casting has been provided which can cope with feed automation of highgrade steel billet size casting powder use ."}
|
Is the categorization of this patent accurate?
| 0.25 |
49d27878671b44cf2401f85d19ab2921adb1331aa61116bc349d37d60692ad6f
| 0.063477 | 0.378906 | 0.078125 | 0.769531 | 0.441406 | 0.429688 |
null |
{"patent": "the preferred embodiments of the present invention will now be explained in detail with reference to the accompanying drawings . in the drawings , the same or similar elements are denoted by the same reference numerals even though they are depicted in different drawings . for the purpose of clarity , a detailed description of well known functions and configurations incorporated herein will be omitted as it may make the subject matter of the present invention unclear . fig2 is a block diagram illustrating the construction of a mobile radio communication system incorporating the present invention . referring to fig2 , a mobile communication exchange 120 performs an exchange function and interacts with another mobile communication , or the exchange 120 connects with a different communication network , such as a psdn , and controls the call termination / origination of a mobile terminal 100 . in addition , the mobile communication exchange 120 informs the charging center 130 of charging information of a telephone call , such as the subscriber &# 39 ; s number , terminating number , call start time , call termination time , discount information and so forth , when the call from the mobile terminal 100 terminates . the charging center 130 calculates and manages a telephone charge for the latest call , an accumulated telephone charge , and a total telephone charge using the charging information , then informs the respective telephone charges to the mobile communication exchange 120 . here , the management of the telephone charges is effected in such a manner that if thirty days elapses , the charging center 130 calculates the accumulated telephone charge for the calls made during the specific period for each mobile terminal subscriber , and notifies the respective mobile terminal subscriber of the total charge via mail . the bill includes a basic charge , a tax charge , and the calculated telephone charge . also , the mobile communication exchange 120 transmits to the mobile terminal 100 the telephone charging information , such as the telephone charge for the latest call , the accumulated telephone charge , and the total telephone charges that are transferred from the charging center 130 . the mobile terminal 100 receives the telephone charging information and displays the information on the display section ( not illustrated ), so that the mobile terminal subscriber can identify it . the mobile communication exchange 120 may be provided with a short message generating section 140 . this short message generating section 140 generates a short message which corresponds to the telephone charging information , such as the telephone charge for the latest call , the accumulated telephone charge , and the total telephone charge that are transferred from the charging center 130 in the form of a short message . the types of telephone charges transferred from the charging center 130 can be selectively determined according to the request by the mobile terminal subscriber . thus , different telephone charge information including the latest call , the accumulated telephone charge , and the total telephone charge can be provided to the subscriber . for instance , the mobile terminal subscriber can be provided with the accumulated telephone charge for a duration of a specified time period that he / she desires from the charging center 130 using the mobile terminal 100 . the type and the different telephone charge information transferred to the mobile terminal 100 may be predetermined or determined by the selection command from the mobile terminal subscriber . for example , the display unit of the mobile terminal can be pre - selected to display only the telephone charge for the latest call or only the total telephone charge , or both the latest call and the total telephone charge . fig3 is a flowchart illustrating a method of informing telephone charges to a mobile terminal for a telephone call and a total telephone charge call when the telephone call terminates according to the embodiment of the present invention . with reference to fig3 , the mobile communication exchange 120 detects whether the telephone call from the mobile terminal 100 terminates ( step 210 ). if the telephone call terminates , the mobile communication exchange 120 informs the charging center 130 of the charging information for the call , such as subscriber number , terminating number , call start time , call termination time , discount information , and so forth ( step 220 ). the charging center 130 calculates the charge for the latest call and adds the calculated charge to the total telephone charge ( step 230 ). the total telephone charge is calculated by adding up the accumulated telephone charge for each mobile terminal as well as the basic charge and the tax charge within a specified time period , for instance , a thirty - day period . the charging center 130 informs the mobile communication exchange 120 of the charge information for the latest call as well as the total charge ( step 240 ). the mobile communication exchange 120 receives the charge information for the latest call and the total charge , converts the received information in the form of a short message using the short message generating section 140 , and transmits the generated short message to the mobile terminal subscriber 100 ( steps 250 and 260 ). the mobile terminal 100 then receives the short message and displays the charge information of the latest call and the total charge on the lcd of the mobile phone . preferably , it may display characters such as \u201c charge for the latest call : 400 won , total charge : 10 , 500 won \u201d. accordingly , the mobile terminal subscriber can immediately confirm the charge for the latest call as well as the total charge calculated to include the latest call fig4 is a flowchart illustrating a method of informing a mobile terminal of the total telephone charge in the case that the mobile terminal requests confirmation of the total telephone charges up to now in a standby state according to another embodiment of the present invention . with reference to fig4 , if the mobile terminal subscriber intends to verify the total telephone charge up to now in a standby state , he / she can request for the total charge , for example , by sequentially pressing keys \u201c*\u201d, \u201c 1 \u201d, \u201c 1 \u201d, and \u201c send \u201d, or a specified key ( step 310 ). if the mobile communication exchange 120 receives the request for the confirmation of the total charge for the mobile terminal 100 through the base station 110 , the exchange 120 in turn requests the confirmation of the total charge of the mobile terminal subscriber to the charging center 130 ( step 320 ). the charging center 130 searches the total charging information corresponding to the requesting mobile terminal subscriber and informs the mobile communication exchange 120 of the total charging information ( step 330 ). the mobile communication exchange 120 generates a short message corresponding to the total charging information through the short message generating section 140 and transmits the short message to the mobile terminal 100 ( steps 340 and 350 ). the mobile terminal 100 receives the short message , and displays the total charge on the lcd . preferably , it may display characters such as \u201c total charge : 10 , 500 won ( i . e . korean currency )\u201d. as described above , it will be apparent that the present invention provides advantages in that the mobile terminal subscriber can be provided with an accurate telephone charge information by enabling the subscriber to immediately verify the charging information calculated by the charging center through the mobile terminal just after a telephone call . also , the mobile terminal subscriber can selectively verify at least one type of charging information out of various charging information managed by the charging center as occasion demands using the mobile terminal . while this invention has been described in connection with what is presently considered to be the most practical and preferred embodiments which display both the charge for the latest call and the total charge or only the total charge , it is to be understood that other modifications thereof may be made without departing from the scope of the invention . for example , the accumulated telephone charge for a specified time period desired by the mobile terminal subscriber , or another type of charge information other than the latest call and the total charge can be selectively displayed . thus , the present invention should not be limited to the disclosed embodiment but should be defined by the scope of the appended claims and their equivalents .", "category": "Electricity"}
|
{"patent": "the preferred embodiments of the present invention will now be explained in detail with reference to the accompanying drawings . in the drawings , the same or similar elements are denoted by the same reference numerals even though they are depicted in different drawings . for the purpose of clarity , a detailed description of well known functions and configurations incorporated herein will be omitted as it may make the subject matter of the present invention unclear . fig2 is a block diagram illustrating the construction of a mobile radio communication system incorporating the present invention . referring to fig2 , a mobile communication exchange 120 performs an exchange function and interacts with another mobile communication , or the exchange 120 connects with a different communication network , such as a psdn , and controls the call termination / origination of a mobile terminal 100 . in addition , the mobile communication exchange 120 informs the charging center 130 of charging information of a telephone call , such as the subscriber &# 39 ; s number , terminating number , call start time , call termination time , discount information and so forth , when the call from the mobile terminal 100 terminates . the charging center 130 calculates and manages a telephone charge for the latest call , an accumulated telephone charge , and a total telephone charge using the charging information , then informs the respective telephone charges to the mobile communication exchange 120 . here , the management of the telephone charges is effected in such a manner that if thirty days elapses , the charging center 130 calculates the accumulated telephone charge for the calls made during the specific period for each mobile terminal subscriber , and notifies the respective mobile terminal subscriber of the total charge via mail . the bill includes a basic charge , a tax charge , and the calculated telephone charge . also , the mobile communication exchange 120 transmits to the mobile terminal 100 the telephone charging information , such as the telephone charge for the latest call , the accumulated telephone charge , and the total telephone charges that are transferred from the charging center 130 . the mobile terminal 100 receives the telephone charging information and displays the information on the display section ( not illustrated ), so that the mobile terminal subscriber can identify it . the mobile communication exchange 120 may be provided with a short message generating section 140 . this short message generating section 140 generates a short message which corresponds to the telephone charging information , such as the telephone charge for the latest call , the accumulated telephone charge , and the total telephone charge that are transferred from the charging center 130 in the form of a short message . the types of telephone charges transferred from the charging center 130 can be selectively determined according to the request by the mobile terminal subscriber . thus , different telephone charge information including the latest call , the accumulated telephone charge , and the total telephone charge can be provided to the subscriber . for instance , the mobile terminal subscriber can be provided with the accumulated telephone charge for a duration of a specified time period that he / she desires from the charging center 130 using the mobile terminal 100 . the type and the different telephone charge information transferred to the mobile terminal 100 may be predetermined or determined by the selection command from the mobile terminal subscriber . for example , the display unit of the mobile terminal can be pre - selected to display only the telephone charge for the latest call or only the total telephone charge , or both the latest call and the total telephone charge . fig3 is a flowchart illustrating a method of informing telephone charges to a mobile terminal for a telephone call and a total telephone charge call when the telephone call terminates according to the embodiment of the present invention . with reference to fig3 , the mobile communication exchange 120 detects whether the telephone call from the mobile terminal 100 terminates ( step 210 ). if the telephone call terminates , the mobile communication exchange 120 informs the charging center 130 of the charging information for the call , such as subscriber number , terminating number , call start time , call termination time , discount information , and so forth ( step 220 ). the charging center 130 calculates the charge for the latest call and adds the calculated charge to the total telephone charge ( step 230 ). the total telephone charge is calculated by adding up the accumulated telephone charge for each mobile terminal as well as the basic charge and the tax charge within a specified time period , for instance , a thirty - day period . the charging center 130 informs the mobile communication exchange 120 of the charge information for the latest call as well as the total charge ( step 240 ). the mobile communication exchange 120 receives the charge information for the latest call and the total charge , converts the received information in the form of a short message using the short message generating section 140 , and transmits the generated short message to the mobile terminal subscriber 100 ( steps 250 and 260 ). the mobile terminal 100 then receives the short message and displays the charge information of the latest call and the total charge on the lcd of the mobile phone . preferably , it may display characters such as \u201c charge for the latest call : 400 won , total charge : 10 , 500 won \u201d. accordingly , the mobile terminal subscriber can immediately confirm the charge for the latest call as well as the total charge calculated to include the latest call fig4 is a flowchart illustrating a method of informing a mobile terminal of the total telephone charge in the case that the mobile terminal requests confirmation of the total telephone charges up to now in a standby state according to another embodiment of the present invention . with reference to fig4 , if the mobile terminal subscriber intends to verify the total telephone charge up to now in a standby state , he / she can request for the total charge , for example , by sequentially pressing keys \u201c*\u201d, \u201c 1 \u201d, \u201c 1 \u201d, and \u201c send \u201d, or a specified key ( step 310 ). if the mobile communication exchange 120 receives the request for the confirmation of the total charge for the mobile terminal 100 through the base station 110 , the exchange 120 in turn requests the confirmation of the total charge of the mobile terminal subscriber to the charging center 130 ( step 320 ). the charging center 130 searches the total charging information corresponding to the requesting mobile terminal subscriber and informs the mobile communication exchange 120 of the total charging information ( step 330 ). the mobile communication exchange 120 generates a short message corresponding to the total charging information through the short message generating section 140 and transmits the short message to the mobile terminal 100 ( steps 340 and 350 ). the mobile terminal 100 receives the short message , and displays the total charge on the lcd . preferably , it may display characters such as \u201c total charge : 10 , 500 won ( i . e . korean currency )\u201d. as described above , it will be apparent that the present invention provides advantages in that the mobile terminal subscriber can be provided with an accurate telephone charge information by enabling the subscriber to immediately verify the charging information calculated by the charging center through the mobile terminal just after a telephone call . also , the mobile terminal subscriber can selectively verify at least one type of charging information out of various charging information managed by the charging center as occasion demands using the mobile terminal . while this invention has been described in connection with what is presently considered to be the most practical and preferred embodiments which display both the charge for the latest call and the total charge or only the total charge , it is to be understood that other modifications thereof may be made without departing from the scope of the invention . for example , the accumulated telephone charge for a specified time period desired by the mobile terminal subscriber , or another type of charge information other than the latest call and the total charge can be selectively displayed . thus , the present invention should not be limited to the disclosed embodiment but should be defined by the scope of the appended claims and their equivalents .", "category": "Human Necessities"}
|
Is the categorization of this patent accurate?
| 0.25 |
c54693020cd566de34695abc4d1ab9b0bb7c2dfdf03740340864fefaee1a46f1
| 0.117676 | 0.00014 | 0.034668 | 0.001282 | 0.068359 | 0.000315 |
null |
{"patent": "the preferred embodiments of the present invention will now be explained in detail with reference to the accompanying drawings . in the drawings , the same or similar elements are denoted by the same reference numerals even though they are depicted in different drawings . for the purpose of clarity , a detailed description of well known functions and configurations incorporated herein will be omitted as it may make the subject matter of the present invention unclear . fig2 is a block diagram illustrating the construction of a mobile radio communication system incorporating the present invention . referring to fig2 , a mobile communication exchange 120 performs an exchange function and interacts with another mobile communication , or the exchange 120 connects with a different communication network , such as a psdn , and controls the call termination / origination of a mobile terminal 100 . in addition , the mobile communication exchange 120 informs the charging center 130 of charging information of a telephone call , such as the subscriber &# 39 ; s number , terminating number , call start time , call termination time , discount information and so forth , when the call from the mobile terminal 100 terminates . the charging center 130 calculates and manages a telephone charge for the latest call , an accumulated telephone charge , and a total telephone charge using the charging information , then informs the respective telephone charges to the mobile communication exchange 120 . here , the management of the telephone charges is effected in such a manner that if thirty days elapses , the charging center 130 calculates the accumulated telephone charge for the calls made during the specific period for each mobile terminal subscriber , and notifies the respective mobile terminal subscriber of the total charge via mail . the bill includes a basic charge , a tax charge , and the calculated telephone charge . also , the mobile communication exchange 120 transmits to the mobile terminal 100 the telephone charging information , such as the telephone charge for the latest call , the accumulated telephone charge , and the total telephone charges that are transferred from the charging center 130 . the mobile terminal 100 receives the telephone charging information and displays the information on the display section ( not illustrated ), so that the mobile terminal subscriber can identify it . the mobile communication exchange 120 may be provided with a short message generating section 140 . this short message generating section 140 generates a short message which corresponds to the telephone charging information , such as the telephone charge for the latest call , the accumulated telephone charge , and the total telephone charge that are transferred from the charging center 130 in the form of a short message . the types of telephone charges transferred from the charging center 130 can be selectively determined according to the request by the mobile terminal subscriber . thus , different telephone charge information including the latest call , the accumulated telephone charge , and the total telephone charge can be provided to the subscriber . for instance , the mobile terminal subscriber can be provided with the accumulated telephone charge for a duration of a specified time period that he / she desires from the charging center 130 using the mobile terminal 100 . the type and the different telephone charge information transferred to the mobile terminal 100 may be predetermined or determined by the selection command from the mobile terminal subscriber . for example , the display unit of the mobile terminal can be pre - selected to display only the telephone charge for the latest call or only the total telephone charge , or both the latest call and the total telephone charge . fig3 is a flowchart illustrating a method of informing telephone charges to a mobile terminal for a telephone call and a total telephone charge call when the telephone call terminates according to the embodiment of the present invention . with reference to fig3 , the mobile communication exchange 120 detects whether the telephone call from the mobile terminal 100 terminates ( step 210 ). if the telephone call terminates , the mobile communication exchange 120 informs the charging center 130 of the charging information for the call , such as subscriber number , terminating number , call start time , call termination time , discount information , and so forth ( step 220 ). the charging center 130 calculates the charge for the latest call and adds the calculated charge to the total telephone charge ( step 230 ). the total telephone charge is calculated by adding up the accumulated telephone charge for each mobile terminal as well as the basic charge and the tax charge within a specified time period , for instance , a thirty - day period . the charging center 130 informs the mobile communication exchange 120 of the charge information for the latest call as well as the total charge ( step 240 ). the mobile communication exchange 120 receives the charge information for the latest call and the total charge , converts the received information in the form of a short message using the short message generating section 140 , and transmits the generated short message to the mobile terminal subscriber 100 ( steps 250 and 260 ). the mobile terminal 100 then receives the short message and displays the charge information of the latest call and the total charge on the lcd of the mobile phone . preferably , it may display characters such as \u201c charge for the latest call : 400 won , total charge : 10 , 500 won \u201d. accordingly , the mobile terminal subscriber can immediately confirm the charge for the latest call as well as the total charge calculated to include the latest call fig4 is a flowchart illustrating a method of informing a mobile terminal of the total telephone charge in the case that the mobile terminal requests confirmation of the total telephone charges up to now in a standby state according to another embodiment of the present invention . with reference to fig4 , if the mobile terminal subscriber intends to verify the total telephone charge up to now in a standby state , he / she can request for the total charge , for example , by sequentially pressing keys \u201c*\u201d, \u201c 1 \u201d, \u201c 1 \u201d, and \u201c send \u201d, or a specified key ( step 310 ). if the mobile communication exchange 120 receives the request for the confirmation of the total charge for the mobile terminal 100 through the base station 110 , the exchange 120 in turn requests the confirmation of the total charge of the mobile terminal subscriber to the charging center 130 ( step 320 ). the charging center 130 searches the total charging information corresponding to the requesting mobile terminal subscriber and informs the mobile communication exchange 120 of the total charging information ( step 330 ). the mobile communication exchange 120 generates a short message corresponding to the total charging information through the short message generating section 140 and transmits the short message to the mobile terminal 100 ( steps 340 and 350 ). the mobile terminal 100 receives the short message , and displays the total charge on the lcd . preferably , it may display characters such as \u201c total charge : 10 , 500 won ( i . e . korean currency )\u201d. as described above , it will be apparent that the present invention provides advantages in that the mobile terminal subscriber can be provided with an accurate telephone charge information by enabling the subscriber to immediately verify the charging information calculated by the charging center through the mobile terminal just after a telephone call . also , the mobile terminal subscriber can selectively verify at least one type of charging information out of various charging information managed by the charging center as occasion demands using the mobile terminal . while this invention has been described in connection with what is presently considered to be the most practical and preferred embodiments which display both the charge for the latest call and the total charge or only the total charge , it is to be understood that other modifications thereof may be made without departing from the scope of the invention . for example , the accumulated telephone charge for a specified time period desired by the mobile terminal subscriber , or another type of charge information other than the latest call and the total charge can be selectively displayed . thus , the present invention should not be limited to the disclosed embodiment but should be defined by the scope of the appended claims and their equivalents .", "category": "Electricity"}
|
{"category": "Performing Operations; Transporting", "patent": "the preferred embodiments of the present invention will now be explained in detail with reference to the accompanying drawings . in the drawings , the same or similar elements are denoted by the same reference numerals even though they are depicted in different drawings . for the purpose of clarity , a detailed description of well known functions and configurations incorporated herein will be omitted as it may make the subject matter of the present invention unclear . fig2 is a block diagram illustrating the construction of a mobile radio communication system incorporating the present invention . referring to fig2 , a mobile communication exchange 120 performs an exchange function and interacts with another mobile communication , or the exchange 120 connects with a different communication network , such as a psdn , and controls the call termination / origination of a mobile terminal 100 . in addition , the mobile communication exchange 120 informs the charging center 130 of charging information of a telephone call , such as the subscriber &# 39 ; s number , terminating number , call start time , call termination time , discount information and so forth , when the call from the mobile terminal 100 terminates . the charging center 130 calculates and manages a telephone charge for the latest call , an accumulated telephone charge , and a total telephone charge using the charging information , then informs the respective telephone charges to the mobile communication exchange 120 . here , the management of the telephone charges is effected in such a manner that if thirty days elapses , the charging center 130 calculates the accumulated telephone charge for the calls made during the specific period for each mobile terminal subscriber , and notifies the respective mobile terminal subscriber of the total charge via mail . the bill includes a basic charge , a tax charge , and the calculated telephone charge . also , the mobile communication exchange 120 transmits to the mobile terminal 100 the telephone charging information , such as the telephone charge for the latest call , the accumulated telephone charge , and the total telephone charges that are transferred from the charging center 130 . the mobile terminal 100 receives the telephone charging information and displays the information on the display section ( not illustrated ), so that the mobile terminal subscriber can identify it . the mobile communication exchange 120 may be provided with a short message generating section 140 . this short message generating section 140 generates a short message which corresponds to the telephone charging information , such as the telephone charge for the latest call , the accumulated telephone charge , and the total telephone charge that are transferred from the charging center 130 in the form of a short message . the types of telephone charges transferred from the charging center 130 can be selectively determined according to the request by the mobile terminal subscriber . thus , different telephone charge information including the latest call , the accumulated telephone charge , and the total telephone charge can be provided to the subscriber . for instance , the mobile terminal subscriber can be provided with the accumulated telephone charge for a duration of a specified time period that he / she desires from the charging center 130 using the mobile terminal 100 . the type and the different telephone charge information transferred to the mobile terminal 100 may be predetermined or determined by the selection command from the mobile terminal subscriber . for example , the display unit of the mobile terminal can be pre - selected to display only the telephone charge for the latest call or only the total telephone charge , or both the latest call and the total telephone charge . fig3 is a flowchart illustrating a method of informing telephone charges to a mobile terminal for a telephone call and a total telephone charge call when the telephone call terminates according to the embodiment of the present invention . with reference to fig3 , the mobile communication exchange 120 detects whether the telephone call from the mobile terminal 100 terminates ( step 210 ). if the telephone call terminates , the mobile communication exchange 120 informs the charging center 130 of the charging information for the call , such as subscriber number , terminating number , call start time , call termination time , discount information , and so forth ( step 220 ). the charging center 130 calculates the charge for the latest call and adds the calculated charge to the total telephone charge ( step 230 ). the total telephone charge is calculated by adding up the accumulated telephone charge for each mobile terminal as well as the basic charge and the tax charge within a specified time period , for instance , a thirty - day period . the charging center 130 informs the mobile communication exchange 120 of the charge information for the latest call as well as the total charge ( step 240 ). the mobile communication exchange 120 receives the charge information for the latest call and the total charge , converts the received information in the form of a short message using the short message generating section 140 , and transmits the generated short message to the mobile terminal subscriber 100 ( steps 250 and 260 ). the mobile terminal 100 then receives the short message and displays the charge information of the latest call and the total charge on the lcd of the mobile phone . preferably , it may display characters such as \u201c charge for the latest call : 400 won , total charge : 10 , 500 won \u201d. accordingly , the mobile terminal subscriber can immediately confirm the charge for the latest call as well as the total charge calculated to include the latest call fig4 is a flowchart illustrating a method of informing a mobile terminal of the total telephone charge in the case that the mobile terminal requests confirmation of the total telephone charges up to now in a standby state according to another embodiment of the present invention . with reference to fig4 , if the mobile terminal subscriber intends to verify the total telephone charge up to now in a standby state , he / she can request for the total charge , for example , by sequentially pressing keys \u201c*\u201d, \u201c 1 \u201d, \u201c 1 \u201d, and \u201c send \u201d, or a specified key ( step 310 ). if the mobile communication exchange 120 receives the request for the confirmation of the total charge for the mobile terminal 100 through the base station 110 , the exchange 120 in turn requests the confirmation of the total charge of the mobile terminal subscriber to the charging center 130 ( step 320 ). the charging center 130 searches the total charging information corresponding to the requesting mobile terminal subscriber and informs the mobile communication exchange 120 of the total charging information ( step 330 ). the mobile communication exchange 120 generates a short message corresponding to the total charging information through the short message generating section 140 and transmits the short message to the mobile terminal 100 ( steps 340 and 350 ). the mobile terminal 100 receives the short message , and displays the total charge on the lcd . preferably , it may display characters such as \u201c total charge : 10 , 500 won ( i . e . korean currency )\u201d. as described above , it will be apparent that the present invention provides advantages in that the mobile terminal subscriber can be provided with an accurate telephone charge information by enabling the subscriber to immediately verify the charging information calculated by the charging center through the mobile terminal just after a telephone call . also , the mobile terminal subscriber can selectively verify at least one type of charging information out of various charging information managed by the charging center as occasion demands using the mobile terminal . while this invention has been described in connection with what is presently considered to be the most practical and preferred embodiments which display both the charge for the latest call and the total charge or only the total charge , it is to be understood that other modifications thereof may be made without departing from the scope of the invention . for example , the accumulated telephone charge for a specified time period desired by the mobile terminal subscriber , or another type of charge information other than the latest call and the total charge can be selectively displayed . thus , the present invention should not be limited to the disclosed embodiment but should be defined by the scope of the appended claims and their equivalents ."}
|
Is the category the most suitable category for the given patent?
| 0.25 |
c54693020cd566de34695abc4d1ab9b0bb7c2dfdf03740340864fefaee1a46f1
| 0.613281 | 0.038574 | 0.017944 | 0.006683 | 0.069336 | 0.21582 |
null |
{"patent": "the preferred embodiments of the present invention will now be explained in detail with reference to the accompanying drawings . in the drawings , the same or similar elements are denoted by the same reference numerals even though they are depicted in different drawings . for the purpose of clarity , a detailed description of well known functions and configurations incorporated herein will be omitted as it may make the subject matter of the present invention unclear . fig2 is a block diagram illustrating the construction of a mobile radio communication system incorporating the present invention . referring to fig2 , a mobile communication exchange 120 performs an exchange function and interacts with another mobile communication , or the exchange 120 connects with a different communication network , such as a psdn , and controls the call termination / origination of a mobile terminal 100 . in addition , the mobile communication exchange 120 informs the charging center 130 of charging information of a telephone call , such as the subscriber &# 39 ; s number , terminating number , call start time , call termination time , discount information and so forth , when the call from the mobile terminal 100 terminates . the charging center 130 calculates and manages a telephone charge for the latest call , an accumulated telephone charge , and a total telephone charge using the charging information , then informs the respective telephone charges to the mobile communication exchange 120 . here , the management of the telephone charges is effected in such a manner that if thirty days elapses , the charging center 130 calculates the accumulated telephone charge for the calls made during the specific period for each mobile terminal subscriber , and notifies the respective mobile terminal subscriber of the total charge via mail . the bill includes a basic charge , a tax charge , and the calculated telephone charge . also , the mobile communication exchange 120 transmits to the mobile terminal 100 the telephone charging information , such as the telephone charge for the latest call , the accumulated telephone charge , and the total telephone charges that are transferred from the charging center 130 . the mobile terminal 100 receives the telephone charging information and displays the information on the display section ( not illustrated ), so that the mobile terminal subscriber can identify it . the mobile communication exchange 120 may be provided with a short message generating section 140 . this short message generating section 140 generates a short message which corresponds to the telephone charging information , such as the telephone charge for the latest call , the accumulated telephone charge , and the total telephone charge that are transferred from the charging center 130 in the form of a short message . the types of telephone charges transferred from the charging center 130 can be selectively determined according to the request by the mobile terminal subscriber . thus , different telephone charge information including the latest call , the accumulated telephone charge , and the total telephone charge can be provided to the subscriber . for instance , the mobile terminal subscriber can be provided with the accumulated telephone charge for a duration of a specified time period that he / she desires from the charging center 130 using the mobile terminal 100 . the type and the different telephone charge information transferred to the mobile terminal 100 may be predetermined or determined by the selection command from the mobile terminal subscriber . for example , the display unit of the mobile terminal can be pre - selected to display only the telephone charge for the latest call or only the total telephone charge , or both the latest call and the total telephone charge . fig3 is a flowchart illustrating a method of informing telephone charges to a mobile terminal for a telephone call and a total telephone charge call when the telephone call terminates according to the embodiment of the present invention . with reference to fig3 , the mobile communication exchange 120 detects whether the telephone call from the mobile terminal 100 terminates ( step 210 ). if the telephone call terminates , the mobile communication exchange 120 informs the charging center 130 of the charging information for the call , such as subscriber number , terminating number , call start time , call termination time , discount information , and so forth ( step 220 ). the charging center 130 calculates the charge for the latest call and adds the calculated charge to the total telephone charge ( step 230 ). the total telephone charge is calculated by adding up the accumulated telephone charge for each mobile terminal as well as the basic charge and the tax charge within a specified time period , for instance , a thirty - day period . the charging center 130 informs the mobile communication exchange 120 of the charge information for the latest call as well as the total charge ( step 240 ). the mobile communication exchange 120 receives the charge information for the latest call and the total charge , converts the received information in the form of a short message using the short message generating section 140 , and transmits the generated short message to the mobile terminal subscriber 100 ( steps 250 and 260 ). the mobile terminal 100 then receives the short message and displays the charge information of the latest call and the total charge on the lcd of the mobile phone . preferably , it may display characters such as \u201c charge for the latest call : 400 won , total charge : 10 , 500 won \u201d. accordingly , the mobile terminal subscriber can immediately confirm the charge for the latest call as well as the total charge calculated to include the latest call fig4 is a flowchart illustrating a method of informing a mobile terminal of the total telephone charge in the case that the mobile terminal requests confirmation of the total telephone charges up to now in a standby state according to another embodiment of the present invention . with reference to fig4 , if the mobile terminal subscriber intends to verify the total telephone charge up to now in a standby state , he / she can request for the total charge , for example , by sequentially pressing keys \u201c*\u201d, \u201c 1 \u201d, \u201c 1 \u201d, and \u201c send \u201d, or a specified key ( step 310 ). if the mobile communication exchange 120 receives the request for the confirmation of the total charge for the mobile terminal 100 through the base station 110 , the exchange 120 in turn requests the confirmation of the total charge of the mobile terminal subscriber to the charging center 130 ( step 320 ). the charging center 130 searches the total charging information corresponding to the requesting mobile terminal subscriber and informs the mobile communication exchange 120 of the total charging information ( step 330 ). the mobile communication exchange 120 generates a short message corresponding to the total charging information through the short message generating section 140 and transmits the short message to the mobile terminal 100 ( steps 340 and 350 ). the mobile terminal 100 receives the short message , and displays the total charge on the lcd . preferably , it may display characters such as \u201c total charge : 10 , 500 won ( i . e . korean currency )\u201d. as described above , it will be apparent that the present invention provides advantages in that the mobile terminal subscriber can be provided with an accurate telephone charge information by enabling the subscriber to immediately verify the charging information calculated by the charging center through the mobile terminal just after a telephone call . also , the mobile terminal subscriber can selectively verify at least one type of charging information out of various charging information managed by the charging center as occasion demands using the mobile terminal . while this invention has been described in connection with what is presently considered to be the most practical and preferred embodiments which display both the charge for the latest call and the total charge or only the total charge , it is to be understood that other modifications thereof may be made without departing from the scope of the invention . for example , the accumulated telephone charge for a specified time period desired by the mobile terminal subscriber , or another type of charge information other than the latest call and the total charge can be selectively displayed . thus , the present invention should not be limited to the disclosed embodiment but should be defined by the scope of the appended claims and their equivalents .", "category": "Electricity"}
|
{"patent": "the preferred embodiments of the present invention will now be explained in detail with reference to the accompanying drawings . in the drawings , the same or similar elements are denoted by the same reference numerals even though they are depicted in different drawings . for the purpose of clarity , a detailed description of well known functions and configurations incorporated herein will be omitted as it may make the subject matter of the present invention unclear . fig2 is a block diagram illustrating the construction of a mobile radio communication system incorporating the present invention . referring to fig2 , a mobile communication exchange 120 performs an exchange function and interacts with another mobile communication , or the exchange 120 connects with a different communication network , such as a psdn , and controls the call termination / origination of a mobile terminal 100 . in addition , the mobile communication exchange 120 informs the charging center 130 of charging information of a telephone call , such as the subscriber &# 39 ; s number , terminating number , call start time , call termination time , discount information and so forth , when the call from the mobile terminal 100 terminates . the charging center 130 calculates and manages a telephone charge for the latest call , an accumulated telephone charge , and a total telephone charge using the charging information , then informs the respective telephone charges to the mobile communication exchange 120 . here , the management of the telephone charges is effected in such a manner that if thirty days elapses , the charging center 130 calculates the accumulated telephone charge for the calls made during the specific period for each mobile terminal subscriber , and notifies the respective mobile terminal subscriber of the total charge via mail . the bill includes a basic charge , a tax charge , and the calculated telephone charge . also , the mobile communication exchange 120 transmits to the mobile terminal 100 the telephone charging information , such as the telephone charge for the latest call , the accumulated telephone charge , and the total telephone charges that are transferred from the charging center 130 . the mobile terminal 100 receives the telephone charging information and displays the information on the display section ( not illustrated ), so that the mobile terminal subscriber can identify it . the mobile communication exchange 120 may be provided with a short message generating section 140 . this short message generating section 140 generates a short message which corresponds to the telephone charging information , such as the telephone charge for the latest call , the accumulated telephone charge , and the total telephone charge that are transferred from the charging center 130 in the form of a short message . the types of telephone charges transferred from the charging center 130 can be selectively determined according to the request by the mobile terminal subscriber . thus , different telephone charge information including the latest call , the accumulated telephone charge , and the total telephone charge can be provided to the subscriber . for instance , the mobile terminal subscriber can be provided with the accumulated telephone charge for a duration of a specified time period that he / she desires from the charging center 130 using the mobile terminal 100 . the type and the different telephone charge information transferred to the mobile terminal 100 may be predetermined or determined by the selection command from the mobile terminal subscriber . for example , the display unit of the mobile terminal can be pre - selected to display only the telephone charge for the latest call or only the total telephone charge , or both the latest call and the total telephone charge . fig3 is a flowchart illustrating a method of informing telephone charges to a mobile terminal for a telephone call and a total telephone charge call when the telephone call terminates according to the embodiment of the present invention . with reference to fig3 , the mobile communication exchange 120 detects whether the telephone call from the mobile terminal 100 terminates ( step 210 ). if the telephone call terminates , the mobile communication exchange 120 informs the charging center 130 of the charging information for the call , such as subscriber number , terminating number , call start time , call termination time , discount information , and so forth ( step 220 ). the charging center 130 calculates the charge for the latest call and adds the calculated charge to the total telephone charge ( step 230 ). the total telephone charge is calculated by adding up the accumulated telephone charge for each mobile terminal as well as the basic charge and the tax charge within a specified time period , for instance , a thirty - day period . the charging center 130 informs the mobile communication exchange 120 of the charge information for the latest call as well as the total charge ( step 240 ). the mobile communication exchange 120 receives the charge information for the latest call and the total charge , converts the received information in the form of a short message using the short message generating section 140 , and transmits the generated short message to the mobile terminal subscriber 100 ( steps 250 and 260 ). the mobile terminal 100 then receives the short message and displays the charge information of the latest call and the total charge on the lcd of the mobile phone . preferably , it may display characters such as \u201c charge for the latest call : 400 won , total charge : 10 , 500 won \u201d. accordingly , the mobile terminal subscriber can immediately confirm the charge for the latest call as well as the total charge calculated to include the latest call fig4 is a flowchart illustrating a method of informing a mobile terminal of the total telephone charge in the case that the mobile terminal requests confirmation of the total telephone charges up to now in a standby state according to another embodiment of the present invention . with reference to fig4 , if the mobile terminal subscriber intends to verify the total telephone charge up to now in a standby state , he / she can request for the total charge , for example , by sequentially pressing keys \u201c*\u201d, \u201c 1 \u201d, \u201c 1 \u201d, and \u201c send \u201d, or a specified key ( step 310 ). if the mobile communication exchange 120 receives the request for the confirmation of the total charge for the mobile terminal 100 through the base station 110 , the exchange 120 in turn requests the confirmation of the total charge of the mobile terminal subscriber to the charging center 130 ( step 320 ). the charging center 130 searches the total charging information corresponding to the requesting mobile terminal subscriber and informs the mobile communication exchange 120 of the total charging information ( step 330 ). the mobile communication exchange 120 generates a short message corresponding to the total charging information through the short message generating section 140 and transmits the short message to the mobile terminal 100 ( steps 340 and 350 ). the mobile terminal 100 receives the short message , and displays the total charge on the lcd . preferably , it may display characters such as \u201c total charge : 10 , 500 won ( i . e . korean currency )\u201d. as described above , it will be apparent that the present invention provides advantages in that the mobile terminal subscriber can be provided with an accurate telephone charge information by enabling the subscriber to immediately verify the charging information calculated by the charging center through the mobile terminal just after a telephone call . also , the mobile terminal subscriber can selectively verify at least one type of charging information out of various charging information managed by the charging center as occasion demands using the mobile terminal . while this invention has been described in connection with what is presently considered to be the most practical and preferred embodiments which display both the charge for the latest call and the total charge or only the total charge , it is to be understood that other modifications thereof may be made without departing from the scope of the invention . for example , the accumulated telephone charge for a specified time period desired by the mobile terminal subscriber , or another type of charge information other than the latest call and the total charge can be selectively displayed . thus , the present invention should not be limited to the disclosed embodiment but should be defined by the scope of the appended claims and their equivalents .", "category": "Chemistry; Metallurgy"}
|
Is the patent correctly categorized?
| 0.25 |
c54693020cd566de34695abc4d1ab9b0bb7c2dfdf03740340864fefaee1a46f1
| 0.060059 | 0.00014 | 0.072754 | 0.00592 | 0.063477 | 0.000345 |
null |
{"category": "Electricity", "patent": "the preferred embodiments of the present invention will now be explained in detail with reference to the accompanying drawings . in the drawings , the same or similar elements are denoted by the same reference numerals even though they are depicted in different drawings . for the purpose of clarity , a detailed description of well known functions and configurations incorporated herein will be omitted as it may make the subject matter of the present invention unclear . fig2 is a block diagram illustrating the construction of a mobile radio communication system incorporating the present invention . referring to fig2 , a mobile communication exchange 120 performs an exchange function and interacts with another mobile communication , or the exchange 120 connects with a different communication network , such as a psdn , and controls the call termination / origination of a mobile terminal 100 . in addition , the mobile communication exchange 120 informs the charging center 130 of charging information of a telephone call , such as the subscriber &# 39 ; s number , terminating number , call start time , call termination time , discount information and so forth , when the call from the mobile terminal 100 terminates . the charging center 130 calculates and manages a telephone charge for the latest call , an accumulated telephone charge , and a total telephone charge using the charging information , then informs the respective telephone charges to the mobile communication exchange 120 . here , the management of the telephone charges is effected in such a manner that if thirty days elapses , the charging center 130 calculates the accumulated telephone charge for the calls made during the specific period for each mobile terminal subscriber , and notifies the respective mobile terminal subscriber of the total charge via mail . the bill includes a basic charge , a tax charge , and the calculated telephone charge . also , the mobile communication exchange 120 transmits to the mobile terminal 100 the telephone charging information , such as the telephone charge for the latest call , the accumulated telephone charge , and the total telephone charges that are transferred from the charging center 130 . the mobile terminal 100 receives the telephone charging information and displays the information on the display section ( not illustrated ), so that the mobile terminal subscriber can identify it . the mobile communication exchange 120 may be provided with a short message generating section 140 . this short message generating section 140 generates a short message which corresponds to the telephone charging information , such as the telephone charge for the latest call , the accumulated telephone charge , and the total telephone charge that are transferred from the charging center 130 in the form of a short message . the types of telephone charges transferred from the charging center 130 can be selectively determined according to the request by the mobile terminal subscriber . thus , different telephone charge information including the latest call , the accumulated telephone charge , and the total telephone charge can be provided to the subscriber . for instance , the mobile terminal subscriber can be provided with the accumulated telephone charge for a duration of a specified time period that he / she desires from the charging center 130 using the mobile terminal 100 . the type and the different telephone charge information transferred to the mobile terminal 100 may be predetermined or determined by the selection command from the mobile terminal subscriber . for example , the display unit of the mobile terminal can be pre - selected to display only the telephone charge for the latest call or only the total telephone charge , or both the latest call and the total telephone charge . fig3 is a flowchart illustrating a method of informing telephone charges to a mobile terminal for a telephone call and a total telephone charge call when the telephone call terminates according to the embodiment of the present invention . with reference to fig3 , the mobile communication exchange 120 detects whether the telephone call from the mobile terminal 100 terminates ( step 210 ). if the telephone call terminates , the mobile communication exchange 120 informs the charging center 130 of the charging information for the call , such as subscriber number , terminating number , call start time , call termination time , discount information , and so forth ( step 220 ). the charging center 130 calculates the charge for the latest call and adds the calculated charge to the total telephone charge ( step 230 ). the total telephone charge is calculated by adding up the accumulated telephone charge for each mobile terminal as well as the basic charge and the tax charge within a specified time period , for instance , a thirty - day period . the charging center 130 informs the mobile communication exchange 120 of the charge information for the latest call as well as the total charge ( step 240 ). the mobile communication exchange 120 receives the charge information for the latest call and the total charge , converts the received information in the form of a short message using the short message generating section 140 , and transmits the generated short message to the mobile terminal subscriber 100 ( steps 250 and 260 ). the mobile terminal 100 then receives the short message and displays the charge information of the latest call and the total charge on the lcd of the mobile phone . preferably , it may display characters such as \u201c charge for the latest call : 400 won , total charge : 10 , 500 won \u201d. accordingly , the mobile terminal subscriber can immediately confirm the charge for the latest call as well as the total charge calculated to include the latest call fig4 is a flowchart illustrating a method of informing a mobile terminal of the total telephone charge in the case that the mobile terminal requests confirmation of the total telephone charges up to now in a standby state according to another embodiment of the present invention . with reference to fig4 , if the mobile terminal subscriber intends to verify the total telephone charge up to now in a standby state , he / she can request for the total charge , for example , by sequentially pressing keys \u201c*\u201d, \u201c 1 \u201d, \u201c 1 \u201d, and \u201c send \u201d, or a specified key ( step 310 ). if the mobile communication exchange 120 receives the request for the confirmation of the total charge for the mobile terminal 100 through the base station 110 , the exchange 120 in turn requests the confirmation of the total charge of the mobile terminal subscriber to the charging center 130 ( step 320 ). the charging center 130 searches the total charging information corresponding to the requesting mobile terminal subscriber and informs the mobile communication exchange 120 of the total charging information ( step 330 ). the mobile communication exchange 120 generates a short message corresponding to the total charging information through the short message generating section 140 and transmits the short message to the mobile terminal 100 ( steps 340 and 350 ). the mobile terminal 100 receives the short message , and displays the total charge on the lcd . preferably , it may display characters such as \u201c total charge : 10 , 500 won ( i . e . korean currency )\u201d. as described above , it will be apparent that the present invention provides advantages in that the mobile terminal subscriber can be provided with an accurate telephone charge information by enabling the subscriber to immediately verify the charging information calculated by the charging center through the mobile terminal just after a telephone call . also , the mobile terminal subscriber can selectively verify at least one type of charging information out of various charging information managed by the charging center as occasion demands using the mobile terminal . while this invention has been described in connection with what is presently considered to be the most practical and preferred embodiments which display both the charge for the latest call and the total charge or only the total charge , it is to be understood that other modifications thereof may be made without departing from the scope of the invention . for example , the accumulated telephone charge for a specified time period desired by the mobile terminal subscriber , or another type of charge information other than the latest call and the total charge can be selectively displayed . thus , the present invention should not be limited to the disclosed embodiment but should be defined by the scope of the appended claims and their equivalents ."}
|
{"patent": "the preferred embodiments of the present invention will now be explained in detail with reference to the accompanying drawings . in the drawings , the same or similar elements are denoted by the same reference numerals even though they are depicted in different drawings . for the purpose of clarity , a detailed description of well known functions and configurations incorporated herein will be omitted as it may make the subject matter of the present invention unclear . fig2 is a block diagram illustrating the construction of a mobile radio communication system incorporating the present invention . referring to fig2 , a mobile communication exchange 120 performs an exchange function and interacts with another mobile communication , or the exchange 120 connects with a different communication network , such as a psdn , and controls the call termination / origination of a mobile terminal 100 . in addition , the mobile communication exchange 120 informs the charging center 130 of charging information of a telephone call , such as the subscriber &# 39 ; s number , terminating number , call start time , call termination time , discount information and so forth , when the call from the mobile terminal 100 terminates . the charging center 130 calculates and manages a telephone charge for the latest call , an accumulated telephone charge , and a total telephone charge using the charging information , then informs the respective telephone charges to the mobile communication exchange 120 . here , the management of the telephone charges is effected in such a manner that if thirty days elapses , the charging center 130 calculates the accumulated telephone charge for the calls made during the specific period for each mobile terminal subscriber , and notifies the respective mobile terminal subscriber of the total charge via mail . the bill includes a basic charge , a tax charge , and the calculated telephone charge . also , the mobile communication exchange 120 transmits to the mobile terminal 100 the telephone charging information , such as the telephone charge for the latest call , the accumulated telephone charge , and the total telephone charges that are transferred from the charging center 130 . the mobile terminal 100 receives the telephone charging information and displays the information on the display section ( not illustrated ), so that the mobile terminal subscriber can identify it . the mobile communication exchange 120 may be provided with a short message generating section 140 . this short message generating section 140 generates a short message which corresponds to the telephone charging information , such as the telephone charge for the latest call , the accumulated telephone charge , and the total telephone charge that are transferred from the charging center 130 in the form of a short message . the types of telephone charges transferred from the charging center 130 can be selectively determined according to the request by the mobile terminal subscriber . thus , different telephone charge information including the latest call , the accumulated telephone charge , and the total telephone charge can be provided to the subscriber . for instance , the mobile terminal subscriber can be provided with the accumulated telephone charge for a duration of a specified time period that he / she desires from the charging center 130 using the mobile terminal 100 . the type and the different telephone charge information transferred to the mobile terminal 100 may be predetermined or determined by the selection command from the mobile terminal subscriber . for example , the display unit of the mobile terminal can be pre - selected to display only the telephone charge for the latest call or only the total telephone charge , or both the latest call and the total telephone charge . fig3 is a flowchart illustrating a method of informing telephone charges to a mobile terminal for a telephone call and a total telephone charge call when the telephone call terminates according to the embodiment of the present invention . with reference to fig3 , the mobile communication exchange 120 detects whether the telephone call from the mobile terminal 100 terminates ( step 210 ). if the telephone call terminates , the mobile communication exchange 120 informs the charging center 130 of the charging information for the call , such as subscriber number , terminating number , call start time , call termination time , discount information , and so forth ( step 220 ). the charging center 130 calculates the charge for the latest call and adds the calculated charge to the total telephone charge ( step 230 ). the total telephone charge is calculated by adding up the accumulated telephone charge for each mobile terminal as well as the basic charge and the tax charge within a specified time period , for instance , a thirty - day period . the charging center 130 informs the mobile communication exchange 120 of the charge information for the latest call as well as the total charge ( step 240 ). the mobile communication exchange 120 receives the charge information for the latest call and the total charge , converts the received information in the form of a short message using the short message generating section 140 , and transmits the generated short message to the mobile terminal subscriber 100 ( steps 250 and 260 ). the mobile terminal 100 then receives the short message and displays the charge information of the latest call and the total charge on the lcd of the mobile phone . preferably , it may display characters such as \u201c charge for the latest call : 400 won , total charge : 10 , 500 won \u201d. accordingly , the mobile terminal subscriber can immediately confirm the charge for the latest call as well as the total charge calculated to include the latest call fig4 is a flowchart illustrating a method of informing a mobile terminal of the total telephone charge in the case that the mobile terminal requests confirmation of the total telephone charges up to now in a standby state according to another embodiment of the present invention . with reference to fig4 , if the mobile terminal subscriber intends to verify the total telephone charge up to now in a standby state , he / she can request for the total charge , for example , by sequentially pressing keys \u201c*\u201d, \u201c 1 \u201d, \u201c 1 \u201d, and \u201c send \u201d, or a specified key ( step 310 ). if the mobile communication exchange 120 receives the request for the confirmation of the total charge for the mobile terminal 100 through the base station 110 , the exchange 120 in turn requests the confirmation of the total charge of the mobile terminal subscriber to the charging center 130 ( step 320 ). the charging center 130 searches the total charging information corresponding to the requesting mobile terminal subscriber and informs the mobile communication exchange 120 of the total charging information ( step 330 ). the mobile communication exchange 120 generates a short message corresponding to the total charging information through the short message generating section 140 and transmits the short message to the mobile terminal 100 ( steps 340 and 350 ). the mobile terminal 100 receives the short message , and displays the total charge on the lcd . preferably , it may display characters such as \u201c total charge : 10 , 500 won ( i . e . korean currency )\u201d. as described above , it will be apparent that the present invention provides advantages in that the mobile terminal subscriber can be provided with an accurate telephone charge information by enabling the subscriber to immediately verify the charging information calculated by the charging center through the mobile terminal just after a telephone call . also , the mobile terminal subscriber can selectively verify at least one type of charging information out of various charging information managed by the charging center as occasion demands using the mobile terminal . while this invention has been described in connection with what is presently considered to be the most practical and preferred embodiments which display both the charge for the latest call and the total charge or only the total charge , it is to be understood that other modifications thereof may be made without departing from the scope of the invention . for example , the accumulated telephone charge for a specified time period desired by the mobile terminal subscriber , or another type of charge information other than the latest call and the total charge can be selectively displayed . thus , the present invention should not be limited to the disclosed embodiment but should be defined by the scope of the appended claims and their equivalents .", "category": "Textiles; Paper"}
|
Is the patent correctly categorized?
| 0.25 |
c54693020cd566de34695abc4d1ab9b0bb7c2dfdf03740340864fefaee1a46f1
| 0.574219 | 0.000103 | 0.410156 | 0.003479 | 0.570313 | 0.000345 |
null |
{"patent": "the preferred embodiments of the present invention will now be explained in detail with reference to the accompanying drawings . in the drawings , the same or similar elements are denoted by the same reference numerals even though they are depicted in different drawings . for the purpose of clarity , a detailed description of well known functions and configurations incorporated herein will be omitted as it may make the subject matter of the present invention unclear . fig2 is a block diagram illustrating the construction of a mobile radio communication system incorporating the present invention . referring to fig2 , a mobile communication exchange 120 performs an exchange function and interacts with another mobile communication , or the exchange 120 connects with a different communication network , such as a psdn , and controls the call termination / origination of a mobile terminal 100 . in addition , the mobile communication exchange 120 informs the charging center 130 of charging information of a telephone call , such as the subscriber &# 39 ; s number , terminating number , call start time , call termination time , discount information and so forth , when the call from the mobile terminal 100 terminates . the charging center 130 calculates and manages a telephone charge for the latest call , an accumulated telephone charge , and a total telephone charge using the charging information , then informs the respective telephone charges to the mobile communication exchange 120 . here , the management of the telephone charges is effected in such a manner that if thirty days elapses , the charging center 130 calculates the accumulated telephone charge for the calls made during the specific period for each mobile terminal subscriber , and notifies the respective mobile terminal subscriber of the total charge via mail . the bill includes a basic charge , a tax charge , and the calculated telephone charge . also , the mobile communication exchange 120 transmits to the mobile terminal 100 the telephone charging information , such as the telephone charge for the latest call , the accumulated telephone charge , and the total telephone charges that are transferred from the charging center 130 . the mobile terminal 100 receives the telephone charging information and displays the information on the display section ( not illustrated ), so that the mobile terminal subscriber can identify it . the mobile communication exchange 120 may be provided with a short message generating section 140 . this short message generating section 140 generates a short message which corresponds to the telephone charging information , such as the telephone charge for the latest call , the accumulated telephone charge , and the total telephone charge that are transferred from the charging center 130 in the form of a short message . the types of telephone charges transferred from the charging center 130 can be selectively determined according to the request by the mobile terminal subscriber . thus , different telephone charge information including the latest call , the accumulated telephone charge , and the total telephone charge can be provided to the subscriber . for instance , the mobile terminal subscriber can be provided with the accumulated telephone charge for a duration of a specified time period that he / she desires from the charging center 130 using the mobile terminal 100 . the type and the different telephone charge information transferred to the mobile terminal 100 may be predetermined or determined by the selection command from the mobile terminal subscriber . for example , the display unit of the mobile terminal can be pre - selected to display only the telephone charge for the latest call or only the total telephone charge , or both the latest call and the total telephone charge . fig3 is a flowchart illustrating a method of informing telephone charges to a mobile terminal for a telephone call and a total telephone charge call when the telephone call terminates according to the embodiment of the present invention . with reference to fig3 , the mobile communication exchange 120 detects whether the telephone call from the mobile terminal 100 terminates ( step 210 ). if the telephone call terminates , the mobile communication exchange 120 informs the charging center 130 of the charging information for the call , such as subscriber number , terminating number , call start time , call termination time , discount information , and so forth ( step 220 ). the charging center 130 calculates the charge for the latest call and adds the calculated charge to the total telephone charge ( step 230 ). the total telephone charge is calculated by adding up the accumulated telephone charge for each mobile terminal as well as the basic charge and the tax charge within a specified time period , for instance , a thirty - day period . the charging center 130 informs the mobile communication exchange 120 of the charge information for the latest call as well as the total charge ( step 240 ). the mobile communication exchange 120 receives the charge information for the latest call and the total charge , converts the received information in the form of a short message using the short message generating section 140 , and transmits the generated short message to the mobile terminal subscriber 100 ( steps 250 and 260 ). the mobile terminal 100 then receives the short message and displays the charge information of the latest call and the total charge on the lcd of the mobile phone . preferably , it may display characters such as \u201c charge for the latest call : 400 won , total charge : 10 , 500 won \u201d. accordingly , the mobile terminal subscriber can immediately confirm the charge for the latest call as well as the total charge calculated to include the latest call fig4 is a flowchart illustrating a method of informing a mobile terminal of the total telephone charge in the case that the mobile terminal requests confirmation of the total telephone charges up to now in a standby state according to another embodiment of the present invention . with reference to fig4 , if the mobile terminal subscriber intends to verify the total telephone charge up to now in a standby state , he / she can request for the total charge , for example , by sequentially pressing keys \u201c*\u201d, \u201c 1 \u201d, \u201c 1 \u201d, and \u201c send \u201d, or a specified key ( step 310 ). if the mobile communication exchange 120 receives the request for the confirmation of the total charge for the mobile terminal 100 through the base station 110 , the exchange 120 in turn requests the confirmation of the total charge of the mobile terminal subscriber to the charging center 130 ( step 320 ). the charging center 130 searches the total charging information corresponding to the requesting mobile terminal subscriber and informs the mobile communication exchange 120 of the total charging information ( step 330 ). the mobile communication exchange 120 generates a short message corresponding to the total charging information through the short message generating section 140 and transmits the short message to the mobile terminal 100 ( steps 340 and 350 ). the mobile terminal 100 receives the short message , and displays the total charge on the lcd . preferably , it may display characters such as \u201c total charge : 10 , 500 won ( i . e . korean currency )\u201d. as described above , it will be apparent that the present invention provides advantages in that the mobile terminal subscriber can be provided with an accurate telephone charge information by enabling the subscriber to immediately verify the charging information calculated by the charging center through the mobile terminal just after a telephone call . also , the mobile terminal subscriber can selectively verify at least one type of charging information out of various charging information managed by the charging center as occasion demands using the mobile terminal . while this invention has been described in connection with what is presently considered to be the most practical and preferred embodiments which display both the charge for the latest call and the total charge or only the total charge , it is to be understood that other modifications thereof may be made without departing from the scope of the invention . for example , the accumulated telephone charge for a specified time period desired by the mobile terminal subscriber , or another type of charge information other than the latest call and the total charge can be selectively displayed . thus , the present invention should not be limited to the disclosed embodiment but should be defined by the scope of the appended claims and their equivalents .", "category": "Electricity"}
|
{"category": "Fixed Constructions", "patent": "the preferred embodiments of the present invention will now be explained in detail with reference to the accompanying drawings . in the drawings , the same or similar elements are denoted by the same reference numerals even though they are depicted in different drawings . for the purpose of clarity , a detailed description of well known functions and configurations incorporated herein will be omitted as it may make the subject matter of the present invention unclear . fig2 is a block diagram illustrating the construction of a mobile radio communication system incorporating the present invention . referring to fig2 , a mobile communication exchange 120 performs an exchange function and interacts with another mobile communication , or the exchange 120 connects with a different communication network , such as a psdn , and controls the call termination / origination of a mobile terminal 100 . in addition , the mobile communication exchange 120 informs the charging center 130 of charging information of a telephone call , such as the subscriber &# 39 ; s number , terminating number , call start time , call termination time , discount information and so forth , when the call from the mobile terminal 100 terminates . the charging center 130 calculates and manages a telephone charge for the latest call , an accumulated telephone charge , and a total telephone charge using the charging information , then informs the respective telephone charges to the mobile communication exchange 120 . here , the management of the telephone charges is effected in such a manner that if thirty days elapses , the charging center 130 calculates the accumulated telephone charge for the calls made during the specific period for each mobile terminal subscriber , and notifies the respective mobile terminal subscriber of the total charge via mail . the bill includes a basic charge , a tax charge , and the calculated telephone charge . also , the mobile communication exchange 120 transmits to the mobile terminal 100 the telephone charging information , such as the telephone charge for the latest call , the accumulated telephone charge , and the total telephone charges that are transferred from the charging center 130 . the mobile terminal 100 receives the telephone charging information and displays the information on the display section ( not illustrated ), so that the mobile terminal subscriber can identify it . the mobile communication exchange 120 may be provided with a short message generating section 140 . this short message generating section 140 generates a short message which corresponds to the telephone charging information , such as the telephone charge for the latest call , the accumulated telephone charge , and the total telephone charge that are transferred from the charging center 130 in the form of a short message . the types of telephone charges transferred from the charging center 130 can be selectively determined according to the request by the mobile terminal subscriber . thus , different telephone charge information including the latest call , the accumulated telephone charge , and the total telephone charge can be provided to the subscriber . for instance , the mobile terminal subscriber can be provided with the accumulated telephone charge for a duration of a specified time period that he / she desires from the charging center 130 using the mobile terminal 100 . the type and the different telephone charge information transferred to the mobile terminal 100 may be predetermined or determined by the selection command from the mobile terminal subscriber . for example , the display unit of the mobile terminal can be pre - selected to display only the telephone charge for the latest call or only the total telephone charge , or both the latest call and the total telephone charge . fig3 is a flowchart illustrating a method of informing telephone charges to a mobile terminal for a telephone call and a total telephone charge call when the telephone call terminates according to the embodiment of the present invention . with reference to fig3 , the mobile communication exchange 120 detects whether the telephone call from the mobile terminal 100 terminates ( step 210 ). if the telephone call terminates , the mobile communication exchange 120 informs the charging center 130 of the charging information for the call , such as subscriber number , terminating number , call start time , call termination time , discount information , and so forth ( step 220 ). the charging center 130 calculates the charge for the latest call and adds the calculated charge to the total telephone charge ( step 230 ). the total telephone charge is calculated by adding up the accumulated telephone charge for each mobile terminal as well as the basic charge and the tax charge within a specified time period , for instance , a thirty - day period . the charging center 130 informs the mobile communication exchange 120 of the charge information for the latest call as well as the total charge ( step 240 ). the mobile communication exchange 120 receives the charge information for the latest call and the total charge , converts the received information in the form of a short message using the short message generating section 140 , and transmits the generated short message to the mobile terminal subscriber 100 ( steps 250 and 260 ). the mobile terminal 100 then receives the short message and displays the charge information of the latest call and the total charge on the lcd of the mobile phone . preferably , it may display characters such as \u201c charge for the latest call : 400 won , total charge : 10 , 500 won \u201d. accordingly , the mobile terminal subscriber can immediately confirm the charge for the latest call as well as the total charge calculated to include the latest call fig4 is a flowchart illustrating a method of informing a mobile terminal of the total telephone charge in the case that the mobile terminal requests confirmation of the total telephone charges up to now in a standby state according to another embodiment of the present invention . with reference to fig4 , if the mobile terminal subscriber intends to verify the total telephone charge up to now in a standby state , he / she can request for the total charge , for example , by sequentially pressing keys \u201c*\u201d, \u201c 1 \u201d, \u201c 1 \u201d, and \u201c send \u201d, or a specified key ( step 310 ). if the mobile communication exchange 120 receives the request for the confirmation of the total charge for the mobile terminal 100 through the base station 110 , the exchange 120 in turn requests the confirmation of the total charge of the mobile terminal subscriber to the charging center 130 ( step 320 ). the charging center 130 searches the total charging information corresponding to the requesting mobile terminal subscriber and informs the mobile communication exchange 120 of the total charging information ( step 330 ). the mobile communication exchange 120 generates a short message corresponding to the total charging information through the short message generating section 140 and transmits the short message to the mobile terminal 100 ( steps 340 and 350 ). the mobile terminal 100 receives the short message , and displays the total charge on the lcd . preferably , it may display characters such as \u201c total charge : 10 , 500 won ( i . e . korean currency )\u201d. as described above , it will be apparent that the present invention provides advantages in that the mobile terminal subscriber can be provided with an accurate telephone charge information by enabling the subscriber to immediately verify the charging information calculated by the charging center through the mobile terminal just after a telephone call . also , the mobile terminal subscriber can selectively verify at least one type of charging information out of various charging information managed by the charging center as occasion demands using the mobile terminal . while this invention has been described in connection with what is presently considered to be the most practical and preferred embodiments which display both the charge for the latest call and the total charge or only the total charge , it is to be understood that other modifications thereof may be made without departing from the scope of the invention . for example , the accumulated telephone charge for a specified time period desired by the mobile terminal subscriber , or another type of charge information other than the latest call and the total charge can be selectively displayed . thus , the present invention should not be limited to the disclosed embodiment but should be defined by the scope of the appended claims and their equivalents ."}
|
Does the patent belong in this category?
| 0.25 |
c54693020cd566de34695abc4d1ab9b0bb7c2dfdf03740340864fefaee1a46f1
| 0.174805 | 0.027222 | 0.039551 | 0.059326 | 0.060059 | 0.039063 |
null |
{"category": "Electricity", "patent": "the preferred embodiments of the present invention will now be explained in detail with reference to the accompanying drawings . in the drawings , the same or similar elements are denoted by the same reference numerals even though they are depicted in different drawings . for the purpose of clarity , a detailed description of well known functions and configurations incorporated herein will be omitted as it may make the subject matter of the present invention unclear . fig2 is a block diagram illustrating the construction of a mobile radio communication system incorporating the present invention . referring to fig2 , a mobile communication exchange 120 performs an exchange function and interacts with another mobile communication , or the exchange 120 connects with a different communication network , such as a psdn , and controls the call termination / origination of a mobile terminal 100 . in addition , the mobile communication exchange 120 informs the charging center 130 of charging information of a telephone call , such as the subscriber &# 39 ; s number , terminating number , call start time , call termination time , discount information and so forth , when the call from the mobile terminal 100 terminates . the charging center 130 calculates and manages a telephone charge for the latest call , an accumulated telephone charge , and a total telephone charge using the charging information , then informs the respective telephone charges to the mobile communication exchange 120 . here , the management of the telephone charges is effected in such a manner that if thirty days elapses , the charging center 130 calculates the accumulated telephone charge for the calls made during the specific period for each mobile terminal subscriber , and notifies the respective mobile terminal subscriber of the total charge via mail . the bill includes a basic charge , a tax charge , and the calculated telephone charge . also , the mobile communication exchange 120 transmits to the mobile terminal 100 the telephone charging information , such as the telephone charge for the latest call , the accumulated telephone charge , and the total telephone charges that are transferred from the charging center 130 . the mobile terminal 100 receives the telephone charging information and displays the information on the display section ( not illustrated ), so that the mobile terminal subscriber can identify it . the mobile communication exchange 120 may be provided with a short message generating section 140 . this short message generating section 140 generates a short message which corresponds to the telephone charging information , such as the telephone charge for the latest call , the accumulated telephone charge , and the total telephone charge that are transferred from the charging center 130 in the form of a short message . the types of telephone charges transferred from the charging center 130 can be selectively determined according to the request by the mobile terminal subscriber . thus , different telephone charge information including the latest call , the accumulated telephone charge , and the total telephone charge can be provided to the subscriber . for instance , the mobile terminal subscriber can be provided with the accumulated telephone charge for a duration of a specified time period that he / she desires from the charging center 130 using the mobile terminal 100 . the type and the different telephone charge information transferred to the mobile terminal 100 may be predetermined or determined by the selection command from the mobile terminal subscriber . for example , the display unit of the mobile terminal can be pre - selected to display only the telephone charge for the latest call or only the total telephone charge , or both the latest call and the total telephone charge . fig3 is a flowchart illustrating a method of informing telephone charges to a mobile terminal for a telephone call and a total telephone charge call when the telephone call terminates according to the embodiment of the present invention . with reference to fig3 , the mobile communication exchange 120 detects whether the telephone call from the mobile terminal 100 terminates ( step 210 ). if the telephone call terminates , the mobile communication exchange 120 informs the charging center 130 of the charging information for the call , such as subscriber number , terminating number , call start time , call termination time , discount information , and so forth ( step 220 ). the charging center 130 calculates the charge for the latest call and adds the calculated charge to the total telephone charge ( step 230 ). the total telephone charge is calculated by adding up the accumulated telephone charge for each mobile terminal as well as the basic charge and the tax charge within a specified time period , for instance , a thirty - day period . the charging center 130 informs the mobile communication exchange 120 of the charge information for the latest call as well as the total charge ( step 240 ). the mobile communication exchange 120 receives the charge information for the latest call and the total charge , converts the received information in the form of a short message using the short message generating section 140 , and transmits the generated short message to the mobile terminal subscriber 100 ( steps 250 and 260 ). the mobile terminal 100 then receives the short message and displays the charge information of the latest call and the total charge on the lcd of the mobile phone . preferably , it may display characters such as \u201c charge for the latest call : 400 won , total charge : 10 , 500 won \u201d. accordingly , the mobile terminal subscriber can immediately confirm the charge for the latest call as well as the total charge calculated to include the latest call fig4 is a flowchart illustrating a method of informing a mobile terminal of the total telephone charge in the case that the mobile terminal requests confirmation of the total telephone charges up to now in a standby state according to another embodiment of the present invention . with reference to fig4 , if the mobile terminal subscriber intends to verify the total telephone charge up to now in a standby state , he / she can request for the total charge , for example , by sequentially pressing keys \u201c*\u201d, \u201c 1 \u201d, \u201c 1 \u201d, and \u201c send \u201d, or a specified key ( step 310 ). if the mobile communication exchange 120 receives the request for the confirmation of the total charge for the mobile terminal 100 through the base station 110 , the exchange 120 in turn requests the confirmation of the total charge of the mobile terminal subscriber to the charging center 130 ( step 320 ). the charging center 130 searches the total charging information corresponding to the requesting mobile terminal subscriber and informs the mobile communication exchange 120 of the total charging information ( step 330 ). the mobile communication exchange 120 generates a short message corresponding to the total charging information through the short message generating section 140 and transmits the short message to the mobile terminal 100 ( steps 340 and 350 ). the mobile terminal 100 receives the short message , and displays the total charge on the lcd . preferably , it may display characters such as \u201c total charge : 10 , 500 won ( i . e . korean currency )\u201d. as described above , it will be apparent that the present invention provides advantages in that the mobile terminal subscriber can be provided with an accurate telephone charge information by enabling the subscriber to immediately verify the charging information calculated by the charging center through the mobile terminal just after a telephone call . also , the mobile terminal subscriber can selectively verify at least one type of charging information out of various charging information managed by the charging center as occasion demands using the mobile terminal . while this invention has been described in connection with what is presently considered to be the most practical and preferred embodiments which display both the charge for the latest call and the total charge or only the total charge , it is to be understood that other modifications thereof may be made without departing from the scope of the invention . for example , the accumulated telephone charge for a specified time period desired by the mobile terminal subscriber , or another type of charge information other than the latest call and the total charge can be selectively displayed . thus , the present invention should not be limited to the disclosed embodiment but should be defined by the scope of the appended claims and their equivalents ."}
|
{"category": "Mechanical Engineering; Lightning; Heating; Weapons; Blasting", "patent": "the preferred embodiments of the present invention will now be explained in detail with reference to the accompanying drawings . in the drawings , the same or similar elements are denoted by the same reference numerals even though they are depicted in different drawings . for the purpose of clarity , a detailed description of well known functions and configurations incorporated herein will be omitted as it may make the subject matter of the present invention unclear . fig2 is a block diagram illustrating the construction of a mobile radio communication system incorporating the present invention . referring to fig2 , a mobile communication exchange 120 performs an exchange function and interacts with another mobile communication , or the exchange 120 connects with a different communication network , such as a psdn , and controls the call termination / origination of a mobile terminal 100 . in addition , the mobile communication exchange 120 informs the charging center 130 of charging information of a telephone call , such as the subscriber &# 39 ; s number , terminating number , call start time , call termination time , discount information and so forth , when the call from the mobile terminal 100 terminates . the charging center 130 calculates and manages a telephone charge for the latest call , an accumulated telephone charge , and a total telephone charge using the charging information , then informs the respective telephone charges to the mobile communication exchange 120 . here , the management of the telephone charges is effected in such a manner that if thirty days elapses , the charging center 130 calculates the accumulated telephone charge for the calls made during the specific period for each mobile terminal subscriber , and notifies the respective mobile terminal subscriber of the total charge via mail . the bill includes a basic charge , a tax charge , and the calculated telephone charge . also , the mobile communication exchange 120 transmits to the mobile terminal 100 the telephone charging information , such as the telephone charge for the latest call , the accumulated telephone charge , and the total telephone charges that are transferred from the charging center 130 . the mobile terminal 100 receives the telephone charging information and displays the information on the display section ( not illustrated ), so that the mobile terminal subscriber can identify it . the mobile communication exchange 120 may be provided with a short message generating section 140 . this short message generating section 140 generates a short message which corresponds to the telephone charging information , such as the telephone charge for the latest call , the accumulated telephone charge , and the total telephone charge that are transferred from the charging center 130 in the form of a short message . the types of telephone charges transferred from the charging center 130 can be selectively determined according to the request by the mobile terminal subscriber . thus , different telephone charge information including the latest call , the accumulated telephone charge , and the total telephone charge can be provided to the subscriber . for instance , the mobile terminal subscriber can be provided with the accumulated telephone charge for a duration of a specified time period that he / she desires from the charging center 130 using the mobile terminal 100 . the type and the different telephone charge information transferred to the mobile terminal 100 may be predetermined or determined by the selection command from the mobile terminal subscriber . for example , the display unit of the mobile terminal can be pre - selected to display only the telephone charge for the latest call or only the total telephone charge , or both the latest call and the total telephone charge . fig3 is a flowchart illustrating a method of informing telephone charges to a mobile terminal for a telephone call and a total telephone charge call when the telephone call terminates according to the embodiment of the present invention . with reference to fig3 , the mobile communication exchange 120 detects whether the telephone call from the mobile terminal 100 terminates ( step 210 ). if the telephone call terminates , the mobile communication exchange 120 informs the charging center 130 of the charging information for the call , such as subscriber number , terminating number , call start time , call termination time , discount information , and so forth ( step 220 ). the charging center 130 calculates the charge for the latest call and adds the calculated charge to the total telephone charge ( step 230 ). the total telephone charge is calculated by adding up the accumulated telephone charge for each mobile terminal as well as the basic charge and the tax charge within a specified time period , for instance , a thirty - day period . the charging center 130 informs the mobile communication exchange 120 of the charge information for the latest call as well as the total charge ( step 240 ). the mobile communication exchange 120 receives the charge information for the latest call and the total charge , converts the received information in the form of a short message using the short message generating section 140 , and transmits the generated short message to the mobile terminal subscriber 100 ( steps 250 and 260 ). the mobile terminal 100 then receives the short message and displays the charge information of the latest call and the total charge on the lcd of the mobile phone . preferably , it may display characters such as \u201c charge for the latest call : 400 won , total charge : 10 , 500 won \u201d. accordingly , the mobile terminal subscriber can immediately confirm the charge for the latest call as well as the total charge calculated to include the latest call fig4 is a flowchart illustrating a method of informing a mobile terminal of the total telephone charge in the case that the mobile terminal requests confirmation of the total telephone charges up to now in a standby state according to another embodiment of the present invention . with reference to fig4 , if the mobile terminal subscriber intends to verify the total telephone charge up to now in a standby state , he / she can request for the total charge , for example , by sequentially pressing keys \u201c*\u201d, \u201c 1 \u201d, \u201c 1 \u201d, and \u201c send \u201d, or a specified key ( step 310 ). if the mobile communication exchange 120 receives the request for the confirmation of the total charge for the mobile terminal 100 through the base station 110 , the exchange 120 in turn requests the confirmation of the total charge of the mobile terminal subscriber to the charging center 130 ( step 320 ). the charging center 130 searches the total charging information corresponding to the requesting mobile terminal subscriber and informs the mobile communication exchange 120 of the total charging information ( step 330 ). the mobile communication exchange 120 generates a short message corresponding to the total charging information through the short message generating section 140 and transmits the short message to the mobile terminal 100 ( steps 340 and 350 ). the mobile terminal 100 receives the short message , and displays the total charge on the lcd . preferably , it may display characters such as \u201c total charge : 10 , 500 won ( i . e . korean currency )\u201d. as described above , it will be apparent that the present invention provides advantages in that the mobile terminal subscriber can be provided with an accurate telephone charge information by enabling the subscriber to immediately verify the charging information calculated by the charging center through the mobile terminal just after a telephone call . also , the mobile terminal subscriber can selectively verify at least one type of charging information out of various charging information managed by the charging center as occasion demands using the mobile terminal . while this invention has been described in connection with what is presently considered to be the most practical and preferred embodiments which display both the charge for the latest call and the total charge or only the total charge , it is to be understood that other modifications thereof may be made without departing from the scope of the invention . for example , the accumulated telephone charge for a specified time period desired by the mobile terminal subscriber , or another type of charge information other than the latest call and the total charge can be selectively displayed . thus , the present invention should not be limited to the disclosed embodiment but should be defined by the scope of the appended claims and their equivalents ."}
|
Is the categorization of this patent accurate?
| 0.25 |
c54693020cd566de34695abc4d1ab9b0bb7c2dfdf03740340864fefaee1a46f1
| 0.570313 | 0.002808 | 0.207031 | 0.001457 | 0.318359 | 0.011658 |
null |
{"patent": "the preferred embodiments of the present invention will now be explained in detail with reference to the accompanying drawings . in the drawings , the same or similar elements are denoted by the same reference numerals even though they are depicted in different drawings . for the purpose of clarity , a detailed description of well known functions and configurations incorporated herein will be omitted as it may make the subject matter of the present invention unclear . fig2 is a block diagram illustrating the construction of a mobile radio communication system incorporating the present invention . referring to fig2 , a mobile communication exchange 120 performs an exchange function and interacts with another mobile communication , or the exchange 120 connects with a different communication network , such as a psdn , and controls the call termination / origination of a mobile terminal 100 . in addition , the mobile communication exchange 120 informs the charging center 130 of charging information of a telephone call , such as the subscriber &# 39 ; s number , terminating number , call start time , call termination time , discount information and so forth , when the call from the mobile terminal 100 terminates . the charging center 130 calculates and manages a telephone charge for the latest call , an accumulated telephone charge , and a total telephone charge using the charging information , then informs the respective telephone charges to the mobile communication exchange 120 . here , the management of the telephone charges is effected in such a manner that if thirty days elapses , the charging center 130 calculates the accumulated telephone charge for the calls made during the specific period for each mobile terminal subscriber , and notifies the respective mobile terminal subscriber of the total charge via mail . the bill includes a basic charge , a tax charge , and the calculated telephone charge . also , the mobile communication exchange 120 transmits to the mobile terminal 100 the telephone charging information , such as the telephone charge for the latest call , the accumulated telephone charge , and the total telephone charges that are transferred from the charging center 130 . the mobile terminal 100 receives the telephone charging information and displays the information on the display section ( not illustrated ), so that the mobile terminal subscriber can identify it . the mobile communication exchange 120 may be provided with a short message generating section 140 . this short message generating section 140 generates a short message which corresponds to the telephone charging information , such as the telephone charge for the latest call , the accumulated telephone charge , and the total telephone charge that are transferred from the charging center 130 in the form of a short message . the types of telephone charges transferred from the charging center 130 can be selectively determined according to the request by the mobile terminal subscriber . thus , different telephone charge information including the latest call , the accumulated telephone charge , and the total telephone charge can be provided to the subscriber . for instance , the mobile terminal subscriber can be provided with the accumulated telephone charge for a duration of a specified time period that he / she desires from the charging center 130 using the mobile terminal 100 . the type and the different telephone charge information transferred to the mobile terminal 100 may be predetermined or determined by the selection command from the mobile terminal subscriber . for example , the display unit of the mobile terminal can be pre - selected to display only the telephone charge for the latest call or only the total telephone charge , or both the latest call and the total telephone charge . fig3 is a flowchart illustrating a method of informing telephone charges to a mobile terminal for a telephone call and a total telephone charge call when the telephone call terminates according to the embodiment of the present invention . with reference to fig3 , the mobile communication exchange 120 detects whether the telephone call from the mobile terminal 100 terminates ( step 210 ). if the telephone call terminates , the mobile communication exchange 120 informs the charging center 130 of the charging information for the call , such as subscriber number , terminating number , call start time , call termination time , discount information , and so forth ( step 220 ). the charging center 130 calculates the charge for the latest call and adds the calculated charge to the total telephone charge ( step 230 ). the total telephone charge is calculated by adding up the accumulated telephone charge for each mobile terminal as well as the basic charge and the tax charge within a specified time period , for instance , a thirty - day period . the charging center 130 informs the mobile communication exchange 120 of the charge information for the latest call as well as the total charge ( step 240 ). the mobile communication exchange 120 receives the charge information for the latest call and the total charge , converts the received information in the form of a short message using the short message generating section 140 , and transmits the generated short message to the mobile terminal subscriber 100 ( steps 250 and 260 ). the mobile terminal 100 then receives the short message and displays the charge information of the latest call and the total charge on the lcd of the mobile phone . preferably , it may display characters such as \u201c charge for the latest call : 400 won , total charge : 10 , 500 won \u201d. accordingly , the mobile terminal subscriber can immediately confirm the charge for the latest call as well as the total charge calculated to include the latest call fig4 is a flowchart illustrating a method of informing a mobile terminal of the total telephone charge in the case that the mobile terminal requests confirmation of the total telephone charges up to now in a standby state according to another embodiment of the present invention . with reference to fig4 , if the mobile terminal subscriber intends to verify the total telephone charge up to now in a standby state , he / she can request for the total charge , for example , by sequentially pressing keys \u201c*\u201d, \u201c 1 \u201d, \u201c 1 \u201d, and \u201c send \u201d, or a specified key ( step 310 ). if the mobile communication exchange 120 receives the request for the confirmation of the total charge for the mobile terminal 100 through the base station 110 , the exchange 120 in turn requests the confirmation of the total charge of the mobile terminal subscriber to the charging center 130 ( step 320 ). the charging center 130 searches the total charging information corresponding to the requesting mobile terminal subscriber and informs the mobile communication exchange 120 of the total charging information ( step 330 ). the mobile communication exchange 120 generates a short message corresponding to the total charging information through the short message generating section 140 and transmits the short message to the mobile terminal 100 ( steps 340 and 350 ). the mobile terminal 100 receives the short message , and displays the total charge on the lcd . preferably , it may display characters such as \u201c total charge : 10 , 500 won ( i . e . korean currency )\u201d. as described above , it will be apparent that the present invention provides advantages in that the mobile terminal subscriber can be provided with an accurate telephone charge information by enabling the subscriber to immediately verify the charging information calculated by the charging center through the mobile terminal just after a telephone call . also , the mobile terminal subscriber can selectively verify at least one type of charging information out of various charging information managed by the charging center as occasion demands using the mobile terminal . while this invention has been described in connection with what is presently considered to be the most practical and preferred embodiments which display both the charge for the latest call and the total charge or only the total charge , it is to be understood that other modifications thereof may be made without departing from the scope of the invention . for example , the accumulated telephone charge for a specified time period desired by the mobile terminal subscriber , or another type of charge information other than the latest call and the total charge can be selectively displayed . thus , the present invention should not be limited to the disclosed embodiment but should be defined by the scope of the appended claims and their equivalents .", "category": "Electricity"}
|
{"patent": "the preferred embodiments of the present invention will now be explained in detail with reference to the accompanying drawings . in the drawings , the same or similar elements are denoted by the same reference numerals even though they are depicted in different drawings . for the purpose of clarity , a detailed description of well known functions and configurations incorporated herein will be omitted as it may make the subject matter of the present invention unclear . fig2 is a block diagram illustrating the construction of a mobile radio communication system incorporating the present invention . referring to fig2 , a mobile communication exchange 120 performs an exchange function and interacts with another mobile communication , or the exchange 120 connects with a different communication network , such as a psdn , and controls the call termination / origination of a mobile terminal 100 . in addition , the mobile communication exchange 120 informs the charging center 130 of charging information of a telephone call , such as the subscriber &# 39 ; s number , terminating number , call start time , call termination time , discount information and so forth , when the call from the mobile terminal 100 terminates . the charging center 130 calculates and manages a telephone charge for the latest call , an accumulated telephone charge , and a total telephone charge using the charging information , then informs the respective telephone charges to the mobile communication exchange 120 . here , the management of the telephone charges is effected in such a manner that if thirty days elapses , the charging center 130 calculates the accumulated telephone charge for the calls made during the specific period for each mobile terminal subscriber , and notifies the respective mobile terminal subscriber of the total charge via mail . the bill includes a basic charge , a tax charge , and the calculated telephone charge . also , the mobile communication exchange 120 transmits to the mobile terminal 100 the telephone charging information , such as the telephone charge for the latest call , the accumulated telephone charge , and the total telephone charges that are transferred from the charging center 130 . the mobile terminal 100 receives the telephone charging information and displays the information on the display section ( not illustrated ), so that the mobile terminal subscriber can identify it . the mobile communication exchange 120 may be provided with a short message generating section 140 . this short message generating section 140 generates a short message which corresponds to the telephone charging information , such as the telephone charge for the latest call , the accumulated telephone charge , and the total telephone charge that are transferred from the charging center 130 in the form of a short message . the types of telephone charges transferred from the charging center 130 can be selectively determined according to the request by the mobile terminal subscriber . thus , different telephone charge information including the latest call , the accumulated telephone charge , and the total telephone charge can be provided to the subscriber . for instance , the mobile terminal subscriber can be provided with the accumulated telephone charge for a duration of a specified time period that he / she desires from the charging center 130 using the mobile terminal 100 . the type and the different telephone charge information transferred to the mobile terminal 100 may be predetermined or determined by the selection command from the mobile terminal subscriber . for example , the display unit of the mobile terminal can be pre - selected to display only the telephone charge for the latest call or only the total telephone charge , or both the latest call and the total telephone charge . fig3 is a flowchart illustrating a method of informing telephone charges to a mobile terminal for a telephone call and a total telephone charge call when the telephone call terminates according to the embodiment of the present invention . with reference to fig3 , the mobile communication exchange 120 detects whether the telephone call from the mobile terminal 100 terminates ( step 210 ). if the telephone call terminates , the mobile communication exchange 120 informs the charging center 130 of the charging information for the call , such as subscriber number , terminating number , call start time , call termination time , discount information , and so forth ( step 220 ). the charging center 130 calculates the charge for the latest call and adds the calculated charge to the total telephone charge ( step 230 ). the total telephone charge is calculated by adding up the accumulated telephone charge for each mobile terminal as well as the basic charge and the tax charge within a specified time period , for instance , a thirty - day period . the charging center 130 informs the mobile communication exchange 120 of the charge information for the latest call as well as the total charge ( step 240 ). the mobile communication exchange 120 receives the charge information for the latest call and the total charge , converts the received information in the form of a short message using the short message generating section 140 , and transmits the generated short message to the mobile terminal subscriber 100 ( steps 250 and 260 ). the mobile terminal 100 then receives the short message and displays the charge information of the latest call and the total charge on the lcd of the mobile phone . preferably , it may display characters such as \u201c charge for the latest call : 400 won , total charge : 10 , 500 won \u201d. accordingly , the mobile terminal subscriber can immediately confirm the charge for the latest call as well as the total charge calculated to include the latest call fig4 is a flowchart illustrating a method of informing a mobile terminal of the total telephone charge in the case that the mobile terminal requests confirmation of the total telephone charges up to now in a standby state according to another embodiment of the present invention . with reference to fig4 , if the mobile terminal subscriber intends to verify the total telephone charge up to now in a standby state , he / she can request for the total charge , for example , by sequentially pressing keys \u201c*\u201d, \u201c 1 \u201d, \u201c 1 \u201d, and \u201c send \u201d, or a specified key ( step 310 ). if the mobile communication exchange 120 receives the request for the confirmation of the total charge for the mobile terminal 100 through the base station 110 , the exchange 120 in turn requests the confirmation of the total charge of the mobile terminal subscriber to the charging center 130 ( step 320 ). the charging center 130 searches the total charging information corresponding to the requesting mobile terminal subscriber and informs the mobile communication exchange 120 of the total charging information ( step 330 ). the mobile communication exchange 120 generates a short message corresponding to the total charging information through the short message generating section 140 and transmits the short message to the mobile terminal 100 ( steps 340 and 350 ). the mobile terminal 100 receives the short message , and displays the total charge on the lcd . preferably , it may display characters such as \u201c total charge : 10 , 500 won ( i . e . korean currency )\u201d. as described above , it will be apparent that the present invention provides advantages in that the mobile terminal subscriber can be provided with an accurate telephone charge information by enabling the subscriber to immediately verify the charging information calculated by the charging center through the mobile terminal just after a telephone call . also , the mobile terminal subscriber can selectively verify at least one type of charging information out of various charging information managed by the charging center as occasion demands using the mobile terminal . while this invention has been described in connection with what is presently considered to be the most practical and preferred embodiments which display both the charge for the latest call and the total charge or only the total charge , it is to be understood that other modifications thereof may be made without departing from the scope of the invention . for example , the accumulated telephone charge for a specified time period desired by the mobile terminal subscriber , or another type of charge information other than the latest call and the total charge can be selectively displayed . thus , the present invention should not be limited to the disclosed embodiment but should be defined by the scope of the appended claims and their equivalents .", "category": "Physics"}
|
Does the category match the content of the patent?
| 0.25 |
c54693020cd566de34695abc4d1ab9b0bb7c2dfdf03740340864fefaee1a46f1
| 0.714844 | 0.01001 | 0.092773 | 0.013611 | 0.069336 | 0.009399 |
null |
{"category": "Electricity", "patent": "the preferred embodiments of the present invention will now be explained in detail with reference to the accompanying drawings . in the drawings , the same or similar elements are denoted by the same reference numerals even though they are depicted in different drawings . for the purpose of clarity , a detailed description of well known functions and configurations incorporated herein will be omitted as it may make the subject matter of the present invention unclear . fig2 is a block diagram illustrating the construction of a mobile radio communication system incorporating the present invention . referring to fig2 , a mobile communication exchange 120 performs an exchange function and interacts with another mobile communication , or the exchange 120 connects with a different communication network , such as a psdn , and controls the call termination / origination of a mobile terminal 100 . in addition , the mobile communication exchange 120 informs the charging center 130 of charging information of a telephone call , such as the subscriber &# 39 ; s number , terminating number , call start time , call termination time , discount information and so forth , when the call from the mobile terminal 100 terminates . the charging center 130 calculates and manages a telephone charge for the latest call , an accumulated telephone charge , and a total telephone charge using the charging information , then informs the respective telephone charges to the mobile communication exchange 120 . here , the management of the telephone charges is effected in such a manner that if thirty days elapses , the charging center 130 calculates the accumulated telephone charge for the calls made during the specific period for each mobile terminal subscriber , and notifies the respective mobile terminal subscriber of the total charge via mail . the bill includes a basic charge , a tax charge , and the calculated telephone charge . also , the mobile communication exchange 120 transmits to the mobile terminal 100 the telephone charging information , such as the telephone charge for the latest call , the accumulated telephone charge , and the total telephone charges that are transferred from the charging center 130 . the mobile terminal 100 receives the telephone charging information and displays the information on the display section ( not illustrated ), so that the mobile terminal subscriber can identify it . the mobile communication exchange 120 may be provided with a short message generating section 140 . this short message generating section 140 generates a short message which corresponds to the telephone charging information , such as the telephone charge for the latest call , the accumulated telephone charge , and the total telephone charge that are transferred from the charging center 130 in the form of a short message . the types of telephone charges transferred from the charging center 130 can be selectively determined according to the request by the mobile terminal subscriber . thus , different telephone charge information including the latest call , the accumulated telephone charge , and the total telephone charge can be provided to the subscriber . for instance , the mobile terminal subscriber can be provided with the accumulated telephone charge for a duration of a specified time period that he / she desires from the charging center 130 using the mobile terminal 100 . the type and the different telephone charge information transferred to the mobile terminal 100 may be predetermined or determined by the selection command from the mobile terminal subscriber . for example , the display unit of the mobile terminal can be pre - selected to display only the telephone charge for the latest call or only the total telephone charge , or both the latest call and the total telephone charge . fig3 is a flowchart illustrating a method of informing telephone charges to a mobile terminal for a telephone call and a total telephone charge call when the telephone call terminates according to the embodiment of the present invention . with reference to fig3 , the mobile communication exchange 120 detects whether the telephone call from the mobile terminal 100 terminates ( step 210 ). if the telephone call terminates , the mobile communication exchange 120 informs the charging center 130 of the charging information for the call , such as subscriber number , terminating number , call start time , call termination time , discount information , and so forth ( step 220 ). the charging center 130 calculates the charge for the latest call and adds the calculated charge to the total telephone charge ( step 230 ). the total telephone charge is calculated by adding up the accumulated telephone charge for each mobile terminal as well as the basic charge and the tax charge within a specified time period , for instance , a thirty - day period . the charging center 130 informs the mobile communication exchange 120 of the charge information for the latest call as well as the total charge ( step 240 ). the mobile communication exchange 120 receives the charge information for the latest call and the total charge , converts the received information in the form of a short message using the short message generating section 140 , and transmits the generated short message to the mobile terminal subscriber 100 ( steps 250 and 260 ). the mobile terminal 100 then receives the short message and displays the charge information of the latest call and the total charge on the lcd of the mobile phone . preferably , it may display characters such as \u201c charge for the latest call : 400 won , total charge : 10 , 500 won \u201d. accordingly , the mobile terminal subscriber can immediately confirm the charge for the latest call as well as the total charge calculated to include the latest call fig4 is a flowchart illustrating a method of informing a mobile terminal of the total telephone charge in the case that the mobile terminal requests confirmation of the total telephone charges up to now in a standby state according to another embodiment of the present invention . with reference to fig4 , if the mobile terminal subscriber intends to verify the total telephone charge up to now in a standby state , he / she can request for the total charge , for example , by sequentially pressing keys \u201c*\u201d, \u201c 1 \u201d, \u201c 1 \u201d, and \u201c send \u201d, or a specified key ( step 310 ). if the mobile communication exchange 120 receives the request for the confirmation of the total charge for the mobile terminal 100 through the base station 110 , the exchange 120 in turn requests the confirmation of the total charge of the mobile terminal subscriber to the charging center 130 ( step 320 ). the charging center 130 searches the total charging information corresponding to the requesting mobile terminal subscriber and informs the mobile communication exchange 120 of the total charging information ( step 330 ). the mobile communication exchange 120 generates a short message corresponding to the total charging information through the short message generating section 140 and transmits the short message to the mobile terminal 100 ( steps 340 and 350 ). the mobile terminal 100 receives the short message , and displays the total charge on the lcd . preferably , it may display characters such as \u201c total charge : 10 , 500 won ( i . e . korean currency )\u201d. as described above , it will be apparent that the present invention provides advantages in that the mobile terminal subscriber can be provided with an accurate telephone charge information by enabling the subscriber to immediately verify the charging information calculated by the charging center through the mobile terminal just after a telephone call . also , the mobile terminal subscriber can selectively verify at least one type of charging information out of various charging information managed by the charging center as occasion demands using the mobile terminal . while this invention has been described in connection with what is presently considered to be the most practical and preferred embodiments which display both the charge for the latest call and the total charge or only the total charge , it is to be understood that other modifications thereof may be made without departing from the scope of the invention . for example , the accumulated telephone charge for a specified time period desired by the mobile terminal subscriber , or another type of charge information other than the latest call and the total charge can be selectively displayed . thus , the present invention should not be limited to the disclosed embodiment but should be defined by the scope of the appended claims and their equivalents ."}
|
{"category": "General tagging of new or cross-sectional technology", "patent": "the preferred embodiments of the present invention will now be explained in detail with reference to the accompanying drawings . in the drawings , the same or similar elements are denoted by the same reference numerals even though they are depicted in different drawings . for the purpose of clarity , a detailed description of well known functions and configurations incorporated herein will be omitted as it may make the subject matter of the present invention unclear . fig2 is a block diagram illustrating the construction of a mobile radio communication system incorporating the present invention . referring to fig2 , a mobile communication exchange 120 performs an exchange function and interacts with another mobile communication , or the exchange 120 connects with a different communication network , such as a psdn , and controls the call termination / origination of a mobile terminal 100 . in addition , the mobile communication exchange 120 informs the charging center 130 of charging information of a telephone call , such as the subscriber &# 39 ; s number , terminating number , call start time , call termination time , discount information and so forth , when the call from the mobile terminal 100 terminates . the charging center 130 calculates and manages a telephone charge for the latest call , an accumulated telephone charge , and a total telephone charge using the charging information , then informs the respective telephone charges to the mobile communication exchange 120 . here , the management of the telephone charges is effected in such a manner that if thirty days elapses , the charging center 130 calculates the accumulated telephone charge for the calls made during the specific period for each mobile terminal subscriber , and notifies the respective mobile terminal subscriber of the total charge via mail . the bill includes a basic charge , a tax charge , and the calculated telephone charge . also , the mobile communication exchange 120 transmits to the mobile terminal 100 the telephone charging information , such as the telephone charge for the latest call , the accumulated telephone charge , and the total telephone charges that are transferred from the charging center 130 . the mobile terminal 100 receives the telephone charging information and displays the information on the display section ( not illustrated ), so that the mobile terminal subscriber can identify it . the mobile communication exchange 120 may be provided with a short message generating section 140 . this short message generating section 140 generates a short message which corresponds to the telephone charging information , such as the telephone charge for the latest call , the accumulated telephone charge , and the total telephone charge that are transferred from the charging center 130 in the form of a short message . the types of telephone charges transferred from the charging center 130 can be selectively determined according to the request by the mobile terminal subscriber . thus , different telephone charge information including the latest call , the accumulated telephone charge , and the total telephone charge can be provided to the subscriber . for instance , the mobile terminal subscriber can be provided with the accumulated telephone charge for a duration of a specified time period that he / she desires from the charging center 130 using the mobile terminal 100 . the type and the different telephone charge information transferred to the mobile terminal 100 may be predetermined or determined by the selection command from the mobile terminal subscriber . for example , the display unit of the mobile terminal can be pre - selected to display only the telephone charge for the latest call or only the total telephone charge , or both the latest call and the total telephone charge . fig3 is a flowchart illustrating a method of informing telephone charges to a mobile terminal for a telephone call and a total telephone charge call when the telephone call terminates according to the embodiment of the present invention . with reference to fig3 , the mobile communication exchange 120 detects whether the telephone call from the mobile terminal 100 terminates ( step 210 ). if the telephone call terminates , the mobile communication exchange 120 informs the charging center 130 of the charging information for the call , such as subscriber number , terminating number , call start time , call termination time , discount information , and so forth ( step 220 ). the charging center 130 calculates the charge for the latest call and adds the calculated charge to the total telephone charge ( step 230 ). the total telephone charge is calculated by adding up the accumulated telephone charge for each mobile terminal as well as the basic charge and the tax charge within a specified time period , for instance , a thirty - day period . the charging center 130 informs the mobile communication exchange 120 of the charge information for the latest call as well as the total charge ( step 240 ). the mobile communication exchange 120 receives the charge information for the latest call and the total charge , converts the received information in the form of a short message using the short message generating section 140 , and transmits the generated short message to the mobile terminal subscriber 100 ( steps 250 and 260 ). the mobile terminal 100 then receives the short message and displays the charge information of the latest call and the total charge on the lcd of the mobile phone . preferably , it may display characters such as \u201c charge for the latest call : 400 won , total charge : 10 , 500 won \u201d. accordingly , the mobile terminal subscriber can immediately confirm the charge for the latest call as well as the total charge calculated to include the latest call fig4 is a flowchart illustrating a method of informing a mobile terminal of the total telephone charge in the case that the mobile terminal requests confirmation of the total telephone charges up to now in a standby state according to another embodiment of the present invention . with reference to fig4 , if the mobile terminal subscriber intends to verify the total telephone charge up to now in a standby state , he / she can request for the total charge , for example , by sequentially pressing keys \u201c*\u201d, \u201c 1 \u201d, \u201c 1 \u201d, and \u201c send \u201d, or a specified key ( step 310 ). if the mobile communication exchange 120 receives the request for the confirmation of the total charge for the mobile terminal 100 through the base station 110 , the exchange 120 in turn requests the confirmation of the total charge of the mobile terminal subscriber to the charging center 130 ( step 320 ). the charging center 130 searches the total charging information corresponding to the requesting mobile terminal subscriber and informs the mobile communication exchange 120 of the total charging information ( step 330 ). the mobile communication exchange 120 generates a short message corresponding to the total charging information through the short message generating section 140 and transmits the short message to the mobile terminal 100 ( steps 340 and 350 ). the mobile terminal 100 receives the short message , and displays the total charge on the lcd . preferably , it may display characters such as \u201c total charge : 10 , 500 won ( i . e . korean currency )\u201d. as described above , it will be apparent that the present invention provides advantages in that the mobile terminal subscriber can be provided with an accurate telephone charge information by enabling the subscriber to immediately verify the charging information calculated by the charging center through the mobile terminal just after a telephone call . also , the mobile terminal subscriber can selectively verify at least one type of charging information out of various charging information managed by the charging center as occasion demands using the mobile terminal . while this invention has been described in connection with what is presently considered to be the most practical and preferred embodiments which display both the charge for the latest call and the total charge or only the total charge , it is to be understood that other modifications thereof may be made without departing from the scope of the invention . for example , the accumulated telephone charge for a specified time period desired by the mobile terminal subscriber , or another type of charge information other than the latest call and the total charge can be selectively displayed . thus , the present invention should not be limited to the disclosed embodiment but should be defined by the scope of the appended claims and their equivalents ."}
|
Is the category the most suitable category for the given patent?
| 0.25 |
c54693020cd566de34695abc4d1ab9b0bb7c2dfdf03740340864fefaee1a46f1
| 0.53125 | 0.083984 | 0.040283 | 0.014038 | 0.304688 | 0.031738 |
null |
{"patent": "all the figures are schematic illustrations from above the head of a user or observer of each device . the left side of the observer is the left side in each figure . in all the figures , identical references designate parts or elements of parts having identical or similar functions . the figures refer to a device cooperating with the left eye of an observer . a device may be symmetrically provided to cooperate with the right eye of an observer . a device according to the invention may also comprise both a device cooperating with the left eye and a device cooperating with the right eye of the observer . fig1 illustrates an optical device 101 of the prior art , based on a method using the stigmatism of two foci of a substantially elliptical dioptric surface . the device , in the order of the optical pathway , chiefly consists of : a light display 1 ; lenses 2 ; and an ocular dioptric surface 3 of substantially elliptical shape represented by a portion of ellipse defined by its major axis \u03b4 , its small axis \u03b4 \u2032, its centre o and its two foci f and f \u2032 located on the major axis \u03b4 , either side of the centre o , as a function of the eccentricity e of the ellipse . in the figures , the axes \u03b4 , \u03b4 \u2032 of the dioptric surface 3 are shown as dot - dashed lines and the central optical path \u03b4 \u2033 is shown as a dotted line . in the example illustrated in fig1 , the device 101 is designed to cooperate with the observer &# 39 ; s left eye 6 . it is arranged on the left side of the observer &# 39 ; s head 7 . in addition , the major axis \u03b4 passes symmetrically through the two eyes of the observer and it is therefore perpendicular to a median plane p of the head 7 . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the central optical path \u03b4 \u2033 passes through one f of the foci of the dioptric surface then , after being reflected on said dioptric surface 3 , through the other focus f \u2032. the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. the spatial arrangements of the parts of device 101 in relation to the central optical path \u03b4 \u2033, on which they are aligned and centred , are inherent in the arrangement of the central optical path \u03b4 \u2033, which necessarily passes through the foci f and f \u2032. it is also ascertained that to observe this arrangement logic , the ocular dioptric surface 3 does not follow the periphery of the observer &# 39 ; s head at the height of the eye as does a conventional pair of spectacles . therefore the centre o of the dioptric surface is located largely outside , on the left of the left eye 6 . in addition the part 1 , 2 of the device upstream of the focus f moves significantly away from the left side of the observer &# 39 ; s head 7 over the distance between the focus f and the display 1 . this positioning of the device 101 in relation to the observer &# 39 ; s head 7 makes the device laterally bulky and of scarcely pleasing design . the device 102 illustrated in fig2 has more reduced lateral bulk than the device in fig1 . to reduce this bulkiness a side fold mirror 4 is arranged on the central optical path \u03b4 \u2033 in the vicinity of the focus f so that it is possible upstream of the mirror 4 to fold or direct the central optical path \u03b4 \u2033 substantially parallel to the median plane p and perpendicular to the major axis \u03b4 . therefore the device 102 in the order of the optical pathway is formed of : a light display 1 ; two groups of lenses 2 ; a side fold mirror 4 of planar , concave or convex shape ; and an ocular dioptric surface 3 of substantially elliptical shape represented by a portion of ellipse e defined by its major axis \u03b4 , its small axis \u03b4 \u2032, its centre o and its two foci f and f \u2032 located on the major axis \u03b4 either side of the centre o , as a function of the eccentricity e of the ellipse . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the side fold mirror 4 is arranged in the vicinity of the focus f so that it reflects the central optical path \u03b4 \u2033 at a chosen angle in the direction of the ocular dioptric surface 3 , so that the central optical path is then reflected in the direction of the observer &# 39 ; s eye , substantially perpendicular to the major axis \u03b4 . the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. in this configuration the bulkiness is significantly reduced since the central optical path \u03b4 \u2033 is folded back along the side of the observer &# 39 ; s head 7 . it is to be noted that the side fold mirror 4 in the example in fig2 is positioned in the immediate vicinity of the focus f on which it can be directed . the positioning thereof , so close to the focus f , is made necessary by the fact that it can offer a surface of minimum reflection . however the side fold mirror 4 can be arranged elsewhere on the central optical central path \u03b4 \u2033 downstream or upstream of the focus f according to the needs of the optical design . a description will now be given with reference to fig3 and 4 of two embodiments of a device according to the invention , in how they differ from the previously illustrated prior art . fig3 is an illustration of a first embodiment of a device according to the invention . the device 103 in fig3 , in the order of the optical pathway , is formed of : a light display 1 ; two groups of lenses 2 ; a side fold mirror 4 ; a back - inverting mirror 5 ; and an ocular dioptric surface 3 of substantially elliptical shape . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the side fold mirror 4 is arranged laterally in the vicinity of the observer &# 39 ; s eye 6 , in front of the observer &# 39 ; s temple , and it reflects the central optical path \u03b4 \u2033 at a chosen angle in the direction of the back - inverting mirror 5 . the back - inverting mirror 5 is arranged in the region of the upper part of the observer &# 39 ; s nose 10 on the right of and at the height of the left eye 6 and again reflects the central optical path \u03b4 \u2033 towards the ocular dioptric surface 3 . the mirrors are arranged such that the central optical path is then reflected by the dioptric surface in the direction of the observer &# 39 ; s eye 6 substantially perpendicular to the major axis \u03b4 . the observer &# 39 ; s eye 6 i . e . the centre of the pupil of the eye is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. according to the new optical scheme of the invention , the foci f and f \u2032 are inverted relative to the observer &# 39 ; s head . the back - inverting mirror 5 located in the vicinity of the focus f allows the virtual placing of that part of the central optical path \u03b4 \u2033 that is incident on the dioptric surface , and the focus f , in the observer &# 39 ; s head . that is to say that the display 1 is virtually placed inside the head 7 . yet in reality the central optical path \u03b4 \u2033 originates from the side of the observer &# 39 ; s head where the central optical path \u03b4 \u2033 was first folded or re - directed by the side fold mirror 4 . the side fold mirror 4 may now be located more distant from the focus f . the focus f is now directly related to the back - inverting mirror 5 . this gives the side fold mirror a much wider range of positioning , and hence of adjustment , the ocular dioptric surface 3 is now better adjusted to the morphological profile , curve , of the observer &# 39 ; s head in the vicinity of the eye , since the outer profile of the said ocular dioptric surface 3 tends to draw close to the side of the observer &# 39 ; s head 7 . the back - inverting mirror 5 is arranged in the vicinity of the upper part of the observer &# 39 ; s nose 10 on the pathway of the central optic path \u03b4 \u2033 between the side fold mirror and the ocular dioptric surface 3 . therefore the back - inverting mirror 5 can be arranged in an area hidden from the view of the observer , called a \u201c blind spot \u201d. the side fold mirror 4 is oriented angularly so that the central optic path \u03b4 \u2033 of the image is returned or reflected back towards the back - inverting mirror 5 and so that the back - inverting mirror 5 is oriented such that the central optic path \u03b4 \u2033 of the image is then returned or reflected towards the ocular dioptric surface 3 , the dioptric surface 3 being oriented so that the central optic path \u03b4 \u2033 of the image is then returned or reflected towards the observer &# 39 ; s eye 6 . in the embodiment in fig4 , the device 104 according to the invention is formed of : a light display 1 ; lenses 2 ; a side fold mirror 4 ; two decentring mirrors 8 and 9 arranged upstream of the fold mirror 4 , so that the they successively reflect the central optic path \u03b4 \u2033; a back - inverting mirror 5 of planar , convex or concave or aspherical shape ; an ocular dioptric surface 3 of substantially elliptical shape . the light display 1 diffuses an image whose pathway is illustrated by a central optic path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optic path \u03b4 \u2033. the central optic path is then reflected towards the right by the first decentring mirror 8 in the direction of the second decentring mirror 9 which is arranged in the vicinity of the observer &# 39 ; s temple . the second decentring mirror 9 then reflects the central path substantially forwardly in the direction of the side fold mirror 4 , which then reflects the same towards the right in the direction of the back - inverting mirror 5 . the back - inverting mirror 5 is arranged substantially against the upper left part of the observer &# 39 ; s nose 10 substantially at the height of the eye 6 . the central optic path \u03b4 \u2033 is then reflected thereat in the direction of the dioptric surface 3 which reflects it back towards the observer &# 39 ; s eye 6 . the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the region of the focus f \u2032. the device 104 in fig4 differs from the device 103 described with reference to fig3 in that it comprises a set of decentring mirrors 8 and 9 . the use of the set of decentring mirrors 8 , 9 has the following advantages : it allows correction of the adjustment of the angle reflection of the optical path \u03b4 \u2033 on the side mirror 4 , to offset positioning of the said side fold mirror 4 when it is brought as close as possible to the observer &# 39 ; s temple or eye , or as close as possible to the ocular dioptric surface 3 , so that it can optionally be integrated therein , for this purpose , it allows a device to be produced that is better adapted to the observer &# 39 ; s morphology since the pathway of the path \u03b4 \u2033 is close to the curve of the observer &# 39 ; s head . in addition , to make the device 103 in fig3 or the device 104 in fig4 more compact , the side fold mirror 4 can be made in a single piece with the ocular dioptric surface 3 . therefore the side fold mirror 4 can be an integral part of the outer end i . e . in the illustrated example of the left end of the ocular dioptric surface 3 . for example , each of the side fold , reflective , back - inverting and / or decentring mirrors may be planar , concave or convex or substantially aspherical in a manner making it possible to improve the overall quality of the optical system . preferably a device of the invention comprises adjustment means , in particular sets of mirror and an ocular dioptric surface capable of adapting the configuration of the device , in particular the optic path , to the observer &# 39 ; s morphology .", "category": "Physics"}
|
{"patent": "all the figures are schematic illustrations from above the head of a user or observer of each device . the left side of the observer is the left side in each figure . in all the figures , identical references designate parts or elements of parts having identical or similar functions . the figures refer to a device cooperating with the left eye of an observer . a device may be symmetrically provided to cooperate with the right eye of an observer . a device according to the invention may also comprise both a device cooperating with the left eye and a device cooperating with the right eye of the observer . fig1 illustrates an optical device 101 of the prior art , based on a method using the stigmatism of two foci of a substantially elliptical dioptric surface . the device , in the order of the optical pathway , chiefly consists of : a light display 1 ; lenses 2 ; and an ocular dioptric surface 3 of substantially elliptical shape represented by a portion of ellipse defined by its major axis \u03b4 , its small axis \u03b4 \u2032, its centre o and its two foci f and f \u2032 located on the major axis \u03b4 , either side of the centre o , as a function of the eccentricity e of the ellipse . in the figures , the axes \u03b4 , \u03b4 \u2032 of the dioptric surface 3 are shown as dot - dashed lines and the central optical path \u03b4 \u2033 is shown as a dotted line . in the example illustrated in fig1 , the device 101 is designed to cooperate with the observer &# 39 ; s left eye 6 . it is arranged on the left side of the observer &# 39 ; s head 7 . in addition , the major axis \u03b4 passes symmetrically through the two eyes of the observer and it is therefore perpendicular to a median plane p of the head 7 . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the central optical path \u03b4 \u2033 passes through one f of the foci of the dioptric surface then , after being reflected on said dioptric surface 3 , through the other focus f \u2032. the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. the spatial arrangements of the parts of device 101 in relation to the central optical path \u03b4 \u2033, on which they are aligned and centred , are inherent in the arrangement of the central optical path \u03b4 \u2033, which necessarily passes through the foci f and f \u2032. it is also ascertained that to observe this arrangement logic , the ocular dioptric surface 3 does not follow the periphery of the observer &# 39 ; s head at the height of the eye as does a conventional pair of spectacles . therefore the centre o of the dioptric surface is located largely outside , on the left of the left eye 6 . in addition the part 1 , 2 of the device upstream of the focus f moves significantly away from the left side of the observer &# 39 ; s head 7 over the distance between the focus f and the display 1 . this positioning of the device 101 in relation to the observer &# 39 ; s head 7 makes the device laterally bulky and of scarcely pleasing design . the device 102 illustrated in fig2 has more reduced lateral bulk than the device in fig1 . to reduce this bulkiness a side fold mirror 4 is arranged on the central optical path \u03b4 \u2033 in the vicinity of the focus f so that it is possible upstream of the mirror 4 to fold or direct the central optical path \u03b4 \u2033 substantially parallel to the median plane p and perpendicular to the major axis \u03b4 . therefore the device 102 in the order of the optical pathway is formed of : a light display 1 ; two groups of lenses 2 ; a side fold mirror 4 of planar , concave or convex shape ; and an ocular dioptric surface 3 of substantially elliptical shape represented by a portion of ellipse e defined by its major axis \u03b4 , its small axis \u03b4 \u2032, its centre o and its two foci f and f \u2032 located on the major axis \u03b4 either side of the centre o , as a function of the eccentricity e of the ellipse . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the side fold mirror 4 is arranged in the vicinity of the focus f so that it reflects the central optical path \u03b4 \u2033 at a chosen angle in the direction of the ocular dioptric surface 3 , so that the central optical path is then reflected in the direction of the observer &# 39 ; s eye , substantially perpendicular to the major axis \u03b4 . the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. in this configuration the bulkiness is significantly reduced since the central optical path \u03b4 \u2033 is folded back along the side of the observer &# 39 ; s head 7 . it is to be noted that the side fold mirror 4 in the example in fig2 is positioned in the immediate vicinity of the focus f on which it can be directed . the positioning thereof , so close to the focus f , is made necessary by the fact that it can offer a surface of minimum reflection . however the side fold mirror 4 can be arranged elsewhere on the central optical central path \u03b4 \u2033 downstream or upstream of the focus f according to the needs of the optical design . a description will now be given with reference to fig3 and 4 of two embodiments of a device according to the invention , in how they differ from the previously illustrated prior art . fig3 is an illustration of a first embodiment of a device according to the invention . the device 103 in fig3 , in the order of the optical pathway , is formed of : a light display 1 ; two groups of lenses 2 ; a side fold mirror 4 ; a back - inverting mirror 5 ; and an ocular dioptric surface 3 of substantially elliptical shape . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the side fold mirror 4 is arranged laterally in the vicinity of the observer &# 39 ; s eye 6 , in front of the observer &# 39 ; s temple , and it reflects the central optical path \u03b4 \u2033 at a chosen angle in the direction of the back - inverting mirror 5 . the back - inverting mirror 5 is arranged in the region of the upper part of the observer &# 39 ; s nose 10 on the right of and at the height of the left eye 6 and again reflects the central optical path \u03b4 \u2033 towards the ocular dioptric surface 3 . the mirrors are arranged such that the central optical path is then reflected by the dioptric surface in the direction of the observer &# 39 ; s eye 6 substantially perpendicular to the major axis \u03b4 . the observer &# 39 ; s eye 6 i . e . the centre of the pupil of the eye is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. according to the new optical scheme of the invention , the foci f and f \u2032 are inverted relative to the observer &# 39 ; s head . the back - inverting mirror 5 located in the vicinity of the focus f allows the virtual placing of that part of the central optical path \u03b4 \u2033 that is incident on the dioptric surface , and the focus f , in the observer &# 39 ; s head . that is to say that the display 1 is virtually placed inside the head 7 . yet in reality the central optical path \u03b4 \u2033 originates from the side of the observer &# 39 ; s head where the central optical path \u03b4 \u2033 was first folded or re - directed by the side fold mirror 4 . the side fold mirror 4 may now be located more distant from the focus f . the focus f is now directly related to the back - inverting mirror 5 . this gives the side fold mirror a much wider range of positioning , and hence of adjustment , the ocular dioptric surface 3 is now better adjusted to the morphological profile , curve , of the observer &# 39 ; s head in the vicinity of the eye , since the outer profile of the said ocular dioptric surface 3 tends to draw close to the side of the observer &# 39 ; s head 7 . the back - inverting mirror 5 is arranged in the vicinity of the upper part of the observer &# 39 ; s nose 10 on the pathway of the central optic path \u03b4 \u2033 between the side fold mirror and the ocular dioptric surface 3 . therefore the back - inverting mirror 5 can be arranged in an area hidden from the view of the observer , called a \u201c blind spot \u201d. the side fold mirror 4 is oriented angularly so that the central optic path \u03b4 \u2033 of the image is returned or reflected back towards the back - inverting mirror 5 and so that the back - inverting mirror 5 is oriented such that the central optic path \u03b4 \u2033 of the image is then returned or reflected towards the ocular dioptric surface 3 , the dioptric surface 3 being oriented so that the central optic path \u03b4 \u2033 of the image is then returned or reflected towards the observer &# 39 ; s eye 6 . in the embodiment in fig4 , the device 104 according to the invention is formed of : a light display 1 ; lenses 2 ; a side fold mirror 4 ; two decentring mirrors 8 and 9 arranged upstream of the fold mirror 4 , so that the they successively reflect the central optic path \u03b4 \u2033; a back - inverting mirror 5 of planar , convex or concave or aspherical shape ; an ocular dioptric surface 3 of substantially elliptical shape . the light display 1 diffuses an image whose pathway is illustrated by a central optic path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optic path \u03b4 \u2033. the central optic path is then reflected towards the right by the first decentring mirror 8 in the direction of the second decentring mirror 9 which is arranged in the vicinity of the observer &# 39 ; s temple . the second decentring mirror 9 then reflects the central path substantially forwardly in the direction of the side fold mirror 4 , which then reflects the same towards the right in the direction of the back - inverting mirror 5 . the back - inverting mirror 5 is arranged substantially against the upper left part of the observer &# 39 ; s nose 10 substantially at the height of the eye 6 . the central optic path \u03b4 \u2033 is then reflected thereat in the direction of the dioptric surface 3 which reflects it back towards the observer &# 39 ; s eye 6 . the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the region of the focus f \u2032. the device 104 in fig4 differs from the device 103 described with reference to fig3 in that it comprises a set of decentring mirrors 8 and 9 . the use of the set of decentring mirrors 8 , 9 has the following advantages : it allows correction of the adjustment of the angle reflection of the optical path \u03b4 \u2033 on the side mirror 4 , to offset positioning of the said side fold mirror 4 when it is brought as close as possible to the observer &# 39 ; s temple or eye , or as close as possible to the ocular dioptric surface 3 , so that it can optionally be integrated therein , for this purpose , it allows a device to be produced that is better adapted to the observer &# 39 ; s morphology since the pathway of the path \u03b4 \u2033 is close to the curve of the observer &# 39 ; s head . in addition , to make the device 103 in fig3 or the device 104 in fig4 more compact , the side fold mirror 4 can be made in a single piece with the ocular dioptric surface 3 . therefore the side fold mirror 4 can be an integral part of the outer end i . e . in the illustrated example of the left end of the ocular dioptric surface 3 . for example , each of the side fold , reflective , back - inverting and / or decentring mirrors may be planar , concave or convex or substantially aspherical in a manner making it possible to improve the overall quality of the optical system . preferably a device of the invention comprises adjustment means , in particular sets of mirror and an ocular dioptric surface capable of adapting the configuration of the device , in particular the optic path , to the observer &# 39 ; s morphology .", "category": "Human Necessities"}
|
Is the category the most suitable category for the given patent?
| 0.25 |
ab7dbe7acb7f81dc7aa04494c82020cf62dbf0961bcc8e2034548c678553ff66
| 0.326172 | 0.029785 | 0.141602 | 0.098145 | 0.345703 | 0.15625 |
null |
{"patent": "all the figures are schematic illustrations from above the head of a user or observer of each device . the left side of the observer is the left side in each figure . in all the figures , identical references designate parts or elements of parts having identical or similar functions . the figures refer to a device cooperating with the left eye of an observer . a device may be symmetrically provided to cooperate with the right eye of an observer . a device according to the invention may also comprise both a device cooperating with the left eye and a device cooperating with the right eye of the observer . fig1 illustrates an optical device 101 of the prior art , based on a method using the stigmatism of two foci of a substantially elliptical dioptric surface . the device , in the order of the optical pathway , chiefly consists of : a light display 1 ; lenses 2 ; and an ocular dioptric surface 3 of substantially elliptical shape represented by a portion of ellipse defined by its major axis \u03b4 , its small axis \u03b4 \u2032, its centre o and its two foci f and f \u2032 located on the major axis \u03b4 , either side of the centre o , as a function of the eccentricity e of the ellipse . in the figures , the axes \u03b4 , \u03b4 \u2032 of the dioptric surface 3 are shown as dot - dashed lines and the central optical path \u03b4 \u2033 is shown as a dotted line . in the example illustrated in fig1 , the device 101 is designed to cooperate with the observer &# 39 ; s left eye 6 . it is arranged on the left side of the observer &# 39 ; s head 7 . in addition , the major axis \u03b4 passes symmetrically through the two eyes of the observer and it is therefore perpendicular to a median plane p of the head 7 . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the central optical path \u03b4 \u2033 passes through one f of the foci of the dioptric surface then , after being reflected on said dioptric surface 3 , through the other focus f \u2032. the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. the spatial arrangements of the parts of device 101 in relation to the central optical path \u03b4 \u2033, on which they are aligned and centred , are inherent in the arrangement of the central optical path \u03b4 \u2033, which necessarily passes through the foci f and f \u2032. it is also ascertained that to observe this arrangement logic , the ocular dioptric surface 3 does not follow the periphery of the observer &# 39 ; s head at the height of the eye as does a conventional pair of spectacles . therefore the centre o of the dioptric surface is located largely outside , on the left of the left eye 6 . in addition the part 1 , 2 of the device upstream of the focus f moves significantly away from the left side of the observer &# 39 ; s head 7 over the distance between the focus f and the display 1 . this positioning of the device 101 in relation to the observer &# 39 ; s head 7 makes the device laterally bulky and of scarcely pleasing design . the device 102 illustrated in fig2 has more reduced lateral bulk than the device in fig1 . to reduce this bulkiness a side fold mirror 4 is arranged on the central optical path \u03b4 \u2033 in the vicinity of the focus f so that it is possible upstream of the mirror 4 to fold or direct the central optical path \u03b4 \u2033 substantially parallel to the median plane p and perpendicular to the major axis \u03b4 . therefore the device 102 in the order of the optical pathway is formed of : a light display 1 ; two groups of lenses 2 ; a side fold mirror 4 of planar , concave or convex shape ; and an ocular dioptric surface 3 of substantially elliptical shape represented by a portion of ellipse e defined by its major axis \u03b4 , its small axis \u03b4 \u2032, its centre o and its two foci f and f \u2032 located on the major axis \u03b4 either side of the centre o , as a function of the eccentricity e of the ellipse . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the side fold mirror 4 is arranged in the vicinity of the focus f so that it reflects the central optical path \u03b4 \u2033 at a chosen angle in the direction of the ocular dioptric surface 3 , so that the central optical path is then reflected in the direction of the observer &# 39 ; s eye , substantially perpendicular to the major axis \u03b4 . the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. in this configuration the bulkiness is significantly reduced since the central optical path \u03b4 \u2033 is folded back along the side of the observer &# 39 ; s head 7 . it is to be noted that the side fold mirror 4 in the example in fig2 is positioned in the immediate vicinity of the focus f on which it can be directed . the positioning thereof , so close to the focus f , is made necessary by the fact that it can offer a surface of minimum reflection . however the side fold mirror 4 can be arranged elsewhere on the central optical central path \u03b4 \u2033 downstream or upstream of the focus f according to the needs of the optical design . a description will now be given with reference to fig3 and 4 of two embodiments of a device according to the invention , in how they differ from the previously illustrated prior art . fig3 is an illustration of a first embodiment of a device according to the invention . the device 103 in fig3 , in the order of the optical pathway , is formed of : a light display 1 ; two groups of lenses 2 ; a side fold mirror 4 ; a back - inverting mirror 5 ; and an ocular dioptric surface 3 of substantially elliptical shape . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the side fold mirror 4 is arranged laterally in the vicinity of the observer &# 39 ; s eye 6 , in front of the observer &# 39 ; s temple , and it reflects the central optical path \u03b4 \u2033 at a chosen angle in the direction of the back - inverting mirror 5 . the back - inverting mirror 5 is arranged in the region of the upper part of the observer &# 39 ; s nose 10 on the right of and at the height of the left eye 6 and again reflects the central optical path \u03b4 \u2033 towards the ocular dioptric surface 3 . the mirrors are arranged such that the central optical path is then reflected by the dioptric surface in the direction of the observer &# 39 ; s eye 6 substantially perpendicular to the major axis \u03b4 . the observer &# 39 ; s eye 6 i . e . the centre of the pupil of the eye is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. according to the new optical scheme of the invention , the foci f and f \u2032 are inverted relative to the observer &# 39 ; s head . the back - inverting mirror 5 located in the vicinity of the focus f allows the virtual placing of that part of the central optical path \u03b4 \u2033 that is incident on the dioptric surface , and the focus f , in the observer &# 39 ; s head . that is to say that the display 1 is virtually placed inside the head 7 . yet in reality the central optical path \u03b4 \u2033 originates from the side of the observer &# 39 ; s head where the central optical path \u03b4 \u2033 was first folded or re - directed by the side fold mirror 4 . the side fold mirror 4 may now be located more distant from the focus f . the focus f is now directly related to the back - inverting mirror 5 . this gives the side fold mirror a much wider range of positioning , and hence of adjustment , the ocular dioptric surface 3 is now better adjusted to the morphological profile , curve , of the observer &# 39 ; s head in the vicinity of the eye , since the outer profile of the said ocular dioptric surface 3 tends to draw close to the side of the observer &# 39 ; s head 7 . the back - inverting mirror 5 is arranged in the vicinity of the upper part of the observer &# 39 ; s nose 10 on the pathway of the central optic path \u03b4 \u2033 between the side fold mirror and the ocular dioptric surface 3 . therefore the back - inverting mirror 5 can be arranged in an area hidden from the view of the observer , called a \u201c blind spot \u201d. the side fold mirror 4 is oriented angularly so that the central optic path \u03b4 \u2033 of the image is returned or reflected back towards the back - inverting mirror 5 and so that the back - inverting mirror 5 is oriented such that the central optic path \u03b4 \u2033 of the image is then returned or reflected towards the ocular dioptric surface 3 , the dioptric surface 3 being oriented so that the central optic path \u03b4 \u2033 of the image is then returned or reflected towards the observer &# 39 ; s eye 6 . in the embodiment in fig4 , the device 104 according to the invention is formed of : a light display 1 ; lenses 2 ; a side fold mirror 4 ; two decentring mirrors 8 and 9 arranged upstream of the fold mirror 4 , so that the they successively reflect the central optic path \u03b4 \u2033; a back - inverting mirror 5 of planar , convex or concave or aspherical shape ; an ocular dioptric surface 3 of substantially elliptical shape . the light display 1 diffuses an image whose pathway is illustrated by a central optic path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optic path \u03b4 \u2033. the central optic path is then reflected towards the right by the first decentring mirror 8 in the direction of the second decentring mirror 9 which is arranged in the vicinity of the observer &# 39 ; s temple . the second decentring mirror 9 then reflects the central path substantially forwardly in the direction of the side fold mirror 4 , which then reflects the same towards the right in the direction of the back - inverting mirror 5 . the back - inverting mirror 5 is arranged substantially against the upper left part of the observer &# 39 ; s nose 10 substantially at the height of the eye 6 . the central optic path \u03b4 \u2033 is then reflected thereat in the direction of the dioptric surface 3 which reflects it back towards the observer &# 39 ; s eye 6 . the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the region of the focus f \u2032. the device 104 in fig4 differs from the device 103 described with reference to fig3 in that it comprises a set of decentring mirrors 8 and 9 . the use of the set of decentring mirrors 8 , 9 has the following advantages : it allows correction of the adjustment of the angle reflection of the optical path \u03b4 \u2033 on the side mirror 4 , to offset positioning of the said side fold mirror 4 when it is brought as close as possible to the observer &# 39 ; s temple or eye , or as close as possible to the ocular dioptric surface 3 , so that it can optionally be integrated therein , for this purpose , it allows a device to be produced that is better adapted to the observer &# 39 ; s morphology since the pathway of the path \u03b4 \u2033 is close to the curve of the observer &# 39 ; s head . in addition , to make the device 103 in fig3 or the device 104 in fig4 more compact , the side fold mirror 4 can be made in a single piece with the ocular dioptric surface 3 . therefore the side fold mirror 4 can be an integral part of the outer end i . e . in the illustrated example of the left end of the ocular dioptric surface 3 . for example , each of the side fold , reflective , back - inverting and / or decentring mirrors may be planar , concave or convex or substantially aspherical in a manner making it possible to improve the overall quality of the optical system . preferably a device of the invention comprises adjustment means , in particular sets of mirror and an ocular dioptric surface capable of adapting the configuration of the device , in particular the optic path , to the observer &# 39 ; s morphology .", "category": "Physics"}
|
{"patent": "all the figures are schematic illustrations from above the head of a user or observer of each device . the left side of the observer is the left side in each figure . in all the figures , identical references designate parts or elements of parts having identical or similar functions . the figures refer to a device cooperating with the left eye of an observer . a device may be symmetrically provided to cooperate with the right eye of an observer . a device according to the invention may also comprise both a device cooperating with the left eye and a device cooperating with the right eye of the observer . fig1 illustrates an optical device 101 of the prior art , based on a method using the stigmatism of two foci of a substantially elliptical dioptric surface . the device , in the order of the optical pathway , chiefly consists of : a light display 1 ; lenses 2 ; and an ocular dioptric surface 3 of substantially elliptical shape represented by a portion of ellipse defined by its major axis \u03b4 , its small axis \u03b4 \u2032, its centre o and its two foci f and f \u2032 located on the major axis \u03b4 , either side of the centre o , as a function of the eccentricity e of the ellipse . in the figures , the axes \u03b4 , \u03b4 \u2032 of the dioptric surface 3 are shown as dot - dashed lines and the central optical path \u03b4 \u2033 is shown as a dotted line . in the example illustrated in fig1 , the device 101 is designed to cooperate with the observer &# 39 ; s left eye 6 . it is arranged on the left side of the observer &# 39 ; s head 7 . in addition , the major axis \u03b4 passes symmetrically through the two eyes of the observer and it is therefore perpendicular to a median plane p of the head 7 . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the central optical path \u03b4 \u2033 passes through one f of the foci of the dioptric surface then , after being reflected on said dioptric surface 3 , through the other focus f \u2032. the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. the spatial arrangements of the parts of device 101 in relation to the central optical path \u03b4 \u2033, on which they are aligned and centred , are inherent in the arrangement of the central optical path \u03b4 \u2033, which necessarily passes through the foci f and f \u2032. it is also ascertained that to observe this arrangement logic , the ocular dioptric surface 3 does not follow the periphery of the observer &# 39 ; s head at the height of the eye as does a conventional pair of spectacles . therefore the centre o of the dioptric surface is located largely outside , on the left of the left eye 6 . in addition the part 1 , 2 of the device upstream of the focus f moves significantly away from the left side of the observer &# 39 ; s head 7 over the distance between the focus f and the display 1 . this positioning of the device 101 in relation to the observer &# 39 ; s head 7 makes the device laterally bulky and of scarcely pleasing design . the device 102 illustrated in fig2 has more reduced lateral bulk than the device in fig1 . to reduce this bulkiness a side fold mirror 4 is arranged on the central optical path \u03b4 \u2033 in the vicinity of the focus f so that it is possible upstream of the mirror 4 to fold or direct the central optical path \u03b4 \u2033 substantially parallel to the median plane p and perpendicular to the major axis \u03b4 . therefore the device 102 in the order of the optical pathway is formed of : a light display 1 ; two groups of lenses 2 ; a side fold mirror 4 of planar , concave or convex shape ; and an ocular dioptric surface 3 of substantially elliptical shape represented by a portion of ellipse e defined by its major axis \u03b4 , its small axis \u03b4 \u2032, its centre o and its two foci f and f \u2032 located on the major axis \u03b4 either side of the centre o , as a function of the eccentricity e of the ellipse . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the side fold mirror 4 is arranged in the vicinity of the focus f so that it reflects the central optical path \u03b4 \u2033 at a chosen angle in the direction of the ocular dioptric surface 3 , so that the central optical path is then reflected in the direction of the observer &# 39 ; s eye , substantially perpendicular to the major axis \u03b4 . the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. in this configuration the bulkiness is significantly reduced since the central optical path \u03b4 \u2033 is folded back along the side of the observer &# 39 ; s head 7 . it is to be noted that the side fold mirror 4 in the example in fig2 is positioned in the immediate vicinity of the focus f on which it can be directed . the positioning thereof , so close to the focus f , is made necessary by the fact that it can offer a surface of minimum reflection . however the side fold mirror 4 can be arranged elsewhere on the central optical central path \u03b4 \u2033 downstream or upstream of the focus f according to the needs of the optical design . a description will now be given with reference to fig3 and 4 of two embodiments of a device according to the invention , in how they differ from the previously illustrated prior art . fig3 is an illustration of a first embodiment of a device according to the invention . the device 103 in fig3 , in the order of the optical pathway , is formed of : a light display 1 ; two groups of lenses 2 ; a side fold mirror 4 ; a back - inverting mirror 5 ; and an ocular dioptric surface 3 of substantially elliptical shape . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the side fold mirror 4 is arranged laterally in the vicinity of the observer &# 39 ; s eye 6 , in front of the observer &# 39 ; s temple , and it reflects the central optical path \u03b4 \u2033 at a chosen angle in the direction of the back - inverting mirror 5 . the back - inverting mirror 5 is arranged in the region of the upper part of the observer &# 39 ; s nose 10 on the right of and at the height of the left eye 6 and again reflects the central optical path \u03b4 \u2033 towards the ocular dioptric surface 3 . the mirrors are arranged such that the central optical path is then reflected by the dioptric surface in the direction of the observer &# 39 ; s eye 6 substantially perpendicular to the major axis \u03b4 . the observer &# 39 ; s eye 6 i . e . the centre of the pupil of the eye is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. according to the new optical scheme of the invention , the foci f and f \u2032 are inverted relative to the observer &# 39 ; s head . the back - inverting mirror 5 located in the vicinity of the focus f allows the virtual placing of that part of the central optical path \u03b4 \u2033 that is incident on the dioptric surface , and the focus f , in the observer &# 39 ; s head . that is to say that the display 1 is virtually placed inside the head 7 . yet in reality the central optical path \u03b4 \u2033 originates from the side of the observer &# 39 ; s head where the central optical path \u03b4 \u2033 was first folded or re - directed by the side fold mirror 4 . the side fold mirror 4 may now be located more distant from the focus f . the focus f is now directly related to the back - inverting mirror 5 . this gives the side fold mirror a much wider range of positioning , and hence of adjustment , the ocular dioptric surface 3 is now better adjusted to the morphological profile , curve , of the observer &# 39 ; s head in the vicinity of the eye , since the outer profile of the said ocular dioptric surface 3 tends to draw close to the side of the observer &# 39 ; s head 7 . the back - inverting mirror 5 is arranged in the vicinity of the upper part of the observer &# 39 ; s nose 10 on the pathway of the central optic path \u03b4 \u2033 between the side fold mirror and the ocular dioptric surface 3 . therefore the back - inverting mirror 5 can be arranged in an area hidden from the view of the observer , called a \u201c blind spot \u201d. the side fold mirror 4 is oriented angularly so that the central optic path \u03b4 \u2033 of the image is returned or reflected back towards the back - inverting mirror 5 and so that the back - inverting mirror 5 is oriented such that the central optic path \u03b4 \u2033 of the image is then returned or reflected towards the ocular dioptric surface 3 , the dioptric surface 3 being oriented so that the central optic path \u03b4 \u2033 of the image is then returned or reflected towards the observer &# 39 ; s eye 6 . in the embodiment in fig4 , the device 104 according to the invention is formed of : a light display 1 ; lenses 2 ; a side fold mirror 4 ; two decentring mirrors 8 and 9 arranged upstream of the fold mirror 4 , so that the they successively reflect the central optic path \u03b4 \u2033; a back - inverting mirror 5 of planar , convex or concave or aspherical shape ; an ocular dioptric surface 3 of substantially elliptical shape . the light display 1 diffuses an image whose pathway is illustrated by a central optic path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optic path \u03b4 \u2033. the central optic path is then reflected towards the right by the first decentring mirror 8 in the direction of the second decentring mirror 9 which is arranged in the vicinity of the observer &# 39 ; s temple . the second decentring mirror 9 then reflects the central path substantially forwardly in the direction of the side fold mirror 4 , which then reflects the same towards the right in the direction of the back - inverting mirror 5 . the back - inverting mirror 5 is arranged substantially against the upper left part of the observer &# 39 ; s nose 10 substantially at the height of the eye 6 . the central optic path \u03b4 \u2033 is then reflected thereat in the direction of the dioptric surface 3 which reflects it back towards the observer &# 39 ; s eye 6 . the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the region of the focus f \u2032. the device 104 in fig4 differs from the device 103 described with reference to fig3 in that it comprises a set of decentring mirrors 8 and 9 . the use of the set of decentring mirrors 8 , 9 has the following advantages : it allows correction of the adjustment of the angle reflection of the optical path \u03b4 \u2033 on the side mirror 4 , to offset positioning of the said side fold mirror 4 when it is brought as close as possible to the observer &# 39 ; s temple or eye , or as close as possible to the ocular dioptric surface 3 , so that it can optionally be integrated therein , for this purpose , it allows a device to be produced that is better adapted to the observer &# 39 ; s morphology since the pathway of the path \u03b4 \u2033 is close to the curve of the observer &# 39 ; s head . in addition , to make the device 103 in fig3 or the device 104 in fig4 more compact , the side fold mirror 4 can be made in a single piece with the ocular dioptric surface 3 . therefore the side fold mirror 4 can be an integral part of the outer end i . e . in the illustrated example of the left end of the ocular dioptric surface 3 . for example , each of the side fold , reflective , back - inverting and / or decentring mirrors may be planar , concave or convex or substantially aspherical in a manner making it possible to improve the overall quality of the optical system . preferably a device of the invention comprises adjustment means , in particular sets of mirror and an ocular dioptric surface capable of adapting the configuration of the device , in particular the optic path , to the observer &# 39 ; s morphology .", "category": "Performing Operations; Transporting"}
|
Is the patent correctly categorized?
| 0.25 |
ab7dbe7acb7f81dc7aa04494c82020cf62dbf0961bcc8e2034548c678553ff66
| 0.070801 | 0.021973 | 0.195313 | 0.130859 | 0.178711 | 0.137695 |
null |
{"patent": "all the figures are schematic illustrations from above the head of a user or observer of each device . the left side of the observer is the left side in each figure . in all the figures , identical references designate parts or elements of parts having identical or similar functions . the figures refer to a device cooperating with the left eye of an observer . a device may be symmetrically provided to cooperate with the right eye of an observer . a device according to the invention may also comprise both a device cooperating with the left eye and a device cooperating with the right eye of the observer . fig1 illustrates an optical device 101 of the prior art , based on a method using the stigmatism of two foci of a substantially elliptical dioptric surface . the device , in the order of the optical pathway , chiefly consists of : a light display 1 ; lenses 2 ; and an ocular dioptric surface 3 of substantially elliptical shape represented by a portion of ellipse defined by its major axis \u03b4 , its small axis \u03b4 \u2032, its centre o and its two foci f and f \u2032 located on the major axis \u03b4 , either side of the centre o , as a function of the eccentricity e of the ellipse . in the figures , the axes \u03b4 , \u03b4 \u2032 of the dioptric surface 3 are shown as dot - dashed lines and the central optical path \u03b4 \u2033 is shown as a dotted line . in the example illustrated in fig1 , the device 101 is designed to cooperate with the observer &# 39 ; s left eye 6 . it is arranged on the left side of the observer &# 39 ; s head 7 . in addition , the major axis \u03b4 passes symmetrically through the two eyes of the observer and it is therefore perpendicular to a median plane p of the head 7 . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the central optical path \u03b4 \u2033 passes through one f of the foci of the dioptric surface then , after being reflected on said dioptric surface 3 , through the other focus f \u2032. the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. the spatial arrangements of the parts of device 101 in relation to the central optical path \u03b4 \u2033, on which they are aligned and centred , are inherent in the arrangement of the central optical path \u03b4 \u2033, which necessarily passes through the foci f and f \u2032. it is also ascertained that to observe this arrangement logic , the ocular dioptric surface 3 does not follow the periphery of the observer &# 39 ; s head at the height of the eye as does a conventional pair of spectacles . therefore the centre o of the dioptric surface is located largely outside , on the left of the left eye 6 . in addition the part 1 , 2 of the device upstream of the focus f moves significantly away from the left side of the observer &# 39 ; s head 7 over the distance between the focus f and the display 1 . this positioning of the device 101 in relation to the observer &# 39 ; s head 7 makes the device laterally bulky and of scarcely pleasing design . the device 102 illustrated in fig2 has more reduced lateral bulk than the device in fig1 . to reduce this bulkiness a side fold mirror 4 is arranged on the central optical path \u03b4 \u2033 in the vicinity of the focus f so that it is possible upstream of the mirror 4 to fold or direct the central optical path \u03b4 \u2033 substantially parallel to the median plane p and perpendicular to the major axis \u03b4 . therefore the device 102 in the order of the optical pathway is formed of : a light display 1 ; two groups of lenses 2 ; a side fold mirror 4 of planar , concave or convex shape ; and an ocular dioptric surface 3 of substantially elliptical shape represented by a portion of ellipse e defined by its major axis \u03b4 , its small axis \u03b4 \u2032, its centre o and its two foci f and f \u2032 located on the major axis \u03b4 either side of the centre o , as a function of the eccentricity e of the ellipse . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the side fold mirror 4 is arranged in the vicinity of the focus f so that it reflects the central optical path \u03b4 \u2033 at a chosen angle in the direction of the ocular dioptric surface 3 , so that the central optical path is then reflected in the direction of the observer &# 39 ; s eye , substantially perpendicular to the major axis \u03b4 . the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. in this configuration the bulkiness is significantly reduced since the central optical path \u03b4 \u2033 is folded back along the side of the observer &# 39 ; s head 7 . it is to be noted that the side fold mirror 4 in the example in fig2 is positioned in the immediate vicinity of the focus f on which it can be directed . the positioning thereof , so close to the focus f , is made necessary by the fact that it can offer a surface of minimum reflection . however the side fold mirror 4 can be arranged elsewhere on the central optical central path \u03b4 \u2033 downstream or upstream of the focus f according to the needs of the optical design . a description will now be given with reference to fig3 and 4 of two embodiments of a device according to the invention , in how they differ from the previously illustrated prior art . fig3 is an illustration of a first embodiment of a device according to the invention . the device 103 in fig3 , in the order of the optical pathway , is formed of : a light display 1 ; two groups of lenses 2 ; a side fold mirror 4 ; a back - inverting mirror 5 ; and an ocular dioptric surface 3 of substantially elliptical shape . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the side fold mirror 4 is arranged laterally in the vicinity of the observer &# 39 ; s eye 6 , in front of the observer &# 39 ; s temple , and it reflects the central optical path \u03b4 \u2033 at a chosen angle in the direction of the back - inverting mirror 5 . the back - inverting mirror 5 is arranged in the region of the upper part of the observer &# 39 ; s nose 10 on the right of and at the height of the left eye 6 and again reflects the central optical path \u03b4 \u2033 towards the ocular dioptric surface 3 . the mirrors are arranged such that the central optical path is then reflected by the dioptric surface in the direction of the observer &# 39 ; s eye 6 substantially perpendicular to the major axis \u03b4 . the observer &# 39 ; s eye 6 i . e . the centre of the pupil of the eye is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. according to the new optical scheme of the invention , the foci f and f \u2032 are inverted relative to the observer &# 39 ; s head . the back - inverting mirror 5 located in the vicinity of the focus f allows the virtual placing of that part of the central optical path \u03b4 \u2033 that is incident on the dioptric surface , and the focus f , in the observer &# 39 ; s head . that is to say that the display 1 is virtually placed inside the head 7 . yet in reality the central optical path \u03b4 \u2033 originates from the side of the observer &# 39 ; s head where the central optical path \u03b4 \u2033 was first folded or re - directed by the side fold mirror 4 . the side fold mirror 4 may now be located more distant from the focus f . the focus f is now directly related to the back - inverting mirror 5 . this gives the side fold mirror a much wider range of positioning , and hence of adjustment , the ocular dioptric surface 3 is now better adjusted to the morphological profile , curve , of the observer &# 39 ; s head in the vicinity of the eye , since the outer profile of the said ocular dioptric surface 3 tends to draw close to the side of the observer &# 39 ; s head 7 . the back - inverting mirror 5 is arranged in the vicinity of the upper part of the observer &# 39 ; s nose 10 on the pathway of the central optic path \u03b4 \u2033 between the side fold mirror and the ocular dioptric surface 3 . therefore the back - inverting mirror 5 can be arranged in an area hidden from the view of the observer , called a \u201c blind spot \u201d. the side fold mirror 4 is oriented angularly so that the central optic path \u03b4 \u2033 of the image is returned or reflected back towards the back - inverting mirror 5 and so that the back - inverting mirror 5 is oriented such that the central optic path \u03b4 \u2033 of the image is then returned or reflected towards the ocular dioptric surface 3 , the dioptric surface 3 being oriented so that the central optic path \u03b4 \u2033 of the image is then returned or reflected towards the observer &# 39 ; s eye 6 . in the embodiment in fig4 , the device 104 according to the invention is formed of : a light display 1 ; lenses 2 ; a side fold mirror 4 ; two decentring mirrors 8 and 9 arranged upstream of the fold mirror 4 , so that the they successively reflect the central optic path \u03b4 \u2033; a back - inverting mirror 5 of planar , convex or concave or aspherical shape ; an ocular dioptric surface 3 of substantially elliptical shape . the light display 1 diffuses an image whose pathway is illustrated by a central optic path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optic path \u03b4 \u2033. the central optic path is then reflected towards the right by the first decentring mirror 8 in the direction of the second decentring mirror 9 which is arranged in the vicinity of the observer &# 39 ; s temple . the second decentring mirror 9 then reflects the central path substantially forwardly in the direction of the side fold mirror 4 , which then reflects the same towards the right in the direction of the back - inverting mirror 5 . the back - inverting mirror 5 is arranged substantially against the upper left part of the observer &# 39 ; s nose 10 substantially at the height of the eye 6 . the central optic path \u03b4 \u2033 is then reflected thereat in the direction of the dioptric surface 3 which reflects it back towards the observer &# 39 ; s eye 6 . the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the region of the focus f \u2032. the device 104 in fig4 differs from the device 103 described with reference to fig3 in that it comprises a set of decentring mirrors 8 and 9 . the use of the set of decentring mirrors 8 , 9 has the following advantages : it allows correction of the adjustment of the angle reflection of the optical path \u03b4 \u2033 on the side mirror 4 , to offset positioning of the said side fold mirror 4 when it is brought as close as possible to the observer &# 39 ; s temple or eye , or as close as possible to the ocular dioptric surface 3 , so that it can optionally be integrated therein , for this purpose , it allows a device to be produced that is better adapted to the observer &# 39 ; s morphology since the pathway of the path \u03b4 \u2033 is close to the curve of the observer &# 39 ; s head . in addition , to make the device 103 in fig3 or the device 104 in fig4 more compact , the side fold mirror 4 can be made in a single piece with the ocular dioptric surface 3 . therefore the side fold mirror 4 can be an integral part of the outer end i . e . in the illustrated example of the left end of the ocular dioptric surface 3 . for example , each of the side fold , reflective , back - inverting and / or decentring mirrors may be planar , concave or convex or substantially aspherical in a manner making it possible to improve the overall quality of the optical system . preferably a device of the invention comprises adjustment means , in particular sets of mirror and an ocular dioptric surface capable of adapting the configuration of the device , in particular the optic path , to the observer &# 39 ; s morphology .", "category": "Physics"}
|
{"category": "Chemistry; Metallurgy", "patent": "all the figures are schematic illustrations from above the head of a user or observer of each device . the left side of the observer is the left side in each figure . in all the figures , identical references designate parts or elements of parts having identical or similar functions . the figures refer to a device cooperating with the left eye of an observer . a device may be symmetrically provided to cooperate with the right eye of an observer . a device according to the invention may also comprise both a device cooperating with the left eye and a device cooperating with the right eye of the observer . fig1 illustrates an optical device 101 of the prior art , based on a method using the stigmatism of two foci of a substantially elliptical dioptric surface . the device , in the order of the optical pathway , chiefly consists of : a light display 1 ; lenses 2 ; and an ocular dioptric surface 3 of substantially elliptical shape represented by a portion of ellipse defined by its major axis \u03b4 , its small axis \u03b4 \u2032, its centre o and its two foci f and f \u2032 located on the major axis \u03b4 , either side of the centre o , as a function of the eccentricity e of the ellipse . in the figures , the axes \u03b4 , \u03b4 \u2032 of the dioptric surface 3 are shown as dot - dashed lines and the central optical path \u03b4 \u2033 is shown as a dotted line . in the example illustrated in fig1 , the device 101 is designed to cooperate with the observer &# 39 ; s left eye 6 . it is arranged on the left side of the observer &# 39 ; s head 7 . in addition , the major axis \u03b4 passes symmetrically through the two eyes of the observer and it is therefore perpendicular to a median plane p of the head 7 . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the central optical path \u03b4 \u2033 passes through one f of the foci of the dioptric surface then , after being reflected on said dioptric surface 3 , through the other focus f \u2032. the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. the spatial arrangements of the parts of device 101 in relation to the central optical path \u03b4 \u2033, on which they are aligned and centred , are inherent in the arrangement of the central optical path \u03b4 \u2033, which necessarily passes through the foci f and f \u2032. it is also ascertained that to observe this arrangement logic , the ocular dioptric surface 3 does not follow the periphery of the observer &# 39 ; s head at the height of the eye as does a conventional pair of spectacles . therefore the centre o of the dioptric surface is located largely outside , on the left of the left eye 6 . in addition the part 1 , 2 of the device upstream of the focus f moves significantly away from the left side of the observer &# 39 ; s head 7 over the distance between the focus f and the display 1 . this positioning of the device 101 in relation to the observer &# 39 ; s head 7 makes the device laterally bulky and of scarcely pleasing design . the device 102 illustrated in fig2 has more reduced lateral bulk than the device in fig1 . to reduce this bulkiness a side fold mirror 4 is arranged on the central optical path \u03b4 \u2033 in the vicinity of the focus f so that it is possible upstream of the mirror 4 to fold or direct the central optical path \u03b4 \u2033 substantially parallel to the median plane p and perpendicular to the major axis \u03b4 . therefore the device 102 in the order of the optical pathway is formed of : a light display 1 ; two groups of lenses 2 ; a side fold mirror 4 of planar , concave or convex shape ; and an ocular dioptric surface 3 of substantially elliptical shape represented by a portion of ellipse e defined by its major axis \u03b4 , its small axis \u03b4 \u2032, its centre o and its two foci f and f \u2032 located on the major axis \u03b4 either side of the centre o , as a function of the eccentricity e of the ellipse . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the side fold mirror 4 is arranged in the vicinity of the focus f so that it reflects the central optical path \u03b4 \u2033 at a chosen angle in the direction of the ocular dioptric surface 3 , so that the central optical path is then reflected in the direction of the observer &# 39 ; s eye , substantially perpendicular to the major axis \u03b4 . the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. in this configuration the bulkiness is significantly reduced since the central optical path \u03b4 \u2033 is folded back along the side of the observer &# 39 ; s head 7 . it is to be noted that the side fold mirror 4 in the example in fig2 is positioned in the immediate vicinity of the focus f on which it can be directed . the positioning thereof , so close to the focus f , is made necessary by the fact that it can offer a surface of minimum reflection . however the side fold mirror 4 can be arranged elsewhere on the central optical central path \u03b4 \u2033 downstream or upstream of the focus f according to the needs of the optical design . a description will now be given with reference to fig3 and 4 of two embodiments of a device according to the invention , in how they differ from the previously illustrated prior art . fig3 is an illustration of a first embodiment of a device according to the invention . the device 103 in fig3 , in the order of the optical pathway , is formed of : a light display 1 ; two groups of lenses 2 ; a side fold mirror 4 ; a back - inverting mirror 5 ; and an ocular dioptric surface 3 of substantially elliptical shape . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the side fold mirror 4 is arranged laterally in the vicinity of the observer &# 39 ; s eye 6 , in front of the observer &# 39 ; s temple , and it reflects the central optical path \u03b4 \u2033 at a chosen angle in the direction of the back - inverting mirror 5 . the back - inverting mirror 5 is arranged in the region of the upper part of the observer &# 39 ; s nose 10 on the right of and at the height of the left eye 6 and again reflects the central optical path \u03b4 \u2033 towards the ocular dioptric surface 3 . the mirrors are arranged such that the central optical path is then reflected by the dioptric surface in the direction of the observer &# 39 ; s eye 6 substantially perpendicular to the major axis \u03b4 . the observer &# 39 ; s eye 6 i . e . the centre of the pupil of the eye is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. according to the new optical scheme of the invention , the foci f and f \u2032 are inverted relative to the observer &# 39 ; s head . the back - inverting mirror 5 located in the vicinity of the focus f allows the virtual placing of that part of the central optical path \u03b4 \u2033 that is incident on the dioptric surface , and the focus f , in the observer &# 39 ; s head . that is to say that the display 1 is virtually placed inside the head 7 . yet in reality the central optical path \u03b4 \u2033 originates from the side of the observer &# 39 ; s head where the central optical path \u03b4 \u2033 was first folded or re - directed by the side fold mirror 4 . the side fold mirror 4 may now be located more distant from the focus f . the focus f is now directly related to the back - inverting mirror 5 . this gives the side fold mirror a much wider range of positioning , and hence of adjustment , the ocular dioptric surface 3 is now better adjusted to the morphological profile , curve , of the observer &# 39 ; s head in the vicinity of the eye , since the outer profile of the said ocular dioptric surface 3 tends to draw close to the side of the observer &# 39 ; s head 7 . the back - inverting mirror 5 is arranged in the vicinity of the upper part of the observer &# 39 ; s nose 10 on the pathway of the central optic path \u03b4 \u2033 between the side fold mirror and the ocular dioptric surface 3 . therefore the back - inverting mirror 5 can be arranged in an area hidden from the view of the observer , called a \u201c blind spot \u201d. the side fold mirror 4 is oriented angularly so that the central optic path \u03b4 \u2033 of the image is returned or reflected back towards the back - inverting mirror 5 and so that the back - inverting mirror 5 is oriented such that the central optic path \u03b4 \u2033 of the image is then returned or reflected towards the ocular dioptric surface 3 , the dioptric surface 3 being oriented so that the central optic path \u03b4 \u2033 of the image is then returned or reflected towards the observer &# 39 ; s eye 6 . in the embodiment in fig4 , the device 104 according to the invention is formed of : a light display 1 ; lenses 2 ; a side fold mirror 4 ; two decentring mirrors 8 and 9 arranged upstream of the fold mirror 4 , so that the they successively reflect the central optic path \u03b4 \u2033; a back - inverting mirror 5 of planar , convex or concave or aspherical shape ; an ocular dioptric surface 3 of substantially elliptical shape . the light display 1 diffuses an image whose pathway is illustrated by a central optic path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optic path \u03b4 \u2033. the central optic path is then reflected towards the right by the first decentring mirror 8 in the direction of the second decentring mirror 9 which is arranged in the vicinity of the observer &# 39 ; s temple . the second decentring mirror 9 then reflects the central path substantially forwardly in the direction of the side fold mirror 4 , which then reflects the same towards the right in the direction of the back - inverting mirror 5 . the back - inverting mirror 5 is arranged substantially against the upper left part of the observer &# 39 ; s nose 10 substantially at the height of the eye 6 . the central optic path \u03b4 \u2033 is then reflected thereat in the direction of the dioptric surface 3 which reflects it back towards the observer &# 39 ; s eye 6 . the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the region of the focus f \u2032. the device 104 in fig4 differs from the device 103 described with reference to fig3 in that it comprises a set of decentring mirrors 8 and 9 . the use of the set of decentring mirrors 8 , 9 has the following advantages : it allows correction of the adjustment of the angle reflection of the optical path \u03b4 \u2033 on the side mirror 4 , to offset positioning of the said side fold mirror 4 when it is brought as close as possible to the observer &# 39 ; s temple or eye , or as close as possible to the ocular dioptric surface 3 , so that it can optionally be integrated therein , for this purpose , it allows a device to be produced that is better adapted to the observer &# 39 ; s morphology since the pathway of the path \u03b4 \u2033 is close to the curve of the observer &# 39 ; s head . in addition , to make the device 103 in fig3 or the device 104 in fig4 more compact , the side fold mirror 4 can be made in a single piece with the ocular dioptric surface 3 . therefore the side fold mirror 4 can be an integral part of the outer end i . e . in the illustrated example of the left end of the ocular dioptric surface 3 . for example , each of the side fold , reflective , back - inverting and / or decentring mirrors may be planar , concave or convex or substantially aspherical in a manner making it possible to improve the overall quality of the optical system . preferably a device of the invention comprises adjustment means , in particular sets of mirror and an ocular dioptric surface capable of adapting the configuration of the device , in particular the optic path , to the observer &# 39 ; s morphology ."}
|
Is the category the most suitable category for the given patent?
| 0.25 |
ab7dbe7acb7f81dc7aa04494c82020cf62dbf0961bcc8e2034548c678553ff66
| 0.326172 | 0.002396 | 0.141602 | 0.001503 | 0.345703 | 0.002884 |
null |
{"patent": "all the figures are schematic illustrations from above the head of a user or observer of each device . the left side of the observer is the left side in each figure . in all the figures , identical references designate parts or elements of parts having identical or similar functions . the figures refer to a device cooperating with the left eye of an observer . a device may be symmetrically provided to cooperate with the right eye of an observer . a device according to the invention may also comprise both a device cooperating with the left eye and a device cooperating with the right eye of the observer . fig1 illustrates an optical device 101 of the prior art , based on a method using the stigmatism of two foci of a substantially elliptical dioptric surface . the device , in the order of the optical pathway , chiefly consists of : a light display 1 ; lenses 2 ; and an ocular dioptric surface 3 of substantially elliptical shape represented by a portion of ellipse defined by its major axis \u03b4 , its small axis \u03b4 \u2032, its centre o and its two foci f and f \u2032 located on the major axis \u03b4 , either side of the centre o , as a function of the eccentricity e of the ellipse . in the figures , the axes \u03b4 , \u03b4 \u2032 of the dioptric surface 3 are shown as dot - dashed lines and the central optical path \u03b4 \u2033 is shown as a dotted line . in the example illustrated in fig1 , the device 101 is designed to cooperate with the observer &# 39 ; s left eye 6 . it is arranged on the left side of the observer &# 39 ; s head 7 . in addition , the major axis \u03b4 passes symmetrically through the two eyes of the observer and it is therefore perpendicular to a median plane p of the head 7 . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the central optical path \u03b4 \u2033 passes through one f of the foci of the dioptric surface then , after being reflected on said dioptric surface 3 , through the other focus f \u2032. the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. the spatial arrangements of the parts of device 101 in relation to the central optical path \u03b4 \u2033, on which they are aligned and centred , are inherent in the arrangement of the central optical path \u03b4 \u2033, which necessarily passes through the foci f and f \u2032. it is also ascertained that to observe this arrangement logic , the ocular dioptric surface 3 does not follow the periphery of the observer &# 39 ; s head at the height of the eye as does a conventional pair of spectacles . therefore the centre o of the dioptric surface is located largely outside , on the left of the left eye 6 . in addition the part 1 , 2 of the device upstream of the focus f moves significantly away from the left side of the observer &# 39 ; s head 7 over the distance between the focus f and the display 1 . this positioning of the device 101 in relation to the observer &# 39 ; s head 7 makes the device laterally bulky and of scarcely pleasing design . the device 102 illustrated in fig2 has more reduced lateral bulk than the device in fig1 . to reduce this bulkiness a side fold mirror 4 is arranged on the central optical path \u03b4 \u2033 in the vicinity of the focus f so that it is possible upstream of the mirror 4 to fold or direct the central optical path \u03b4 \u2033 substantially parallel to the median plane p and perpendicular to the major axis \u03b4 . therefore the device 102 in the order of the optical pathway is formed of : a light display 1 ; two groups of lenses 2 ; a side fold mirror 4 of planar , concave or convex shape ; and an ocular dioptric surface 3 of substantially elliptical shape represented by a portion of ellipse e defined by its major axis \u03b4 , its small axis \u03b4 \u2032, its centre o and its two foci f and f \u2032 located on the major axis \u03b4 either side of the centre o , as a function of the eccentricity e of the ellipse . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the side fold mirror 4 is arranged in the vicinity of the focus f so that it reflects the central optical path \u03b4 \u2033 at a chosen angle in the direction of the ocular dioptric surface 3 , so that the central optical path is then reflected in the direction of the observer &# 39 ; s eye , substantially perpendicular to the major axis \u03b4 . the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. in this configuration the bulkiness is significantly reduced since the central optical path \u03b4 \u2033 is folded back along the side of the observer &# 39 ; s head 7 . it is to be noted that the side fold mirror 4 in the example in fig2 is positioned in the immediate vicinity of the focus f on which it can be directed . the positioning thereof , so close to the focus f , is made necessary by the fact that it can offer a surface of minimum reflection . however the side fold mirror 4 can be arranged elsewhere on the central optical central path \u03b4 \u2033 downstream or upstream of the focus f according to the needs of the optical design . a description will now be given with reference to fig3 and 4 of two embodiments of a device according to the invention , in how they differ from the previously illustrated prior art . fig3 is an illustration of a first embodiment of a device according to the invention . the device 103 in fig3 , in the order of the optical pathway , is formed of : a light display 1 ; two groups of lenses 2 ; a side fold mirror 4 ; a back - inverting mirror 5 ; and an ocular dioptric surface 3 of substantially elliptical shape . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the side fold mirror 4 is arranged laterally in the vicinity of the observer &# 39 ; s eye 6 , in front of the observer &# 39 ; s temple , and it reflects the central optical path \u03b4 \u2033 at a chosen angle in the direction of the back - inverting mirror 5 . the back - inverting mirror 5 is arranged in the region of the upper part of the observer &# 39 ; s nose 10 on the right of and at the height of the left eye 6 and again reflects the central optical path \u03b4 \u2033 towards the ocular dioptric surface 3 . the mirrors are arranged such that the central optical path is then reflected by the dioptric surface in the direction of the observer &# 39 ; s eye 6 substantially perpendicular to the major axis \u03b4 . the observer &# 39 ; s eye 6 i . e . the centre of the pupil of the eye is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. according to the new optical scheme of the invention , the foci f and f \u2032 are inverted relative to the observer &# 39 ; s head . the back - inverting mirror 5 located in the vicinity of the focus f allows the virtual placing of that part of the central optical path \u03b4 \u2033 that is incident on the dioptric surface , and the focus f , in the observer &# 39 ; s head . that is to say that the display 1 is virtually placed inside the head 7 . yet in reality the central optical path \u03b4 \u2033 originates from the side of the observer &# 39 ; s head where the central optical path \u03b4 \u2033 was first folded or re - directed by the side fold mirror 4 . the side fold mirror 4 may now be located more distant from the focus f . the focus f is now directly related to the back - inverting mirror 5 . this gives the side fold mirror a much wider range of positioning , and hence of adjustment , the ocular dioptric surface 3 is now better adjusted to the morphological profile , curve , of the observer &# 39 ; s head in the vicinity of the eye , since the outer profile of the said ocular dioptric surface 3 tends to draw close to the side of the observer &# 39 ; s head 7 . the back - inverting mirror 5 is arranged in the vicinity of the upper part of the observer &# 39 ; s nose 10 on the pathway of the central optic path \u03b4 \u2033 between the side fold mirror and the ocular dioptric surface 3 . therefore the back - inverting mirror 5 can be arranged in an area hidden from the view of the observer , called a \u201c blind spot \u201d. the side fold mirror 4 is oriented angularly so that the central optic path \u03b4 \u2033 of the image is returned or reflected back towards the back - inverting mirror 5 and so that the back - inverting mirror 5 is oriented such that the central optic path \u03b4 \u2033 of the image is then returned or reflected towards the ocular dioptric surface 3 , the dioptric surface 3 being oriented so that the central optic path \u03b4 \u2033 of the image is then returned or reflected towards the observer &# 39 ; s eye 6 . in the embodiment in fig4 , the device 104 according to the invention is formed of : a light display 1 ; lenses 2 ; a side fold mirror 4 ; two decentring mirrors 8 and 9 arranged upstream of the fold mirror 4 , so that the they successively reflect the central optic path \u03b4 \u2033; a back - inverting mirror 5 of planar , convex or concave or aspherical shape ; an ocular dioptric surface 3 of substantially elliptical shape . the light display 1 diffuses an image whose pathway is illustrated by a central optic path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optic path \u03b4 \u2033. the central optic path is then reflected towards the right by the first decentring mirror 8 in the direction of the second decentring mirror 9 which is arranged in the vicinity of the observer &# 39 ; s temple . the second decentring mirror 9 then reflects the central path substantially forwardly in the direction of the side fold mirror 4 , which then reflects the same towards the right in the direction of the back - inverting mirror 5 . the back - inverting mirror 5 is arranged substantially against the upper left part of the observer &# 39 ; s nose 10 substantially at the height of the eye 6 . the central optic path \u03b4 \u2033 is then reflected thereat in the direction of the dioptric surface 3 which reflects it back towards the observer &# 39 ; s eye 6 . the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the region of the focus f \u2032. the device 104 in fig4 differs from the device 103 described with reference to fig3 in that it comprises a set of decentring mirrors 8 and 9 . the use of the set of decentring mirrors 8 , 9 has the following advantages : it allows correction of the adjustment of the angle reflection of the optical path \u03b4 \u2033 on the side mirror 4 , to offset positioning of the said side fold mirror 4 when it is brought as close as possible to the observer &# 39 ; s temple or eye , or as close as possible to the ocular dioptric surface 3 , so that it can optionally be integrated therein , for this purpose , it allows a device to be produced that is better adapted to the observer &# 39 ; s morphology since the pathway of the path \u03b4 \u2033 is close to the curve of the observer &# 39 ; s head . in addition , to make the device 103 in fig3 or the device 104 in fig4 more compact , the side fold mirror 4 can be made in a single piece with the ocular dioptric surface 3 . therefore the side fold mirror 4 can be an integral part of the outer end i . e . in the illustrated example of the left end of the ocular dioptric surface 3 . for example , each of the side fold , reflective , back - inverting and / or decentring mirrors may be planar , concave or convex or substantially aspherical in a manner making it possible to improve the overall quality of the optical system . preferably a device of the invention comprises adjustment means , in particular sets of mirror and an ocular dioptric surface capable of adapting the configuration of the device , in particular the optic path , to the observer &# 39 ; s morphology .", "category": "Physics"}
|
{"patent": "all the figures are schematic illustrations from above the head of a user or observer of each device . the left side of the observer is the left side in each figure . in all the figures , identical references designate parts or elements of parts having identical or similar functions . the figures refer to a device cooperating with the left eye of an observer . a device may be symmetrically provided to cooperate with the right eye of an observer . a device according to the invention may also comprise both a device cooperating with the left eye and a device cooperating with the right eye of the observer . fig1 illustrates an optical device 101 of the prior art , based on a method using the stigmatism of two foci of a substantially elliptical dioptric surface . the device , in the order of the optical pathway , chiefly consists of : a light display 1 ; lenses 2 ; and an ocular dioptric surface 3 of substantially elliptical shape represented by a portion of ellipse defined by its major axis \u03b4 , its small axis \u03b4 \u2032, its centre o and its two foci f and f \u2032 located on the major axis \u03b4 , either side of the centre o , as a function of the eccentricity e of the ellipse . in the figures , the axes \u03b4 , \u03b4 \u2032 of the dioptric surface 3 are shown as dot - dashed lines and the central optical path \u03b4 \u2033 is shown as a dotted line . in the example illustrated in fig1 , the device 101 is designed to cooperate with the observer &# 39 ; s left eye 6 . it is arranged on the left side of the observer &# 39 ; s head 7 . in addition , the major axis \u03b4 passes symmetrically through the two eyes of the observer and it is therefore perpendicular to a median plane p of the head 7 . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the central optical path \u03b4 \u2033 passes through one f of the foci of the dioptric surface then , after being reflected on said dioptric surface 3 , through the other focus f \u2032. the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. the spatial arrangements of the parts of device 101 in relation to the central optical path \u03b4 \u2033, on which they are aligned and centred , are inherent in the arrangement of the central optical path \u03b4 \u2033, which necessarily passes through the foci f and f \u2032. it is also ascertained that to observe this arrangement logic , the ocular dioptric surface 3 does not follow the periphery of the observer &# 39 ; s head at the height of the eye as does a conventional pair of spectacles . therefore the centre o of the dioptric surface is located largely outside , on the left of the left eye 6 . in addition the part 1 , 2 of the device upstream of the focus f moves significantly away from the left side of the observer &# 39 ; s head 7 over the distance between the focus f and the display 1 . this positioning of the device 101 in relation to the observer &# 39 ; s head 7 makes the device laterally bulky and of scarcely pleasing design . the device 102 illustrated in fig2 has more reduced lateral bulk than the device in fig1 . to reduce this bulkiness a side fold mirror 4 is arranged on the central optical path \u03b4 \u2033 in the vicinity of the focus f so that it is possible upstream of the mirror 4 to fold or direct the central optical path \u03b4 \u2033 substantially parallel to the median plane p and perpendicular to the major axis \u03b4 . therefore the device 102 in the order of the optical pathway is formed of : a light display 1 ; two groups of lenses 2 ; a side fold mirror 4 of planar , concave or convex shape ; and an ocular dioptric surface 3 of substantially elliptical shape represented by a portion of ellipse e defined by its major axis \u03b4 , its small axis \u03b4 \u2032, its centre o and its two foci f and f \u2032 located on the major axis \u03b4 either side of the centre o , as a function of the eccentricity e of the ellipse . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the side fold mirror 4 is arranged in the vicinity of the focus f so that it reflects the central optical path \u03b4 \u2033 at a chosen angle in the direction of the ocular dioptric surface 3 , so that the central optical path is then reflected in the direction of the observer &# 39 ; s eye , substantially perpendicular to the major axis \u03b4 . the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. in this configuration the bulkiness is significantly reduced since the central optical path \u03b4 \u2033 is folded back along the side of the observer &# 39 ; s head 7 . it is to be noted that the side fold mirror 4 in the example in fig2 is positioned in the immediate vicinity of the focus f on which it can be directed . the positioning thereof , so close to the focus f , is made necessary by the fact that it can offer a surface of minimum reflection . however the side fold mirror 4 can be arranged elsewhere on the central optical central path \u03b4 \u2033 downstream or upstream of the focus f according to the needs of the optical design . a description will now be given with reference to fig3 and 4 of two embodiments of a device according to the invention , in how they differ from the previously illustrated prior art . fig3 is an illustration of a first embodiment of a device according to the invention . the device 103 in fig3 , in the order of the optical pathway , is formed of : a light display 1 ; two groups of lenses 2 ; a side fold mirror 4 ; a back - inverting mirror 5 ; and an ocular dioptric surface 3 of substantially elliptical shape . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the side fold mirror 4 is arranged laterally in the vicinity of the observer &# 39 ; s eye 6 , in front of the observer &# 39 ; s temple , and it reflects the central optical path \u03b4 \u2033 at a chosen angle in the direction of the back - inverting mirror 5 . the back - inverting mirror 5 is arranged in the region of the upper part of the observer &# 39 ; s nose 10 on the right of and at the height of the left eye 6 and again reflects the central optical path \u03b4 \u2033 towards the ocular dioptric surface 3 . the mirrors are arranged such that the central optical path is then reflected by the dioptric surface in the direction of the observer &# 39 ; s eye 6 substantially perpendicular to the major axis \u03b4 . the observer &# 39 ; s eye 6 i . e . the centre of the pupil of the eye is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. according to the new optical scheme of the invention , the foci f and f \u2032 are inverted relative to the observer &# 39 ; s head . the back - inverting mirror 5 located in the vicinity of the focus f allows the virtual placing of that part of the central optical path \u03b4 \u2033 that is incident on the dioptric surface , and the focus f , in the observer &# 39 ; s head . that is to say that the display 1 is virtually placed inside the head 7 . yet in reality the central optical path \u03b4 \u2033 originates from the side of the observer &# 39 ; s head where the central optical path \u03b4 \u2033 was first folded or re - directed by the side fold mirror 4 . the side fold mirror 4 may now be located more distant from the focus f . the focus f is now directly related to the back - inverting mirror 5 . this gives the side fold mirror a much wider range of positioning , and hence of adjustment , the ocular dioptric surface 3 is now better adjusted to the morphological profile , curve , of the observer &# 39 ; s head in the vicinity of the eye , since the outer profile of the said ocular dioptric surface 3 tends to draw close to the side of the observer &# 39 ; s head 7 . the back - inverting mirror 5 is arranged in the vicinity of the upper part of the observer &# 39 ; s nose 10 on the pathway of the central optic path \u03b4 \u2033 between the side fold mirror and the ocular dioptric surface 3 . therefore the back - inverting mirror 5 can be arranged in an area hidden from the view of the observer , called a \u201c blind spot \u201d. the side fold mirror 4 is oriented angularly so that the central optic path \u03b4 \u2033 of the image is returned or reflected back towards the back - inverting mirror 5 and so that the back - inverting mirror 5 is oriented such that the central optic path \u03b4 \u2033 of the image is then returned or reflected towards the ocular dioptric surface 3 , the dioptric surface 3 being oriented so that the central optic path \u03b4 \u2033 of the image is then returned or reflected towards the observer &# 39 ; s eye 6 . in the embodiment in fig4 , the device 104 according to the invention is formed of : a light display 1 ; lenses 2 ; a side fold mirror 4 ; two decentring mirrors 8 and 9 arranged upstream of the fold mirror 4 , so that the they successively reflect the central optic path \u03b4 \u2033; a back - inverting mirror 5 of planar , convex or concave or aspherical shape ; an ocular dioptric surface 3 of substantially elliptical shape . the light display 1 diffuses an image whose pathway is illustrated by a central optic path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optic path \u03b4 \u2033. the central optic path is then reflected towards the right by the first decentring mirror 8 in the direction of the second decentring mirror 9 which is arranged in the vicinity of the observer &# 39 ; s temple . the second decentring mirror 9 then reflects the central path substantially forwardly in the direction of the side fold mirror 4 , which then reflects the same towards the right in the direction of the back - inverting mirror 5 . the back - inverting mirror 5 is arranged substantially against the upper left part of the observer &# 39 ; s nose 10 substantially at the height of the eye 6 . the central optic path \u03b4 \u2033 is then reflected thereat in the direction of the dioptric surface 3 which reflects it back towards the observer &# 39 ; s eye 6 . the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the region of the focus f \u2032. the device 104 in fig4 differs from the device 103 described with reference to fig3 in that it comprises a set of decentring mirrors 8 and 9 . the use of the set of decentring mirrors 8 , 9 has the following advantages : it allows correction of the adjustment of the angle reflection of the optical path \u03b4 \u2033 on the side mirror 4 , to offset positioning of the said side fold mirror 4 when it is brought as close as possible to the observer &# 39 ; s temple or eye , or as close as possible to the ocular dioptric surface 3 , so that it can optionally be integrated therein , for this purpose , it allows a device to be produced that is better adapted to the observer &# 39 ; s morphology since the pathway of the path \u03b4 \u2033 is close to the curve of the observer &# 39 ; s head . in addition , to make the device 103 in fig3 or the device 104 in fig4 more compact , the side fold mirror 4 can be made in a single piece with the ocular dioptric surface 3 . therefore the side fold mirror 4 can be an integral part of the outer end i . e . in the illustrated example of the left end of the ocular dioptric surface 3 . for example , each of the side fold , reflective , back - inverting and / or decentring mirrors may be planar , concave or convex or substantially aspherical in a manner making it possible to improve the overall quality of the optical system . preferably a device of the invention comprises adjustment means , in particular sets of mirror and an ocular dioptric surface capable of adapting the configuration of the device , in particular the optic path , to the observer &# 39 ; s morphology .", "category": "Textiles; Paper"}
|
Is the patent correctly categorized?
| 0.25 |
ab7dbe7acb7f81dc7aa04494c82020cf62dbf0961bcc8e2034548c678553ff66
| 0.070801 | 0.006897 | 0.195313 | 0.008301 | 0.178711 | 0.001869 |
null |
{"patent": "all the figures are schematic illustrations from above the head of a user or observer of each device . the left side of the observer is the left side in each figure . in all the figures , identical references designate parts or elements of parts having identical or similar functions . the figures refer to a device cooperating with the left eye of an observer . a device may be symmetrically provided to cooperate with the right eye of an observer . a device according to the invention may also comprise both a device cooperating with the left eye and a device cooperating with the right eye of the observer . fig1 illustrates an optical device 101 of the prior art , based on a method using the stigmatism of two foci of a substantially elliptical dioptric surface . the device , in the order of the optical pathway , chiefly consists of : a light display 1 ; lenses 2 ; and an ocular dioptric surface 3 of substantially elliptical shape represented by a portion of ellipse defined by its major axis \u03b4 , its small axis \u03b4 \u2032, its centre o and its two foci f and f \u2032 located on the major axis \u03b4 , either side of the centre o , as a function of the eccentricity e of the ellipse . in the figures , the axes \u03b4 , \u03b4 \u2032 of the dioptric surface 3 are shown as dot - dashed lines and the central optical path \u03b4 \u2033 is shown as a dotted line . in the example illustrated in fig1 , the device 101 is designed to cooperate with the observer &# 39 ; s left eye 6 . it is arranged on the left side of the observer &# 39 ; s head 7 . in addition , the major axis \u03b4 passes symmetrically through the two eyes of the observer and it is therefore perpendicular to a median plane p of the head 7 . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the central optical path \u03b4 \u2033 passes through one f of the foci of the dioptric surface then , after being reflected on said dioptric surface 3 , through the other focus f \u2032. the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. the spatial arrangements of the parts of device 101 in relation to the central optical path \u03b4 \u2033, on which they are aligned and centred , are inherent in the arrangement of the central optical path \u03b4 \u2033, which necessarily passes through the foci f and f \u2032. it is also ascertained that to observe this arrangement logic , the ocular dioptric surface 3 does not follow the periphery of the observer &# 39 ; s head at the height of the eye as does a conventional pair of spectacles . therefore the centre o of the dioptric surface is located largely outside , on the left of the left eye 6 . in addition the part 1 , 2 of the device upstream of the focus f moves significantly away from the left side of the observer &# 39 ; s head 7 over the distance between the focus f and the display 1 . this positioning of the device 101 in relation to the observer &# 39 ; s head 7 makes the device laterally bulky and of scarcely pleasing design . the device 102 illustrated in fig2 has more reduced lateral bulk than the device in fig1 . to reduce this bulkiness a side fold mirror 4 is arranged on the central optical path \u03b4 \u2033 in the vicinity of the focus f so that it is possible upstream of the mirror 4 to fold or direct the central optical path \u03b4 \u2033 substantially parallel to the median plane p and perpendicular to the major axis \u03b4 . therefore the device 102 in the order of the optical pathway is formed of : a light display 1 ; two groups of lenses 2 ; a side fold mirror 4 of planar , concave or convex shape ; and an ocular dioptric surface 3 of substantially elliptical shape represented by a portion of ellipse e defined by its major axis \u03b4 , its small axis \u03b4 \u2032, its centre o and its two foci f and f \u2032 located on the major axis \u03b4 either side of the centre o , as a function of the eccentricity e of the ellipse . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the side fold mirror 4 is arranged in the vicinity of the focus f so that it reflects the central optical path \u03b4 \u2033 at a chosen angle in the direction of the ocular dioptric surface 3 , so that the central optical path is then reflected in the direction of the observer &# 39 ; s eye , substantially perpendicular to the major axis \u03b4 . the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. in this configuration the bulkiness is significantly reduced since the central optical path \u03b4 \u2033 is folded back along the side of the observer &# 39 ; s head 7 . it is to be noted that the side fold mirror 4 in the example in fig2 is positioned in the immediate vicinity of the focus f on which it can be directed . the positioning thereof , so close to the focus f , is made necessary by the fact that it can offer a surface of minimum reflection . however the side fold mirror 4 can be arranged elsewhere on the central optical central path \u03b4 \u2033 downstream or upstream of the focus f according to the needs of the optical design . a description will now be given with reference to fig3 and 4 of two embodiments of a device according to the invention , in how they differ from the previously illustrated prior art . fig3 is an illustration of a first embodiment of a device according to the invention . the device 103 in fig3 , in the order of the optical pathway , is formed of : a light display 1 ; two groups of lenses 2 ; a side fold mirror 4 ; a back - inverting mirror 5 ; and an ocular dioptric surface 3 of substantially elliptical shape . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the side fold mirror 4 is arranged laterally in the vicinity of the observer &# 39 ; s eye 6 , in front of the observer &# 39 ; s temple , and it reflects the central optical path \u03b4 \u2033 at a chosen angle in the direction of the back - inverting mirror 5 . the back - inverting mirror 5 is arranged in the region of the upper part of the observer &# 39 ; s nose 10 on the right of and at the height of the left eye 6 and again reflects the central optical path \u03b4 \u2033 towards the ocular dioptric surface 3 . the mirrors are arranged such that the central optical path is then reflected by the dioptric surface in the direction of the observer &# 39 ; s eye 6 substantially perpendicular to the major axis \u03b4 . the observer &# 39 ; s eye 6 i . e . the centre of the pupil of the eye is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. according to the new optical scheme of the invention , the foci f and f \u2032 are inverted relative to the observer &# 39 ; s head . the back - inverting mirror 5 located in the vicinity of the focus f allows the virtual placing of that part of the central optical path \u03b4 \u2033 that is incident on the dioptric surface , and the focus f , in the observer &# 39 ; s head . that is to say that the display 1 is virtually placed inside the head 7 . yet in reality the central optical path \u03b4 \u2033 originates from the side of the observer &# 39 ; s head where the central optical path \u03b4 \u2033 was first folded or re - directed by the side fold mirror 4 . the side fold mirror 4 may now be located more distant from the focus f . the focus f is now directly related to the back - inverting mirror 5 . this gives the side fold mirror a much wider range of positioning , and hence of adjustment , the ocular dioptric surface 3 is now better adjusted to the morphological profile , curve , of the observer &# 39 ; s head in the vicinity of the eye , since the outer profile of the said ocular dioptric surface 3 tends to draw close to the side of the observer &# 39 ; s head 7 . the back - inverting mirror 5 is arranged in the vicinity of the upper part of the observer &# 39 ; s nose 10 on the pathway of the central optic path \u03b4 \u2033 between the side fold mirror and the ocular dioptric surface 3 . therefore the back - inverting mirror 5 can be arranged in an area hidden from the view of the observer , called a \u201c blind spot \u201d. the side fold mirror 4 is oriented angularly so that the central optic path \u03b4 \u2033 of the image is returned or reflected back towards the back - inverting mirror 5 and so that the back - inverting mirror 5 is oriented such that the central optic path \u03b4 \u2033 of the image is then returned or reflected towards the ocular dioptric surface 3 , the dioptric surface 3 being oriented so that the central optic path \u03b4 \u2033 of the image is then returned or reflected towards the observer &# 39 ; s eye 6 . in the embodiment in fig4 , the device 104 according to the invention is formed of : a light display 1 ; lenses 2 ; a side fold mirror 4 ; two decentring mirrors 8 and 9 arranged upstream of the fold mirror 4 , so that the they successively reflect the central optic path \u03b4 \u2033; a back - inverting mirror 5 of planar , convex or concave or aspherical shape ; an ocular dioptric surface 3 of substantially elliptical shape . the light display 1 diffuses an image whose pathway is illustrated by a central optic path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optic path \u03b4 \u2033. the central optic path is then reflected towards the right by the first decentring mirror 8 in the direction of the second decentring mirror 9 which is arranged in the vicinity of the observer &# 39 ; s temple . the second decentring mirror 9 then reflects the central path substantially forwardly in the direction of the side fold mirror 4 , which then reflects the same towards the right in the direction of the back - inverting mirror 5 . the back - inverting mirror 5 is arranged substantially against the upper left part of the observer &# 39 ; s nose 10 substantially at the height of the eye 6 . the central optic path \u03b4 \u2033 is then reflected thereat in the direction of the dioptric surface 3 which reflects it back towards the observer &# 39 ; s eye 6 . the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the region of the focus f \u2032. the device 104 in fig4 differs from the device 103 described with reference to fig3 in that it comprises a set of decentring mirrors 8 and 9 . the use of the set of decentring mirrors 8 , 9 has the following advantages : it allows correction of the adjustment of the angle reflection of the optical path \u03b4 \u2033 on the side mirror 4 , to offset positioning of the said side fold mirror 4 when it is brought as close as possible to the observer &# 39 ; s temple or eye , or as close as possible to the ocular dioptric surface 3 , so that it can optionally be integrated therein , for this purpose , it allows a device to be produced that is better adapted to the observer &# 39 ; s morphology since the pathway of the path \u03b4 \u2033 is close to the curve of the observer &# 39 ; s head . in addition , to make the device 103 in fig3 or the device 104 in fig4 more compact , the side fold mirror 4 can be made in a single piece with the ocular dioptric surface 3 . therefore the side fold mirror 4 can be an integral part of the outer end i . e . in the illustrated example of the left end of the ocular dioptric surface 3 . for example , each of the side fold , reflective , back - inverting and / or decentring mirrors may be planar , concave or convex or substantially aspherical in a manner making it possible to improve the overall quality of the optical system . preferably a device of the invention comprises adjustment means , in particular sets of mirror and an ocular dioptric surface capable of adapting the configuration of the device , in particular the optic path , to the observer &# 39 ; s morphology .", "category": "Physics"}
|
{"category": "Fixed Constructions", "patent": "all the figures are schematic illustrations from above the head of a user or observer of each device . the left side of the observer is the left side in each figure . in all the figures , identical references designate parts or elements of parts having identical or similar functions . the figures refer to a device cooperating with the left eye of an observer . a device may be symmetrically provided to cooperate with the right eye of an observer . a device according to the invention may also comprise both a device cooperating with the left eye and a device cooperating with the right eye of the observer . fig1 illustrates an optical device 101 of the prior art , based on a method using the stigmatism of two foci of a substantially elliptical dioptric surface . the device , in the order of the optical pathway , chiefly consists of : a light display 1 ; lenses 2 ; and an ocular dioptric surface 3 of substantially elliptical shape represented by a portion of ellipse defined by its major axis \u03b4 , its small axis \u03b4 \u2032, its centre o and its two foci f and f \u2032 located on the major axis \u03b4 , either side of the centre o , as a function of the eccentricity e of the ellipse . in the figures , the axes \u03b4 , \u03b4 \u2032 of the dioptric surface 3 are shown as dot - dashed lines and the central optical path \u03b4 \u2033 is shown as a dotted line . in the example illustrated in fig1 , the device 101 is designed to cooperate with the observer &# 39 ; s left eye 6 . it is arranged on the left side of the observer &# 39 ; s head 7 . in addition , the major axis \u03b4 passes symmetrically through the two eyes of the observer and it is therefore perpendicular to a median plane p of the head 7 . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the central optical path \u03b4 \u2033 passes through one f of the foci of the dioptric surface then , after being reflected on said dioptric surface 3 , through the other focus f \u2032. the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. the spatial arrangements of the parts of device 101 in relation to the central optical path \u03b4 \u2033, on which they are aligned and centred , are inherent in the arrangement of the central optical path \u03b4 \u2033, which necessarily passes through the foci f and f \u2032. it is also ascertained that to observe this arrangement logic , the ocular dioptric surface 3 does not follow the periphery of the observer &# 39 ; s head at the height of the eye as does a conventional pair of spectacles . therefore the centre o of the dioptric surface is located largely outside , on the left of the left eye 6 . in addition the part 1 , 2 of the device upstream of the focus f moves significantly away from the left side of the observer &# 39 ; s head 7 over the distance between the focus f and the display 1 . this positioning of the device 101 in relation to the observer &# 39 ; s head 7 makes the device laterally bulky and of scarcely pleasing design . the device 102 illustrated in fig2 has more reduced lateral bulk than the device in fig1 . to reduce this bulkiness a side fold mirror 4 is arranged on the central optical path \u03b4 \u2033 in the vicinity of the focus f so that it is possible upstream of the mirror 4 to fold or direct the central optical path \u03b4 \u2033 substantially parallel to the median plane p and perpendicular to the major axis \u03b4 . therefore the device 102 in the order of the optical pathway is formed of : a light display 1 ; two groups of lenses 2 ; a side fold mirror 4 of planar , concave or convex shape ; and an ocular dioptric surface 3 of substantially elliptical shape represented by a portion of ellipse e defined by its major axis \u03b4 , its small axis \u03b4 \u2032, its centre o and its two foci f and f \u2032 located on the major axis \u03b4 either side of the centre o , as a function of the eccentricity e of the ellipse . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the side fold mirror 4 is arranged in the vicinity of the focus f so that it reflects the central optical path \u03b4 \u2033 at a chosen angle in the direction of the ocular dioptric surface 3 , so that the central optical path is then reflected in the direction of the observer &# 39 ; s eye , substantially perpendicular to the major axis \u03b4 . the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. in this configuration the bulkiness is significantly reduced since the central optical path \u03b4 \u2033 is folded back along the side of the observer &# 39 ; s head 7 . it is to be noted that the side fold mirror 4 in the example in fig2 is positioned in the immediate vicinity of the focus f on which it can be directed . the positioning thereof , so close to the focus f , is made necessary by the fact that it can offer a surface of minimum reflection . however the side fold mirror 4 can be arranged elsewhere on the central optical central path \u03b4 \u2033 downstream or upstream of the focus f according to the needs of the optical design . a description will now be given with reference to fig3 and 4 of two embodiments of a device according to the invention , in how they differ from the previously illustrated prior art . fig3 is an illustration of a first embodiment of a device according to the invention . the device 103 in fig3 , in the order of the optical pathway , is formed of : a light display 1 ; two groups of lenses 2 ; a side fold mirror 4 ; a back - inverting mirror 5 ; and an ocular dioptric surface 3 of substantially elliptical shape . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the side fold mirror 4 is arranged laterally in the vicinity of the observer &# 39 ; s eye 6 , in front of the observer &# 39 ; s temple , and it reflects the central optical path \u03b4 \u2033 at a chosen angle in the direction of the back - inverting mirror 5 . the back - inverting mirror 5 is arranged in the region of the upper part of the observer &# 39 ; s nose 10 on the right of and at the height of the left eye 6 and again reflects the central optical path \u03b4 \u2033 towards the ocular dioptric surface 3 . the mirrors are arranged such that the central optical path is then reflected by the dioptric surface in the direction of the observer &# 39 ; s eye 6 substantially perpendicular to the major axis \u03b4 . the observer &# 39 ; s eye 6 i . e . the centre of the pupil of the eye is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. according to the new optical scheme of the invention , the foci f and f \u2032 are inverted relative to the observer &# 39 ; s head . the back - inverting mirror 5 located in the vicinity of the focus f allows the virtual placing of that part of the central optical path \u03b4 \u2033 that is incident on the dioptric surface , and the focus f , in the observer &# 39 ; s head . that is to say that the display 1 is virtually placed inside the head 7 . yet in reality the central optical path \u03b4 \u2033 originates from the side of the observer &# 39 ; s head where the central optical path \u03b4 \u2033 was first folded or re - directed by the side fold mirror 4 . the side fold mirror 4 may now be located more distant from the focus f . the focus f is now directly related to the back - inverting mirror 5 . this gives the side fold mirror a much wider range of positioning , and hence of adjustment , the ocular dioptric surface 3 is now better adjusted to the morphological profile , curve , of the observer &# 39 ; s head in the vicinity of the eye , since the outer profile of the said ocular dioptric surface 3 tends to draw close to the side of the observer &# 39 ; s head 7 . the back - inverting mirror 5 is arranged in the vicinity of the upper part of the observer &# 39 ; s nose 10 on the pathway of the central optic path \u03b4 \u2033 between the side fold mirror and the ocular dioptric surface 3 . therefore the back - inverting mirror 5 can be arranged in an area hidden from the view of the observer , called a \u201c blind spot \u201d. the side fold mirror 4 is oriented angularly so that the central optic path \u03b4 \u2033 of the image is returned or reflected back towards the back - inverting mirror 5 and so that the back - inverting mirror 5 is oriented such that the central optic path \u03b4 \u2033 of the image is then returned or reflected towards the ocular dioptric surface 3 , the dioptric surface 3 being oriented so that the central optic path \u03b4 \u2033 of the image is then returned or reflected towards the observer &# 39 ; s eye 6 . in the embodiment in fig4 , the device 104 according to the invention is formed of : a light display 1 ; lenses 2 ; a side fold mirror 4 ; two decentring mirrors 8 and 9 arranged upstream of the fold mirror 4 , so that the they successively reflect the central optic path \u03b4 \u2033; a back - inverting mirror 5 of planar , convex or concave or aspherical shape ; an ocular dioptric surface 3 of substantially elliptical shape . the light display 1 diffuses an image whose pathway is illustrated by a central optic path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optic path \u03b4 \u2033. the central optic path is then reflected towards the right by the first decentring mirror 8 in the direction of the second decentring mirror 9 which is arranged in the vicinity of the observer &# 39 ; s temple . the second decentring mirror 9 then reflects the central path substantially forwardly in the direction of the side fold mirror 4 , which then reflects the same towards the right in the direction of the back - inverting mirror 5 . the back - inverting mirror 5 is arranged substantially against the upper left part of the observer &# 39 ; s nose 10 substantially at the height of the eye 6 . the central optic path \u03b4 \u2033 is then reflected thereat in the direction of the dioptric surface 3 which reflects it back towards the observer &# 39 ; s eye 6 . the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the region of the focus f \u2032. the device 104 in fig4 differs from the device 103 described with reference to fig3 in that it comprises a set of decentring mirrors 8 and 9 . the use of the set of decentring mirrors 8 , 9 has the following advantages : it allows correction of the adjustment of the angle reflection of the optical path \u03b4 \u2033 on the side mirror 4 , to offset positioning of the said side fold mirror 4 when it is brought as close as possible to the observer &# 39 ; s temple or eye , or as close as possible to the ocular dioptric surface 3 , so that it can optionally be integrated therein , for this purpose , it allows a device to be produced that is better adapted to the observer &# 39 ; s morphology since the pathway of the path \u03b4 \u2033 is close to the curve of the observer &# 39 ; s head . in addition , to make the device 103 in fig3 or the device 104 in fig4 more compact , the side fold mirror 4 can be made in a single piece with the ocular dioptric surface 3 . therefore the side fold mirror 4 can be an integral part of the outer end i . e . in the illustrated example of the left end of the ocular dioptric surface 3 . for example , each of the side fold , reflective , back - inverting and / or decentring mirrors may be planar , concave or convex or substantially aspherical in a manner making it possible to improve the overall quality of the optical system . preferably a device of the invention comprises adjustment means , in particular sets of mirror and an ocular dioptric surface capable of adapting the configuration of the device , in particular the optic path , to the observer &# 39 ; s morphology ."}
|
Is the categorization of this patent accurate?
| 0.25 |
ab7dbe7acb7f81dc7aa04494c82020cf62dbf0961bcc8e2034548c678553ff66
| 0.079102 | 0.306641 | 0.124023 | 0.699219 | 0.228516 | 0.396484 |
null |
{"category": "Physics", "patent": "all the figures are schematic illustrations from above the head of a user or observer of each device . the left side of the observer is the left side in each figure . in all the figures , identical references designate parts or elements of parts having identical or similar functions . the figures refer to a device cooperating with the left eye of an observer . a device may be symmetrically provided to cooperate with the right eye of an observer . a device according to the invention may also comprise both a device cooperating with the left eye and a device cooperating with the right eye of the observer . fig1 illustrates an optical device 101 of the prior art , based on a method using the stigmatism of two foci of a substantially elliptical dioptric surface . the device , in the order of the optical pathway , chiefly consists of : a light display 1 ; lenses 2 ; and an ocular dioptric surface 3 of substantially elliptical shape represented by a portion of ellipse defined by its major axis \u03b4 , its small axis \u03b4 \u2032, its centre o and its two foci f and f \u2032 located on the major axis \u03b4 , either side of the centre o , as a function of the eccentricity e of the ellipse . in the figures , the axes \u03b4 , \u03b4 \u2032 of the dioptric surface 3 are shown as dot - dashed lines and the central optical path \u03b4 \u2033 is shown as a dotted line . in the example illustrated in fig1 , the device 101 is designed to cooperate with the observer &# 39 ; s left eye 6 . it is arranged on the left side of the observer &# 39 ; s head 7 . in addition , the major axis \u03b4 passes symmetrically through the two eyes of the observer and it is therefore perpendicular to a median plane p of the head 7 . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the central optical path \u03b4 \u2033 passes through one f of the foci of the dioptric surface then , after being reflected on said dioptric surface 3 , through the other focus f \u2032. the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. the spatial arrangements of the parts of device 101 in relation to the central optical path \u03b4 \u2033, on which they are aligned and centred , are inherent in the arrangement of the central optical path \u03b4 \u2033, which necessarily passes through the foci f and f \u2032. it is also ascertained that to observe this arrangement logic , the ocular dioptric surface 3 does not follow the periphery of the observer &# 39 ; s head at the height of the eye as does a conventional pair of spectacles . therefore the centre o of the dioptric surface is located largely outside , on the left of the left eye 6 . in addition the part 1 , 2 of the device upstream of the focus f moves significantly away from the left side of the observer &# 39 ; s head 7 over the distance between the focus f and the display 1 . this positioning of the device 101 in relation to the observer &# 39 ; s head 7 makes the device laterally bulky and of scarcely pleasing design . the device 102 illustrated in fig2 has more reduced lateral bulk than the device in fig1 . to reduce this bulkiness a side fold mirror 4 is arranged on the central optical path \u03b4 \u2033 in the vicinity of the focus f so that it is possible upstream of the mirror 4 to fold or direct the central optical path \u03b4 \u2033 substantially parallel to the median plane p and perpendicular to the major axis \u03b4 . therefore the device 102 in the order of the optical pathway is formed of : a light display 1 ; two groups of lenses 2 ; a side fold mirror 4 of planar , concave or convex shape ; and an ocular dioptric surface 3 of substantially elliptical shape represented by a portion of ellipse e defined by its major axis \u03b4 , its small axis \u03b4 \u2032, its centre o and its two foci f and f \u2032 located on the major axis \u03b4 either side of the centre o , as a function of the eccentricity e of the ellipse . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the side fold mirror 4 is arranged in the vicinity of the focus f so that it reflects the central optical path \u03b4 \u2033 at a chosen angle in the direction of the ocular dioptric surface 3 , so that the central optical path is then reflected in the direction of the observer &# 39 ; s eye , substantially perpendicular to the major axis \u03b4 . the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. in this configuration the bulkiness is significantly reduced since the central optical path \u03b4 \u2033 is folded back along the side of the observer &# 39 ; s head 7 . it is to be noted that the side fold mirror 4 in the example in fig2 is positioned in the immediate vicinity of the focus f on which it can be directed . the positioning thereof , so close to the focus f , is made necessary by the fact that it can offer a surface of minimum reflection . however the side fold mirror 4 can be arranged elsewhere on the central optical central path \u03b4 \u2033 downstream or upstream of the focus f according to the needs of the optical design . a description will now be given with reference to fig3 and 4 of two embodiments of a device according to the invention , in how they differ from the previously illustrated prior art . fig3 is an illustration of a first embodiment of a device according to the invention . the device 103 in fig3 , in the order of the optical pathway , is formed of : a light display 1 ; two groups of lenses 2 ; a side fold mirror 4 ; a back - inverting mirror 5 ; and an ocular dioptric surface 3 of substantially elliptical shape . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the side fold mirror 4 is arranged laterally in the vicinity of the observer &# 39 ; s eye 6 , in front of the observer &# 39 ; s temple , and it reflects the central optical path \u03b4 \u2033 at a chosen angle in the direction of the back - inverting mirror 5 . the back - inverting mirror 5 is arranged in the region of the upper part of the observer &# 39 ; s nose 10 on the right of and at the height of the left eye 6 and again reflects the central optical path \u03b4 \u2033 towards the ocular dioptric surface 3 . the mirrors are arranged such that the central optical path is then reflected by the dioptric surface in the direction of the observer &# 39 ; s eye 6 substantially perpendicular to the major axis \u03b4 . the observer &# 39 ; s eye 6 i . e . the centre of the pupil of the eye is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. according to the new optical scheme of the invention , the foci f and f \u2032 are inverted relative to the observer &# 39 ; s head . the back - inverting mirror 5 located in the vicinity of the focus f allows the virtual placing of that part of the central optical path \u03b4 \u2033 that is incident on the dioptric surface , and the focus f , in the observer &# 39 ; s head . that is to say that the display 1 is virtually placed inside the head 7 . yet in reality the central optical path \u03b4 \u2033 originates from the side of the observer &# 39 ; s head where the central optical path \u03b4 \u2033 was first folded or re - directed by the side fold mirror 4 . the side fold mirror 4 may now be located more distant from the focus f . the focus f is now directly related to the back - inverting mirror 5 . this gives the side fold mirror a much wider range of positioning , and hence of adjustment , the ocular dioptric surface 3 is now better adjusted to the morphological profile , curve , of the observer &# 39 ; s head in the vicinity of the eye , since the outer profile of the said ocular dioptric surface 3 tends to draw close to the side of the observer &# 39 ; s head 7 . the back - inverting mirror 5 is arranged in the vicinity of the upper part of the observer &# 39 ; s nose 10 on the pathway of the central optic path \u03b4 \u2033 between the side fold mirror and the ocular dioptric surface 3 . therefore the back - inverting mirror 5 can be arranged in an area hidden from the view of the observer , called a \u201c blind spot \u201d. the side fold mirror 4 is oriented angularly so that the central optic path \u03b4 \u2033 of the image is returned or reflected back towards the back - inverting mirror 5 and so that the back - inverting mirror 5 is oriented such that the central optic path \u03b4 \u2033 of the image is then returned or reflected towards the ocular dioptric surface 3 , the dioptric surface 3 being oriented so that the central optic path \u03b4 \u2033 of the image is then returned or reflected towards the observer &# 39 ; s eye 6 . in the embodiment in fig4 , the device 104 according to the invention is formed of : a light display 1 ; lenses 2 ; a side fold mirror 4 ; two decentring mirrors 8 and 9 arranged upstream of the fold mirror 4 , so that the they successively reflect the central optic path \u03b4 \u2033; a back - inverting mirror 5 of planar , convex or concave or aspherical shape ; an ocular dioptric surface 3 of substantially elliptical shape . the light display 1 diffuses an image whose pathway is illustrated by a central optic path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optic path \u03b4 \u2033. the central optic path is then reflected towards the right by the first decentring mirror 8 in the direction of the second decentring mirror 9 which is arranged in the vicinity of the observer &# 39 ; s temple . the second decentring mirror 9 then reflects the central path substantially forwardly in the direction of the side fold mirror 4 , which then reflects the same towards the right in the direction of the back - inverting mirror 5 . the back - inverting mirror 5 is arranged substantially against the upper left part of the observer &# 39 ; s nose 10 substantially at the height of the eye 6 . the central optic path \u03b4 \u2033 is then reflected thereat in the direction of the dioptric surface 3 which reflects it back towards the observer &# 39 ; s eye 6 . the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the region of the focus f \u2032. the device 104 in fig4 differs from the device 103 described with reference to fig3 in that it comprises a set of decentring mirrors 8 and 9 . the use of the set of decentring mirrors 8 , 9 has the following advantages : it allows correction of the adjustment of the angle reflection of the optical path \u03b4 \u2033 on the side mirror 4 , to offset positioning of the said side fold mirror 4 when it is brought as close as possible to the observer &# 39 ; s temple or eye , or as close as possible to the ocular dioptric surface 3 , so that it can optionally be integrated therein , for this purpose , it allows a device to be produced that is better adapted to the observer &# 39 ; s morphology since the pathway of the path \u03b4 \u2033 is close to the curve of the observer &# 39 ; s head . in addition , to make the device 103 in fig3 or the device 104 in fig4 more compact , the side fold mirror 4 can be made in a single piece with the ocular dioptric surface 3 . therefore the side fold mirror 4 can be an integral part of the outer end i . e . in the illustrated example of the left end of the ocular dioptric surface 3 . for example , each of the side fold , reflective , back - inverting and / or decentring mirrors may be planar , concave or convex or substantially aspherical in a manner making it possible to improve the overall quality of the optical system . preferably a device of the invention comprises adjustment means , in particular sets of mirror and an ocular dioptric surface capable of adapting the configuration of the device , in particular the optic path , to the observer &# 39 ; s morphology ."}
|
{"category": "Mechanical Engineering; Lightning; Heating; Weapons; Blasting", "patent": "all the figures are schematic illustrations from above the head of a user or observer of each device . the left side of the observer is the left side in each figure . in all the figures , identical references designate parts or elements of parts having identical or similar functions . the figures refer to a device cooperating with the left eye of an observer . a device may be symmetrically provided to cooperate with the right eye of an observer . a device according to the invention may also comprise both a device cooperating with the left eye and a device cooperating with the right eye of the observer . fig1 illustrates an optical device 101 of the prior art , based on a method using the stigmatism of two foci of a substantially elliptical dioptric surface . the device , in the order of the optical pathway , chiefly consists of : a light display 1 ; lenses 2 ; and an ocular dioptric surface 3 of substantially elliptical shape represented by a portion of ellipse defined by its major axis \u03b4 , its small axis \u03b4 \u2032, its centre o and its two foci f and f \u2032 located on the major axis \u03b4 , either side of the centre o , as a function of the eccentricity e of the ellipse . in the figures , the axes \u03b4 , \u03b4 \u2032 of the dioptric surface 3 are shown as dot - dashed lines and the central optical path \u03b4 \u2033 is shown as a dotted line . in the example illustrated in fig1 , the device 101 is designed to cooperate with the observer &# 39 ; s left eye 6 . it is arranged on the left side of the observer &# 39 ; s head 7 . in addition , the major axis \u03b4 passes symmetrically through the two eyes of the observer and it is therefore perpendicular to a median plane p of the head 7 . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the central optical path \u03b4 \u2033 passes through one f of the foci of the dioptric surface then , after being reflected on said dioptric surface 3 , through the other focus f \u2032. the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. the spatial arrangements of the parts of device 101 in relation to the central optical path \u03b4 \u2033, on which they are aligned and centred , are inherent in the arrangement of the central optical path \u03b4 \u2033, which necessarily passes through the foci f and f \u2032. it is also ascertained that to observe this arrangement logic , the ocular dioptric surface 3 does not follow the periphery of the observer &# 39 ; s head at the height of the eye as does a conventional pair of spectacles . therefore the centre o of the dioptric surface is located largely outside , on the left of the left eye 6 . in addition the part 1 , 2 of the device upstream of the focus f moves significantly away from the left side of the observer &# 39 ; s head 7 over the distance between the focus f and the display 1 . this positioning of the device 101 in relation to the observer &# 39 ; s head 7 makes the device laterally bulky and of scarcely pleasing design . the device 102 illustrated in fig2 has more reduced lateral bulk than the device in fig1 . to reduce this bulkiness a side fold mirror 4 is arranged on the central optical path \u03b4 \u2033 in the vicinity of the focus f so that it is possible upstream of the mirror 4 to fold or direct the central optical path \u03b4 \u2033 substantially parallel to the median plane p and perpendicular to the major axis \u03b4 . therefore the device 102 in the order of the optical pathway is formed of : a light display 1 ; two groups of lenses 2 ; a side fold mirror 4 of planar , concave or convex shape ; and an ocular dioptric surface 3 of substantially elliptical shape represented by a portion of ellipse e defined by its major axis \u03b4 , its small axis \u03b4 \u2032, its centre o and its two foci f and f \u2032 located on the major axis \u03b4 either side of the centre o , as a function of the eccentricity e of the ellipse . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the side fold mirror 4 is arranged in the vicinity of the focus f so that it reflects the central optical path \u03b4 \u2033 at a chosen angle in the direction of the ocular dioptric surface 3 , so that the central optical path is then reflected in the direction of the observer &# 39 ; s eye , substantially perpendicular to the major axis \u03b4 . the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. in this configuration the bulkiness is significantly reduced since the central optical path \u03b4 \u2033 is folded back along the side of the observer &# 39 ; s head 7 . it is to be noted that the side fold mirror 4 in the example in fig2 is positioned in the immediate vicinity of the focus f on which it can be directed . the positioning thereof , so close to the focus f , is made necessary by the fact that it can offer a surface of minimum reflection . however the side fold mirror 4 can be arranged elsewhere on the central optical central path \u03b4 \u2033 downstream or upstream of the focus f according to the needs of the optical design . a description will now be given with reference to fig3 and 4 of two embodiments of a device according to the invention , in how they differ from the previously illustrated prior art . fig3 is an illustration of a first embodiment of a device according to the invention . the device 103 in fig3 , in the order of the optical pathway , is formed of : a light display 1 ; two groups of lenses 2 ; a side fold mirror 4 ; a back - inverting mirror 5 ; and an ocular dioptric surface 3 of substantially elliptical shape . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the side fold mirror 4 is arranged laterally in the vicinity of the observer &# 39 ; s eye 6 , in front of the observer &# 39 ; s temple , and it reflects the central optical path \u03b4 \u2033 at a chosen angle in the direction of the back - inverting mirror 5 . the back - inverting mirror 5 is arranged in the region of the upper part of the observer &# 39 ; s nose 10 on the right of and at the height of the left eye 6 and again reflects the central optical path \u03b4 \u2033 towards the ocular dioptric surface 3 . the mirrors are arranged such that the central optical path is then reflected by the dioptric surface in the direction of the observer &# 39 ; s eye 6 substantially perpendicular to the major axis \u03b4 . the observer &# 39 ; s eye 6 i . e . the centre of the pupil of the eye is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. according to the new optical scheme of the invention , the foci f and f \u2032 are inverted relative to the observer &# 39 ; s head . the back - inverting mirror 5 located in the vicinity of the focus f allows the virtual placing of that part of the central optical path \u03b4 \u2033 that is incident on the dioptric surface , and the focus f , in the observer &# 39 ; s head . that is to say that the display 1 is virtually placed inside the head 7 . yet in reality the central optical path \u03b4 \u2033 originates from the side of the observer &# 39 ; s head where the central optical path \u03b4 \u2033 was first folded or re - directed by the side fold mirror 4 . the side fold mirror 4 may now be located more distant from the focus f . the focus f is now directly related to the back - inverting mirror 5 . this gives the side fold mirror a much wider range of positioning , and hence of adjustment , the ocular dioptric surface 3 is now better adjusted to the morphological profile , curve , of the observer &# 39 ; s head in the vicinity of the eye , since the outer profile of the said ocular dioptric surface 3 tends to draw close to the side of the observer &# 39 ; s head 7 . the back - inverting mirror 5 is arranged in the vicinity of the upper part of the observer &# 39 ; s nose 10 on the pathway of the central optic path \u03b4 \u2033 between the side fold mirror and the ocular dioptric surface 3 . therefore the back - inverting mirror 5 can be arranged in an area hidden from the view of the observer , called a \u201c blind spot \u201d. the side fold mirror 4 is oriented angularly so that the central optic path \u03b4 \u2033 of the image is returned or reflected back towards the back - inverting mirror 5 and so that the back - inverting mirror 5 is oriented such that the central optic path \u03b4 \u2033 of the image is then returned or reflected towards the ocular dioptric surface 3 , the dioptric surface 3 being oriented so that the central optic path \u03b4 \u2033 of the image is then returned or reflected towards the observer &# 39 ; s eye 6 . in the embodiment in fig4 , the device 104 according to the invention is formed of : a light display 1 ; lenses 2 ; a side fold mirror 4 ; two decentring mirrors 8 and 9 arranged upstream of the fold mirror 4 , so that the they successively reflect the central optic path \u03b4 \u2033; a back - inverting mirror 5 of planar , convex or concave or aspherical shape ; an ocular dioptric surface 3 of substantially elliptical shape . the light display 1 diffuses an image whose pathway is illustrated by a central optic path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optic path \u03b4 \u2033. the central optic path is then reflected towards the right by the first decentring mirror 8 in the direction of the second decentring mirror 9 which is arranged in the vicinity of the observer &# 39 ; s temple . the second decentring mirror 9 then reflects the central path substantially forwardly in the direction of the side fold mirror 4 , which then reflects the same towards the right in the direction of the back - inverting mirror 5 . the back - inverting mirror 5 is arranged substantially against the upper left part of the observer &# 39 ; s nose 10 substantially at the height of the eye 6 . the central optic path \u03b4 \u2033 is then reflected thereat in the direction of the dioptric surface 3 which reflects it back towards the observer &# 39 ; s eye 6 . the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the region of the focus f \u2032. the device 104 in fig4 differs from the device 103 described with reference to fig3 in that it comprises a set of decentring mirrors 8 and 9 . the use of the set of decentring mirrors 8 , 9 has the following advantages : it allows correction of the adjustment of the angle reflection of the optical path \u03b4 \u2033 on the side mirror 4 , to offset positioning of the said side fold mirror 4 when it is brought as close as possible to the observer &# 39 ; s temple or eye , or as close as possible to the ocular dioptric surface 3 , so that it can optionally be integrated therein , for this purpose , it allows a device to be produced that is better adapted to the observer &# 39 ; s morphology since the pathway of the path \u03b4 \u2033 is close to the curve of the observer &# 39 ; s head . in addition , to make the device 103 in fig3 or the device 104 in fig4 more compact , the side fold mirror 4 can be made in a single piece with the ocular dioptric surface 3 . therefore the side fold mirror 4 can be an integral part of the outer end i . e . in the illustrated example of the left end of the ocular dioptric surface 3 . for example , each of the side fold , reflective , back - inverting and / or decentring mirrors may be planar , concave or convex or substantially aspherical in a manner making it possible to improve the overall quality of the optical system . preferably a device of the invention comprises adjustment means , in particular sets of mirror and an ocular dioptric surface capable of adapting the configuration of the device , in particular the optic path , to the observer &# 39 ; s morphology ."}
|
Is the patent correctly categorized?
| 0.25 |
ab7dbe7acb7f81dc7aa04494c82020cf62dbf0961bcc8e2034548c678553ff66
| 0.722656 | 0.02002 | 0.925781 | 0.005737 | 0.890625 | 0.19043 |
null |
{"category": "Physics", "patent": "all the figures are schematic illustrations from above the head of a user or observer of each device . the left side of the observer is the left side in each figure . in all the figures , identical references designate parts or elements of parts having identical or similar functions . the figures refer to a device cooperating with the left eye of an observer . a device may be symmetrically provided to cooperate with the right eye of an observer . a device according to the invention may also comprise both a device cooperating with the left eye and a device cooperating with the right eye of the observer . fig1 illustrates an optical device 101 of the prior art , based on a method using the stigmatism of two foci of a substantially elliptical dioptric surface . the device , in the order of the optical pathway , chiefly consists of : a light display 1 ; lenses 2 ; and an ocular dioptric surface 3 of substantially elliptical shape represented by a portion of ellipse defined by its major axis \u03b4 , its small axis \u03b4 \u2032, its centre o and its two foci f and f \u2032 located on the major axis \u03b4 , either side of the centre o , as a function of the eccentricity e of the ellipse . in the figures , the axes \u03b4 , \u03b4 \u2032 of the dioptric surface 3 are shown as dot - dashed lines and the central optical path \u03b4 \u2033 is shown as a dotted line . in the example illustrated in fig1 , the device 101 is designed to cooperate with the observer &# 39 ; s left eye 6 . it is arranged on the left side of the observer &# 39 ; s head 7 . in addition , the major axis \u03b4 passes symmetrically through the two eyes of the observer and it is therefore perpendicular to a median plane p of the head 7 . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the central optical path \u03b4 \u2033 passes through one f of the foci of the dioptric surface then , after being reflected on said dioptric surface 3 , through the other focus f \u2032. the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. the spatial arrangements of the parts of device 101 in relation to the central optical path \u03b4 \u2033, on which they are aligned and centred , are inherent in the arrangement of the central optical path \u03b4 \u2033, which necessarily passes through the foci f and f \u2032. it is also ascertained that to observe this arrangement logic , the ocular dioptric surface 3 does not follow the periphery of the observer &# 39 ; s head at the height of the eye as does a conventional pair of spectacles . therefore the centre o of the dioptric surface is located largely outside , on the left of the left eye 6 . in addition the part 1 , 2 of the device upstream of the focus f moves significantly away from the left side of the observer &# 39 ; s head 7 over the distance between the focus f and the display 1 . this positioning of the device 101 in relation to the observer &# 39 ; s head 7 makes the device laterally bulky and of scarcely pleasing design . the device 102 illustrated in fig2 has more reduced lateral bulk than the device in fig1 . to reduce this bulkiness a side fold mirror 4 is arranged on the central optical path \u03b4 \u2033 in the vicinity of the focus f so that it is possible upstream of the mirror 4 to fold or direct the central optical path \u03b4 \u2033 substantially parallel to the median plane p and perpendicular to the major axis \u03b4 . therefore the device 102 in the order of the optical pathway is formed of : a light display 1 ; two groups of lenses 2 ; a side fold mirror 4 of planar , concave or convex shape ; and an ocular dioptric surface 3 of substantially elliptical shape represented by a portion of ellipse e defined by its major axis \u03b4 , its small axis \u03b4 \u2032, its centre o and its two foci f and f \u2032 located on the major axis \u03b4 either side of the centre o , as a function of the eccentricity e of the ellipse . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the side fold mirror 4 is arranged in the vicinity of the focus f so that it reflects the central optical path \u03b4 \u2033 at a chosen angle in the direction of the ocular dioptric surface 3 , so that the central optical path is then reflected in the direction of the observer &# 39 ; s eye , substantially perpendicular to the major axis \u03b4 . the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. in this configuration the bulkiness is significantly reduced since the central optical path \u03b4 \u2033 is folded back along the side of the observer &# 39 ; s head 7 . it is to be noted that the side fold mirror 4 in the example in fig2 is positioned in the immediate vicinity of the focus f on which it can be directed . the positioning thereof , so close to the focus f , is made necessary by the fact that it can offer a surface of minimum reflection . however the side fold mirror 4 can be arranged elsewhere on the central optical central path \u03b4 \u2033 downstream or upstream of the focus f according to the needs of the optical design . a description will now be given with reference to fig3 and 4 of two embodiments of a device according to the invention , in how they differ from the previously illustrated prior art . fig3 is an illustration of a first embodiment of a device according to the invention . the device 103 in fig3 , in the order of the optical pathway , is formed of : a light display 1 ; two groups of lenses 2 ; a side fold mirror 4 ; a back - inverting mirror 5 ; and an ocular dioptric surface 3 of substantially elliptical shape . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the side fold mirror 4 is arranged laterally in the vicinity of the observer &# 39 ; s eye 6 , in front of the observer &# 39 ; s temple , and it reflects the central optical path \u03b4 \u2033 at a chosen angle in the direction of the back - inverting mirror 5 . the back - inverting mirror 5 is arranged in the region of the upper part of the observer &# 39 ; s nose 10 on the right of and at the height of the left eye 6 and again reflects the central optical path \u03b4 \u2033 towards the ocular dioptric surface 3 . the mirrors are arranged such that the central optical path is then reflected by the dioptric surface in the direction of the observer &# 39 ; s eye 6 substantially perpendicular to the major axis \u03b4 . the observer &# 39 ; s eye 6 i . e . the centre of the pupil of the eye is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. according to the new optical scheme of the invention , the foci f and f \u2032 are inverted relative to the observer &# 39 ; s head . the back - inverting mirror 5 located in the vicinity of the focus f allows the virtual placing of that part of the central optical path \u03b4 \u2033 that is incident on the dioptric surface , and the focus f , in the observer &# 39 ; s head . that is to say that the display 1 is virtually placed inside the head 7 . yet in reality the central optical path \u03b4 \u2033 originates from the side of the observer &# 39 ; s head where the central optical path \u03b4 \u2033 was first folded or re - directed by the side fold mirror 4 . the side fold mirror 4 may now be located more distant from the focus f . the focus f is now directly related to the back - inverting mirror 5 . this gives the side fold mirror a much wider range of positioning , and hence of adjustment , the ocular dioptric surface 3 is now better adjusted to the morphological profile , curve , of the observer &# 39 ; s head in the vicinity of the eye , since the outer profile of the said ocular dioptric surface 3 tends to draw close to the side of the observer &# 39 ; s head 7 . the back - inverting mirror 5 is arranged in the vicinity of the upper part of the observer &# 39 ; s nose 10 on the pathway of the central optic path \u03b4 \u2033 between the side fold mirror and the ocular dioptric surface 3 . therefore the back - inverting mirror 5 can be arranged in an area hidden from the view of the observer , called a \u201c blind spot \u201d. the side fold mirror 4 is oriented angularly so that the central optic path \u03b4 \u2033 of the image is returned or reflected back towards the back - inverting mirror 5 and so that the back - inverting mirror 5 is oriented such that the central optic path \u03b4 \u2033 of the image is then returned or reflected towards the ocular dioptric surface 3 , the dioptric surface 3 being oriented so that the central optic path \u03b4 \u2033 of the image is then returned or reflected towards the observer &# 39 ; s eye 6 . in the embodiment in fig4 , the device 104 according to the invention is formed of : a light display 1 ; lenses 2 ; a side fold mirror 4 ; two decentring mirrors 8 and 9 arranged upstream of the fold mirror 4 , so that the they successively reflect the central optic path \u03b4 \u2033; a back - inverting mirror 5 of planar , convex or concave or aspherical shape ; an ocular dioptric surface 3 of substantially elliptical shape . the light display 1 diffuses an image whose pathway is illustrated by a central optic path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optic path \u03b4 \u2033. the central optic path is then reflected towards the right by the first decentring mirror 8 in the direction of the second decentring mirror 9 which is arranged in the vicinity of the observer &# 39 ; s temple . the second decentring mirror 9 then reflects the central path substantially forwardly in the direction of the side fold mirror 4 , which then reflects the same towards the right in the direction of the back - inverting mirror 5 . the back - inverting mirror 5 is arranged substantially against the upper left part of the observer &# 39 ; s nose 10 substantially at the height of the eye 6 . the central optic path \u03b4 \u2033 is then reflected thereat in the direction of the dioptric surface 3 which reflects it back towards the observer &# 39 ; s eye 6 . the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the region of the focus f \u2032. the device 104 in fig4 differs from the device 103 described with reference to fig3 in that it comprises a set of decentring mirrors 8 and 9 . the use of the set of decentring mirrors 8 , 9 has the following advantages : it allows correction of the adjustment of the angle reflection of the optical path \u03b4 \u2033 on the side mirror 4 , to offset positioning of the said side fold mirror 4 when it is brought as close as possible to the observer &# 39 ; s temple or eye , or as close as possible to the ocular dioptric surface 3 , so that it can optionally be integrated therein , for this purpose , it allows a device to be produced that is better adapted to the observer &# 39 ; s morphology since the pathway of the path \u03b4 \u2033 is close to the curve of the observer &# 39 ; s head . in addition , to make the device 103 in fig3 or the device 104 in fig4 more compact , the side fold mirror 4 can be made in a single piece with the ocular dioptric surface 3 . therefore the side fold mirror 4 can be an integral part of the outer end i . e . in the illustrated example of the left end of the ocular dioptric surface 3 . for example , each of the side fold , reflective , back - inverting and / or decentring mirrors may be planar , concave or convex or substantially aspherical in a manner making it possible to improve the overall quality of the optical system . preferably a device of the invention comprises adjustment means , in particular sets of mirror and an ocular dioptric surface capable of adapting the configuration of the device , in particular the optic path , to the observer &# 39 ; s morphology ."}
|
{"category": "Electricity", "patent": "all the figures are schematic illustrations from above the head of a user or observer of each device . the left side of the observer is the left side in each figure . in all the figures , identical references designate parts or elements of parts having identical or similar functions . the figures refer to a device cooperating with the left eye of an observer . a device may be symmetrically provided to cooperate with the right eye of an observer . a device according to the invention may also comprise both a device cooperating with the left eye and a device cooperating with the right eye of the observer . fig1 illustrates an optical device 101 of the prior art , based on a method using the stigmatism of two foci of a substantially elliptical dioptric surface . the device , in the order of the optical pathway , chiefly consists of : a light display 1 ; lenses 2 ; and an ocular dioptric surface 3 of substantially elliptical shape represented by a portion of ellipse defined by its major axis \u03b4 , its small axis \u03b4 \u2032, its centre o and its two foci f and f \u2032 located on the major axis \u03b4 , either side of the centre o , as a function of the eccentricity e of the ellipse . in the figures , the axes \u03b4 , \u03b4 \u2032 of the dioptric surface 3 are shown as dot - dashed lines and the central optical path \u03b4 \u2033 is shown as a dotted line . in the example illustrated in fig1 , the device 101 is designed to cooperate with the observer &# 39 ; s left eye 6 . it is arranged on the left side of the observer &# 39 ; s head 7 . in addition , the major axis \u03b4 passes symmetrically through the two eyes of the observer and it is therefore perpendicular to a median plane p of the head 7 . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the central optical path \u03b4 \u2033 passes through one f of the foci of the dioptric surface then , after being reflected on said dioptric surface 3 , through the other focus f \u2032. the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. the spatial arrangements of the parts of device 101 in relation to the central optical path \u03b4 \u2033, on which they are aligned and centred , are inherent in the arrangement of the central optical path \u03b4 \u2033, which necessarily passes through the foci f and f \u2032. it is also ascertained that to observe this arrangement logic , the ocular dioptric surface 3 does not follow the periphery of the observer &# 39 ; s head at the height of the eye as does a conventional pair of spectacles . therefore the centre o of the dioptric surface is located largely outside , on the left of the left eye 6 . in addition the part 1 , 2 of the device upstream of the focus f moves significantly away from the left side of the observer &# 39 ; s head 7 over the distance between the focus f and the display 1 . this positioning of the device 101 in relation to the observer &# 39 ; s head 7 makes the device laterally bulky and of scarcely pleasing design . the device 102 illustrated in fig2 has more reduced lateral bulk than the device in fig1 . to reduce this bulkiness a side fold mirror 4 is arranged on the central optical path \u03b4 \u2033 in the vicinity of the focus f so that it is possible upstream of the mirror 4 to fold or direct the central optical path \u03b4 \u2033 substantially parallel to the median plane p and perpendicular to the major axis \u03b4 . therefore the device 102 in the order of the optical pathway is formed of : a light display 1 ; two groups of lenses 2 ; a side fold mirror 4 of planar , concave or convex shape ; and an ocular dioptric surface 3 of substantially elliptical shape represented by a portion of ellipse e defined by its major axis \u03b4 , its small axis \u03b4 \u2032, its centre o and its two foci f and f \u2032 located on the major axis \u03b4 either side of the centre o , as a function of the eccentricity e of the ellipse . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the side fold mirror 4 is arranged in the vicinity of the focus f so that it reflects the central optical path \u03b4 \u2033 at a chosen angle in the direction of the ocular dioptric surface 3 , so that the central optical path is then reflected in the direction of the observer &# 39 ; s eye , substantially perpendicular to the major axis \u03b4 . the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. in this configuration the bulkiness is significantly reduced since the central optical path \u03b4 \u2033 is folded back along the side of the observer &# 39 ; s head 7 . it is to be noted that the side fold mirror 4 in the example in fig2 is positioned in the immediate vicinity of the focus f on which it can be directed . the positioning thereof , so close to the focus f , is made necessary by the fact that it can offer a surface of minimum reflection . however the side fold mirror 4 can be arranged elsewhere on the central optical central path \u03b4 \u2033 downstream or upstream of the focus f according to the needs of the optical design . a description will now be given with reference to fig3 and 4 of two embodiments of a device according to the invention , in how they differ from the previously illustrated prior art . fig3 is an illustration of a first embodiment of a device according to the invention . the device 103 in fig3 , in the order of the optical pathway , is formed of : a light display 1 ; two groups of lenses 2 ; a side fold mirror 4 ; a back - inverting mirror 5 ; and an ocular dioptric surface 3 of substantially elliptical shape . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the side fold mirror 4 is arranged laterally in the vicinity of the observer &# 39 ; s eye 6 , in front of the observer &# 39 ; s temple , and it reflects the central optical path \u03b4 \u2033 at a chosen angle in the direction of the back - inverting mirror 5 . the back - inverting mirror 5 is arranged in the region of the upper part of the observer &# 39 ; s nose 10 on the right of and at the height of the left eye 6 and again reflects the central optical path \u03b4 \u2033 towards the ocular dioptric surface 3 . the mirrors are arranged such that the central optical path is then reflected by the dioptric surface in the direction of the observer &# 39 ; s eye 6 substantially perpendicular to the major axis \u03b4 . the observer &# 39 ; s eye 6 i . e . the centre of the pupil of the eye is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. according to the new optical scheme of the invention , the foci f and f \u2032 are inverted relative to the observer &# 39 ; s head . the back - inverting mirror 5 located in the vicinity of the focus f allows the virtual placing of that part of the central optical path \u03b4 \u2033 that is incident on the dioptric surface , and the focus f , in the observer &# 39 ; s head . that is to say that the display 1 is virtually placed inside the head 7 . yet in reality the central optical path \u03b4 \u2033 originates from the side of the observer &# 39 ; s head where the central optical path \u03b4 \u2033 was first folded or re - directed by the side fold mirror 4 . the side fold mirror 4 may now be located more distant from the focus f . the focus f is now directly related to the back - inverting mirror 5 . this gives the side fold mirror a much wider range of positioning , and hence of adjustment , the ocular dioptric surface 3 is now better adjusted to the morphological profile , curve , of the observer &# 39 ; s head in the vicinity of the eye , since the outer profile of the said ocular dioptric surface 3 tends to draw close to the side of the observer &# 39 ; s head 7 . the back - inverting mirror 5 is arranged in the vicinity of the upper part of the observer &# 39 ; s nose 10 on the pathway of the central optic path \u03b4 \u2033 between the side fold mirror and the ocular dioptric surface 3 . therefore the back - inverting mirror 5 can be arranged in an area hidden from the view of the observer , called a \u201c blind spot \u201d. the side fold mirror 4 is oriented angularly so that the central optic path \u03b4 \u2033 of the image is returned or reflected back towards the back - inverting mirror 5 and so that the back - inverting mirror 5 is oriented such that the central optic path \u03b4 \u2033 of the image is then returned or reflected towards the ocular dioptric surface 3 , the dioptric surface 3 being oriented so that the central optic path \u03b4 \u2033 of the image is then returned or reflected towards the observer &# 39 ; s eye 6 . in the embodiment in fig4 , the device 104 according to the invention is formed of : a light display 1 ; lenses 2 ; a side fold mirror 4 ; two decentring mirrors 8 and 9 arranged upstream of the fold mirror 4 , so that the they successively reflect the central optic path \u03b4 \u2033; a back - inverting mirror 5 of planar , convex or concave or aspherical shape ; an ocular dioptric surface 3 of substantially elliptical shape . the light display 1 diffuses an image whose pathway is illustrated by a central optic path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optic path \u03b4 \u2033. the central optic path is then reflected towards the right by the first decentring mirror 8 in the direction of the second decentring mirror 9 which is arranged in the vicinity of the observer &# 39 ; s temple . the second decentring mirror 9 then reflects the central path substantially forwardly in the direction of the side fold mirror 4 , which then reflects the same towards the right in the direction of the back - inverting mirror 5 . the back - inverting mirror 5 is arranged substantially against the upper left part of the observer &# 39 ; s nose 10 substantially at the height of the eye 6 . the central optic path \u03b4 \u2033 is then reflected thereat in the direction of the dioptric surface 3 which reflects it back towards the observer &# 39 ; s eye 6 . the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the region of the focus f \u2032. the device 104 in fig4 differs from the device 103 described with reference to fig3 in that it comprises a set of decentring mirrors 8 and 9 . the use of the set of decentring mirrors 8 , 9 has the following advantages : it allows correction of the adjustment of the angle reflection of the optical path \u03b4 \u2033 on the side mirror 4 , to offset positioning of the said side fold mirror 4 when it is brought as close as possible to the observer &# 39 ; s temple or eye , or as close as possible to the ocular dioptric surface 3 , so that it can optionally be integrated therein , for this purpose , it allows a device to be produced that is better adapted to the observer &# 39 ; s morphology since the pathway of the path \u03b4 \u2033 is close to the curve of the observer &# 39 ; s head . in addition , to make the device 103 in fig3 or the device 104 in fig4 more compact , the side fold mirror 4 can be made in a single piece with the ocular dioptric surface 3 . therefore the side fold mirror 4 can be an integral part of the outer end i . e . in the illustrated example of the left end of the ocular dioptric surface 3 . for example , each of the side fold , reflective , back - inverting and / or decentring mirrors may be planar , concave or convex or substantially aspherical in a manner making it possible to improve the overall quality of the optical system . preferably a device of the invention comprises adjustment means , in particular sets of mirror and an ocular dioptric surface capable of adapting the configuration of the device , in particular the optic path , to the observer &# 39 ; s morphology ."}
|
Is the category the most suitable category for the given patent?
| 0.25 |
ab7dbe7acb7f81dc7aa04494c82020cf62dbf0961bcc8e2034548c678553ff66
| 0.699219 | 0.017456 | 0.235352 | 0.015869 | 0.738281 | 0.016357 |
null |
{"patent": "all the figures are schematic illustrations from above the head of a user or observer of each device . the left side of the observer is the left side in each figure . in all the figures , identical references designate parts or elements of parts having identical or similar functions . the figures refer to a device cooperating with the left eye of an observer . a device may be symmetrically provided to cooperate with the right eye of an observer . a device according to the invention may also comprise both a device cooperating with the left eye and a device cooperating with the right eye of the observer . fig1 illustrates an optical device 101 of the prior art , based on a method using the stigmatism of two foci of a substantially elliptical dioptric surface . the device , in the order of the optical pathway , chiefly consists of : a light display 1 ; lenses 2 ; and an ocular dioptric surface 3 of substantially elliptical shape represented by a portion of ellipse defined by its major axis \u03b4 , its small axis \u03b4 \u2032, its centre o and its two foci f and f \u2032 located on the major axis \u03b4 , either side of the centre o , as a function of the eccentricity e of the ellipse . in the figures , the axes \u03b4 , \u03b4 \u2032 of the dioptric surface 3 are shown as dot - dashed lines and the central optical path \u03b4 \u2033 is shown as a dotted line . in the example illustrated in fig1 , the device 101 is designed to cooperate with the observer &# 39 ; s left eye 6 . it is arranged on the left side of the observer &# 39 ; s head 7 . in addition , the major axis \u03b4 passes symmetrically through the two eyes of the observer and it is therefore perpendicular to a median plane p of the head 7 . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the central optical path \u03b4 \u2033 passes through one f of the foci of the dioptric surface then , after being reflected on said dioptric surface 3 , through the other focus f \u2032. the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. the spatial arrangements of the parts of device 101 in relation to the central optical path \u03b4 \u2033, on which they are aligned and centred , are inherent in the arrangement of the central optical path \u03b4 \u2033, which necessarily passes through the foci f and f \u2032. it is also ascertained that to observe this arrangement logic , the ocular dioptric surface 3 does not follow the periphery of the observer &# 39 ; s head at the height of the eye as does a conventional pair of spectacles . therefore the centre o of the dioptric surface is located largely outside , on the left of the left eye 6 . in addition the part 1 , 2 of the device upstream of the focus f moves significantly away from the left side of the observer &# 39 ; s head 7 over the distance between the focus f and the display 1 . this positioning of the device 101 in relation to the observer &# 39 ; s head 7 makes the device laterally bulky and of scarcely pleasing design . the device 102 illustrated in fig2 has more reduced lateral bulk than the device in fig1 . to reduce this bulkiness a side fold mirror 4 is arranged on the central optical path \u03b4 \u2033 in the vicinity of the focus f so that it is possible upstream of the mirror 4 to fold or direct the central optical path \u03b4 \u2033 substantially parallel to the median plane p and perpendicular to the major axis \u03b4 . therefore the device 102 in the order of the optical pathway is formed of : a light display 1 ; two groups of lenses 2 ; a side fold mirror 4 of planar , concave or convex shape ; and an ocular dioptric surface 3 of substantially elliptical shape represented by a portion of ellipse e defined by its major axis \u03b4 , its small axis \u03b4 \u2032, its centre o and its two foci f and f \u2032 located on the major axis \u03b4 either side of the centre o , as a function of the eccentricity e of the ellipse . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the side fold mirror 4 is arranged in the vicinity of the focus f so that it reflects the central optical path \u03b4 \u2033 at a chosen angle in the direction of the ocular dioptric surface 3 , so that the central optical path is then reflected in the direction of the observer &# 39 ; s eye , substantially perpendicular to the major axis \u03b4 . the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. in this configuration the bulkiness is significantly reduced since the central optical path \u03b4 \u2033 is folded back along the side of the observer &# 39 ; s head 7 . it is to be noted that the side fold mirror 4 in the example in fig2 is positioned in the immediate vicinity of the focus f on which it can be directed . the positioning thereof , so close to the focus f , is made necessary by the fact that it can offer a surface of minimum reflection . however the side fold mirror 4 can be arranged elsewhere on the central optical central path \u03b4 \u2033 downstream or upstream of the focus f according to the needs of the optical design . a description will now be given with reference to fig3 and 4 of two embodiments of a device according to the invention , in how they differ from the previously illustrated prior art . fig3 is an illustration of a first embodiment of a device according to the invention . the device 103 in fig3 , in the order of the optical pathway , is formed of : a light display 1 ; two groups of lenses 2 ; a side fold mirror 4 ; a back - inverting mirror 5 ; and an ocular dioptric surface 3 of substantially elliptical shape . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the side fold mirror 4 is arranged laterally in the vicinity of the observer &# 39 ; s eye 6 , in front of the observer &# 39 ; s temple , and it reflects the central optical path \u03b4 \u2033 at a chosen angle in the direction of the back - inverting mirror 5 . the back - inverting mirror 5 is arranged in the region of the upper part of the observer &# 39 ; s nose 10 on the right of and at the height of the left eye 6 and again reflects the central optical path \u03b4 \u2033 towards the ocular dioptric surface 3 . the mirrors are arranged such that the central optical path is then reflected by the dioptric surface in the direction of the observer &# 39 ; s eye 6 substantially perpendicular to the major axis \u03b4 . the observer &# 39 ; s eye 6 i . e . the centre of the pupil of the eye is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. according to the new optical scheme of the invention , the foci f and f \u2032 are inverted relative to the observer &# 39 ; s head . the back - inverting mirror 5 located in the vicinity of the focus f allows the virtual placing of that part of the central optical path \u03b4 \u2033 that is incident on the dioptric surface , and the focus f , in the observer &# 39 ; s head . that is to say that the display 1 is virtually placed inside the head 7 . yet in reality the central optical path \u03b4 \u2033 originates from the side of the observer &# 39 ; s head where the central optical path \u03b4 \u2033 was first folded or re - directed by the side fold mirror 4 . the side fold mirror 4 may now be located more distant from the focus f . the focus f is now directly related to the back - inverting mirror 5 . this gives the side fold mirror a much wider range of positioning , and hence of adjustment , the ocular dioptric surface 3 is now better adjusted to the morphological profile , curve , of the observer &# 39 ; s head in the vicinity of the eye , since the outer profile of the said ocular dioptric surface 3 tends to draw close to the side of the observer &# 39 ; s head 7 . the back - inverting mirror 5 is arranged in the vicinity of the upper part of the observer &# 39 ; s nose 10 on the pathway of the central optic path \u03b4 \u2033 between the side fold mirror and the ocular dioptric surface 3 . therefore the back - inverting mirror 5 can be arranged in an area hidden from the view of the observer , called a \u201c blind spot \u201d. the side fold mirror 4 is oriented angularly so that the central optic path \u03b4 \u2033 of the image is returned or reflected back towards the back - inverting mirror 5 and so that the back - inverting mirror 5 is oriented such that the central optic path \u03b4 \u2033 of the image is then returned or reflected towards the ocular dioptric surface 3 , the dioptric surface 3 being oriented so that the central optic path \u03b4 \u2033 of the image is then returned or reflected towards the observer &# 39 ; s eye 6 . in the embodiment in fig4 , the device 104 according to the invention is formed of : a light display 1 ; lenses 2 ; a side fold mirror 4 ; two decentring mirrors 8 and 9 arranged upstream of the fold mirror 4 , so that the they successively reflect the central optic path \u03b4 \u2033; a back - inverting mirror 5 of planar , convex or concave or aspherical shape ; an ocular dioptric surface 3 of substantially elliptical shape . the light display 1 diffuses an image whose pathway is illustrated by a central optic path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optic path \u03b4 \u2033. the central optic path is then reflected towards the right by the first decentring mirror 8 in the direction of the second decentring mirror 9 which is arranged in the vicinity of the observer &# 39 ; s temple . the second decentring mirror 9 then reflects the central path substantially forwardly in the direction of the side fold mirror 4 , which then reflects the same towards the right in the direction of the back - inverting mirror 5 . the back - inverting mirror 5 is arranged substantially against the upper left part of the observer &# 39 ; s nose 10 substantially at the height of the eye 6 . the central optic path \u03b4 \u2033 is then reflected thereat in the direction of the dioptric surface 3 which reflects it back towards the observer &# 39 ; s eye 6 . the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the region of the focus f \u2032. the device 104 in fig4 differs from the device 103 described with reference to fig3 in that it comprises a set of decentring mirrors 8 and 9 . the use of the set of decentring mirrors 8 , 9 has the following advantages : it allows correction of the adjustment of the angle reflection of the optical path \u03b4 \u2033 on the side mirror 4 , to offset positioning of the said side fold mirror 4 when it is brought as close as possible to the observer &# 39 ; s temple or eye , or as close as possible to the ocular dioptric surface 3 , so that it can optionally be integrated therein , for this purpose , it allows a device to be produced that is better adapted to the observer &# 39 ; s morphology since the pathway of the path \u03b4 \u2033 is close to the curve of the observer &# 39 ; s head . in addition , to make the device 103 in fig3 or the device 104 in fig4 more compact , the side fold mirror 4 can be made in a single piece with the ocular dioptric surface 3 . therefore the side fold mirror 4 can be an integral part of the outer end i . e . in the illustrated example of the left end of the ocular dioptric surface 3 . for example , each of the side fold , reflective , back - inverting and / or decentring mirrors may be planar , concave or convex or substantially aspherical in a manner making it possible to improve the overall quality of the optical system . preferably a device of the invention comprises adjustment means , in particular sets of mirror and an ocular dioptric surface capable of adapting the configuration of the device , in particular the optic path , to the observer &# 39 ; s morphology .", "category": "Physics"}
|
{"category": "General tagging of new or cross-sectional technology", "patent": "all the figures are schematic illustrations from above the head of a user or observer of each device . the left side of the observer is the left side in each figure . in all the figures , identical references designate parts or elements of parts having identical or similar functions . the figures refer to a device cooperating with the left eye of an observer . a device may be symmetrically provided to cooperate with the right eye of an observer . a device according to the invention may also comprise both a device cooperating with the left eye and a device cooperating with the right eye of the observer . fig1 illustrates an optical device 101 of the prior art , based on a method using the stigmatism of two foci of a substantially elliptical dioptric surface . the device , in the order of the optical pathway , chiefly consists of : a light display 1 ; lenses 2 ; and an ocular dioptric surface 3 of substantially elliptical shape represented by a portion of ellipse defined by its major axis \u03b4 , its small axis \u03b4 \u2032, its centre o and its two foci f and f \u2032 located on the major axis \u03b4 , either side of the centre o , as a function of the eccentricity e of the ellipse . in the figures , the axes \u03b4 , \u03b4 \u2032 of the dioptric surface 3 are shown as dot - dashed lines and the central optical path \u03b4 \u2033 is shown as a dotted line . in the example illustrated in fig1 , the device 101 is designed to cooperate with the observer &# 39 ; s left eye 6 . it is arranged on the left side of the observer &# 39 ; s head 7 . in addition , the major axis \u03b4 passes symmetrically through the two eyes of the observer and it is therefore perpendicular to a median plane p of the head 7 . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the central optical path \u03b4 \u2033 passes through one f of the foci of the dioptric surface then , after being reflected on said dioptric surface 3 , through the other focus f \u2032. the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. the spatial arrangements of the parts of device 101 in relation to the central optical path \u03b4 \u2033, on which they are aligned and centred , are inherent in the arrangement of the central optical path \u03b4 \u2033, which necessarily passes through the foci f and f \u2032. it is also ascertained that to observe this arrangement logic , the ocular dioptric surface 3 does not follow the periphery of the observer &# 39 ; s head at the height of the eye as does a conventional pair of spectacles . therefore the centre o of the dioptric surface is located largely outside , on the left of the left eye 6 . in addition the part 1 , 2 of the device upstream of the focus f moves significantly away from the left side of the observer &# 39 ; s head 7 over the distance between the focus f and the display 1 . this positioning of the device 101 in relation to the observer &# 39 ; s head 7 makes the device laterally bulky and of scarcely pleasing design . the device 102 illustrated in fig2 has more reduced lateral bulk than the device in fig1 . to reduce this bulkiness a side fold mirror 4 is arranged on the central optical path \u03b4 \u2033 in the vicinity of the focus f so that it is possible upstream of the mirror 4 to fold or direct the central optical path \u03b4 \u2033 substantially parallel to the median plane p and perpendicular to the major axis \u03b4 . therefore the device 102 in the order of the optical pathway is formed of : a light display 1 ; two groups of lenses 2 ; a side fold mirror 4 of planar , concave or convex shape ; and an ocular dioptric surface 3 of substantially elliptical shape represented by a portion of ellipse e defined by its major axis \u03b4 , its small axis \u03b4 \u2032, its centre o and its two foci f and f \u2032 located on the major axis \u03b4 either side of the centre o , as a function of the eccentricity e of the ellipse . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the side fold mirror 4 is arranged in the vicinity of the focus f so that it reflects the central optical path \u03b4 \u2033 at a chosen angle in the direction of the ocular dioptric surface 3 , so that the central optical path is then reflected in the direction of the observer &# 39 ; s eye , substantially perpendicular to the major axis \u03b4 . the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. in this configuration the bulkiness is significantly reduced since the central optical path \u03b4 \u2033 is folded back along the side of the observer &# 39 ; s head 7 . it is to be noted that the side fold mirror 4 in the example in fig2 is positioned in the immediate vicinity of the focus f on which it can be directed . the positioning thereof , so close to the focus f , is made necessary by the fact that it can offer a surface of minimum reflection . however the side fold mirror 4 can be arranged elsewhere on the central optical central path \u03b4 \u2033 downstream or upstream of the focus f according to the needs of the optical design . a description will now be given with reference to fig3 and 4 of two embodiments of a device according to the invention , in how they differ from the previously illustrated prior art . fig3 is an illustration of a first embodiment of a device according to the invention . the device 103 in fig3 , in the order of the optical pathway , is formed of : a light display 1 ; two groups of lenses 2 ; a side fold mirror 4 ; a back - inverting mirror 5 ; and an ocular dioptric surface 3 of substantially elliptical shape . the light display 1 diffuses an image whose pathway is represented by a central optical path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optical path \u03b4 \u2033. the side fold mirror 4 is arranged laterally in the vicinity of the observer &# 39 ; s eye 6 , in front of the observer &# 39 ; s temple , and it reflects the central optical path \u03b4 \u2033 at a chosen angle in the direction of the back - inverting mirror 5 . the back - inverting mirror 5 is arranged in the region of the upper part of the observer &# 39 ; s nose 10 on the right of and at the height of the left eye 6 and again reflects the central optical path \u03b4 \u2033 towards the ocular dioptric surface 3 . the mirrors are arranged such that the central optical path is then reflected by the dioptric surface in the direction of the observer &# 39 ; s eye 6 substantially perpendicular to the major axis \u03b4 . the observer &# 39 ; s eye 6 i . e . the centre of the pupil of the eye is approximately aligned and centred on the central optical path \u03b4 \u2033 in the vicinity of the focus f \u2032. according to the new optical scheme of the invention , the foci f and f \u2032 are inverted relative to the observer &# 39 ; s head . the back - inverting mirror 5 located in the vicinity of the focus f allows the virtual placing of that part of the central optical path \u03b4 \u2033 that is incident on the dioptric surface , and the focus f , in the observer &# 39 ; s head . that is to say that the display 1 is virtually placed inside the head 7 . yet in reality the central optical path \u03b4 \u2033 originates from the side of the observer &# 39 ; s head where the central optical path \u03b4 \u2033 was first folded or re - directed by the side fold mirror 4 . the side fold mirror 4 may now be located more distant from the focus f . the focus f is now directly related to the back - inverting mirror 5 . this gives the side fold mirror a much wider range of positioning , and hence of adjustment , the ocular dioptric surface 3 is now better adjusted to the morphological profile , curve , of the observer &# 39 ; s head in the vicinity of the eye , since the outer profile of the said ocular dioptric surface 3 tends to draw close to the side of the observer &# 39 ; s head 7 . the back - inverting mirror 5 is arranged in the vicinity of the upper part of the observer &# 39 ; s nose 10 on the pathway of the central optic path \u03b4 \u2033 between the side fold mirror and the ocular dioptric surface 3 . therefore the back - inverting mirror 5 can be arranged in an area hidden from the view of the observer , called a \u201c blind spot \u201d. the side fold mirror 4 is oriented angularly so that the central optic path \u03b4 \u2033 of the image is returned or reflected back towards the back - inverting mirror 5 and so that the back - inverting mirror 5 is oriented such that the central optic path \u03b4 \u2033 of the image is then returned or reflected towards the ocular dioptric surface 3 , the dioptric surface 3 being oriented so that the central optic path \u03b4 \u2033 of the image is then returned or reflected towards the observer &# 39 ; s eye 6 . in the embodiment in fig4 , the device 104 according to the invention is formed of : a light display 1 ; lenses 2 ; a side fold mirror 4 ; two decentring mirrors 8 and 9 arranged upstream of the fold mirror 4 , so that the they successively reflect the central optic path \u03b4 \u2033; a back - inverting mirror 5 of planar , convex or concave or aspherical shape ; an ocular dioptric surface 3 of substantially elliptical shape . the light display 1 diffuses an image whose pathway is illustrated by a central optic path \u03b4 \u2033. the lenses 2 are aligned and centred on the central optic path \u03b4 \u2033. the central optic path is then reflected towards the right by the first decentring mirror 8 in the direction of the second decentring mirror 9 which is arranged in the vicinity of the observer &# 39 ; s temple . the second decentring mirror 9 then reflects the central path substantially forwardly in the direction of the side fold mirror 4 , which then reflects the same towards the right in the direction of the back - inverting mirror 5 . the back - inverting mirror 5 is arranged substantially against the upper left part of the observer &# 39 ; s nose 10 substantially at the height of the eye 6 . the central optic path \u03b4 \u2033 is then reflected thereat in the direction of the dioptric surface 3 which reflects it back towards the observer &# 39 ; s eye 6 . the observer &# 39 ; s eye 6 is approximately aligned and centred on the central optical path \u03b4 \u2033 in the region of the focus f \u2032. the device 104 in fig4 differs from the device 103 described with reference to fig3 in that it comprises a set of decentring mirrors 8 and 9 . the use of the set of decentring mirrors 8 , 9 has the following advantages : it allows correction of the adjustment of the angle reflection of the optical path \u03b4 \u2033 on the side mirror 4 , to offset positioning of the said side fold mirror 4 when it is brought as close as possible to the observer &# 39 ; s temple or eye , or as close as possible to the ocular dioptric surface 3 , so that it can optionally be integrated therein , for this purpose , it allows a device to be produced that is better adapted to the observer &# 39 ; s morphology since the pathway of the path \u03b4 \u2033 is close to the curve of the observer &# 39 ; s head . in addition , to make the device 103 in fig3 or the device 104 in fig4 more compact , the side fold mirror 4 can be made in a single piece with the ocular dioptric surface 3 . therefore the side fold mirror 4 can be an integral part of the outer end i . e . in the illustrated example of the left end of the ocular dioptric surface 3 . for example , each of the side fold , reflective , back - inverting and / or decentring mirrors may be planar , concave or convex or substantially aspherical in a manner making it possible to improve the overall quality of the optical system . preferably a device of the invention comprises adjustment means , in particular sets of mirror and an ocular dioptric surface capable of adapting the configuration of the device , in particular the optic path , to the observer &# 39 ; s morphology ."}
|
Does the patent belong in this category?
| 0.25 |
ab7dbe7acb7f81dc7aa04494c82020cf62dbf0961bcc8e2034548c678553ff66
| 0.133789 | 0.621094 | 0.102539 | 0.789063 | 0.3125 | 0.808594 |
null |
{"patent": "a spare web roll , for example a spare paper web roll 1 for a web - fed rotary printing press , has a continuous tube 3 in the area of its axis of rotation 2 , on whose tube body a web 4 of material has been wound , all as seen in fig1 and 2 . a star - shaped , for example three - armed holder 7 is arranged in the end area of the tube , and in particular in the area of the interior diameter , or respectively interior 6 of the tube 3 , whose arms 8 , 9 , 10 rest resiliently against the interior circumference 6 . for example , the holder 7 supports a cylinder - shaped or disk - shaped code carrier 12 , which can be activated , i . e . can be coded and decoded , coaxially with the axis of rotation 2 . the code carrier 12 is an electronic component , which can be coded by inductive means . the holder 7 can be inserted , for example into the end area of the tube 3 of a spare paper web 1 roll , i . e . into the interior 6 of the tube 3 , in such a way that it remains in a temporary first position b , i . e . close to the lateral edges 13 of the web of material 4 as \u2014 represented in dashed lines in fig1 . only when the spare paper web 1 roll is placed on the shaft , is the holder 7 pushed into a second position c by means of a clamping cone , not represented . the holder 7 will now be in the position shown in solid lines in fig1 . after the roll is used up , for example , the holder 7 can again be taken out of the tube 3 by removing it . for this purpose the arms 8 , 9 , 10 can have bores 14 , 16 , 17 , for example , behind which a suitable tool can reach . a number of code carriers 18 , 21 , 23 , 26 are represented in respectively different fastening positions in fig3 . however , only one of which is used . a code carrier 18 is arranged in the front face 19 of the tube 3 , i . e . in the tube body , and may for example be pressed into a bore . a code carrier 21 is fastened to the inner surface 22 of the tube 3 , i . e . in the interior 6 of the tube 3 , for example by gluing . a code carrier 23 is arranged in the tube body 3 . in this case , the tube wall 24 can have a bore into which the code carrier 23 is pressed . it is also possible to arrange a code carrier 26 in the webs of material 4 which are close to the tube 3 , i . e . the inner or lower ones , i . e . to introduce it from the direction of the lateral edges 13 of the webs of material 4 into an area close to the edge , for example to press it in . therefore the code carrier 26 is outside of and in the vicinity of the tube 3 . the code carriers 12 , 18 , 21 , 23 , 26 arranged in the interior of the tube 3 or in the vicinity of the tube 3 , for example the code carrier 18 located in the front face 19 of the tube 3 , are connected with one of the recording and reading heads 27 , 28 , 29 . such recording and reading heads 27 , 28 , 29 are , as shown in fig4 stationed along a conveyance path for movable spare paper web transport carriages , for example the recording and reading head 27 in the vicinity of a storage facility for spare paper web rolls 1 , the recording and reading head 28 is locked in the vicinity of a preparation station for spare paper web , and the recording and reading head 29 is located in the vicinity of a roll changer . each recording and reading head 27 to 29 consists of an electronic component , which transmits energy 37 and information 38 in a contactless manner to the code carrier 18 . on the other end , all recording and reading heads 27 to 29 are connected with each other via a data bus 39 , as well as with an evaluation unit 41 and a control device 42 , as well as a master computer 43 , all as shown in fig4 . the master computer 43 has an input station 46 . the code carrier 18 , as well as the other code carriers 12 , 21 , 23 , 26 , are essentially embodied as data memory devices , and respectively consist of an electronic component with a memory and logical control device . the above mentioned code carriers are , for example , designed as eeprom versions , and can be written on and read out , for example , with up to eight kilobytes . if , for example , a spare paper web roll 1 is moved out of the storage facility \u2014 the transport carriages are , for example , pulled by means of under - floor or driverless transport systems \u2014 the code carrier 18 located in the tube 3 of the spare paper web roll 1 passes by the recording / reading head 27 . as soon as the code carrier 18 comes into the active area of the recording / reading head 27 , the energy 37 required for data transmission is built up and the data 38 are transmitted by the recording / reading head 27 to the code carrier 18 , and the data 44 are transmitted by the code carrier 18 to the recording / reading head 27 , each in a frequency range of , for example , 70 kilohertz . such data can be , besides the data entered by the manufacturer : new destination , for example the spare paper web preparation station , remaining length of paper , particularities of the paper , for example a \u201c beating \u201d roll . the data , which are inductively coupled in by the code carrier 18 , are converted into a digital energy signal and conducted to the evaluation unit 41 . the evaluation unit 41 manages the data transfer 37 , 44 between the recording / reading head 28 and the code carrier 18 and is used as an intermediate memory . the evaluation unit 41 is the connecting member between the master computer 43 , or respectively the control device 42 , and the code carrier 18 . data can be erased or new data can be added , such as for example , a new destination , the gluing preparation station and the type of the glue tip . following the gluing preparation of the spare paper web roll 1 , it is possible to enter a further destination , for example a roll changer , by means of the recording / reading head 28 . after a portion of the spare paper web roll 1 has been used , another entry to the roll data takes place by means of the recording / reading head 29 , as well as a new destination , for example a storage facility . it is also possible to arrange only stationary reading devices for the purpose of reading out the destination at appropriate positions on the transport path , for example at switches . in accordance with a further preferred embodiment , the recording / reading head 29 can be arranged in the roll changer , i . e . for example in a roll cone of a support arm of the roll changer , so that a data exchange can take place between the recording / reading head 29 and a code carrier 18 while the spare paper web roll 1 is still on the shaft . alternatively it is also possible to fasten the recording / reading head 29 on the end of the support arm of the roll changer . the data exchange preferably takes place during the run - out of the spare paper web roll 1 , i . e . while the spare paper web roll 1 turns slowly . it is also possible to utilize a code carrier 12 arranged in the area of the imagined axis of rotation 2 of the spare paper web roll 1 in the interior 6 of the tube 3 . in this case the recording / reading head 12 is arranged in the direction of an extended axis of rotation 2 on the exterior of a support arm of the roll changer . here , the clamping cone is designed to be hollow . while preferred embodiments of a spare paper roll in accordance with the present invention have been set forth fully and completely hereinabove , it will be apparent to one of skill in the art that a number of changes in , for example , the material used to construct the tube , the type of press with which the spare paper roll will be used , and the like may be made without departing from the true spirit and scope of the present invention which is accordingly to be limited only by the following claims .", "category": "Performing Operations; Transporting"}
|
{"category": "Human Necessities", "patent": "a spare web roll , for example a spare paper web roll 1 for a web - fed rotary printing press , has a continuous tube 3 in the area of its axis of rotation 2 , on whose tube body a web 4 of material has been wound , all as seen in fig1 and 2 . a star - shaped , for example three - armed holder 7 is arranged in the end area of the tube , and in particular in the area of the interior diameter , or respectively interior 6 of the tube 3 , whose arms 8 , 9 , 10 rest resiliently against the interior circumference 6 . for example , the holder 7 supports a cylinder - shaped or disk - shaped code carrier 12 , which can be activated , i . e . can be coded and decoded , coaxially with the axis of rotation 2 . the code carrier 12 is an electronic component , which can be coded by inductive means . the holder 7 can be inserted , for example into the end area of the tube 3 of a spare paper web 1 roll , i . e . into the interior 6 of the tube 3 , in such a way that it remains in a temporary first position b , i . e . close to the lateral edges 13 of the web of material 4 as \u2014 represented in dashed lines in fig1 . only when the spare paper web 1 roll is placed on the shaft , is the holder 7 pushed into a second position c by means of a clamping cone , not represented . the holder 7 will now be in the position shown in solid lines in fig1 . after the roll is used up , for example , the holder 7 can again be taken out of the tube 3 by removing it . for this purpose the arms 8 , 9 , 10 can have bores 14 , 16 , 17 , for example , behind which a suitable tool can reach . a number of code carriers 18 , 21 , 23 , 26 are represented in respectively different fastening positions in fig3 . however , only one of which is used . a code carrier 18 is arranged in the front face 19 of the tube 3 , i . e . in the tube body , and may for example be pressed into a bore . a code carrier 21 is fastened to the inner surface 22 of the tube 3 , i . e . in the interior 6 of the tube 3 , for example by gluing . a code carrier 23 is arranged in the tube body 3 . in this case , the tube wall 24 can have a bore into which the code carrier 23 is pressed . it is also possible to arrange a code carrier 26 in the webs of material 4 which are close to the tube 3 , i . e . the inner or lower ones , i . e . to introduce it from the direction of the lateral edges 13 of the webs of material 4 into an area close to the edge , for example to press it in . therefore the code carrier 26 is outside of and in the vicinity of the tube 3 . the code carriers 12 , 18 , 21 , 23 , 26 arranged in the interior of the tube 3 or in the vicinity of the tube 3 , for example the code carrier 18 located in the front face 19 of the tube 3 , are connected with one of the recording and reading heads 27 , 28 , 29 . such recording and reading heads 27 , 28 , 29 are , as shown in fig4 stationed along a conveyance path for movable spare paper web transport carriages , for example the recording and reading head 27 in the vicinity of a storage facility for spare paper web rolls 1 , the recording and reading head 28 is locked in the vicinity of a preparation station for spare paper web , and the recording and reading head 29 is located in the vicinity of a roll changer . each recording and reading head 27 to 29 consists of an electronic component , which transmits energy 37 and information 38 in a contactless manner to the code carrier 18 . on the other end , all recording and reading heads 27 to 29 are connected with each other via a data bus 39 , as well as with an evaluation unit 41 and a control device 42 , as well as a master computer 43 , all as shown in fig4 . the master computer 43 has an input station 46 . the code carrier 18 , as well as the other code carriers 12 , 21 , 23 , 26 , are essentially embodied as data memory devices , and respectively consist of an electronic component with a memory and logical control device . the above mentioned code carriers are , for example , designed as eeprom versions , and can be written on and read out , for example , with up to eight kilobytes . if , for example , a spare paper web roll 1 is moved out of the storage facility \u2014 the transport carriages are , for example , pulled by means of under - floor or driverless transport systems \u2014 the code carrier 18 located in the tube 3 of the spare paper web roll 1 passes by the recording / reading head 27 . as soon as the code carrier 18 comes into the active area of the recording / reading head 27 , the energy 37 required for data transmission is built up and the data 38 are transmitted by the recording / reading head 27 to the code carrier 18 , and the data 44 are transmitted by the code carrier 18 to the recording / reading head 27 , each in a frequency range of , for example , 70 kilohertz . such data can be , besides the data entered by the manufacturer : new destination , for example the spare paper web preparation station , remaining length of paper , particularities of the paper , for example a \u201c beating \u201d roll . the data , which are inductively coupled in by the code carrier 18 , are converted into a digital energy signal and conducted to the evaluation unit 41 . the evaluation unit 41 manages the data transfer 37 , 44 between the recording / reading head 28 and the code carrier 18 and is used as an intermediate memory . the evaluation unit 41 is the connecting member between the master computer 43 , or respectively the control device 42 , and the code carrier 18 . data can be erased or new data can be added , such as for example , a new destination , the gluing preparation station and the type of the glue tip . following the gluing preparation of the spare paper web roll 1 , it is possible to enter a further destination , for example a roll changer , by means of the recording / reading head 28 . after a portion of the spare paper web roll 1 has been used , another entry to the roll data takes place by means of the recording / reading head 29 , as well as a new destination , for example a storage facility . it is also possible to arrange only stationary reading devices for the purpose of reading out the destination at appropriate positions on the transport path , for example at switches . in accordance with a further preferred embodiment , the recording / reading head 29 can be arranged in the roll changer , i . e . for example in a roll cone of a support arm of the roll changer , so that a data exchange can take place between the recording / reading head 29 and a code carrier 18 while the spare paper web roll 1 is still on the shaft . alternatively it is also possible to fasten the recording / reading head 29 on the end of the support arm of the roll changer . the data exchange preferably takes place during the run - out of the spare paper web roll 1 , i . e . while the spare paper web roll 1 turns slowly . it is also possible to utilize a code carrier 12 arranged in the area of the imagined axis of rotation 2 of the spare paper web roll 1 in the interior 6 of the tube 3 . in this case the recording / reading head 12 is arranged in the direction of an extended axis of rotation 2 on the exterior of a support arm of the roll changer . here , the clamping cone is designed to be hollow . while preferred embodiments of a spare paper roll in accordance with the present invention have been set forth fully and completely hereinabove , it will be apparent to one of skill in the art that a number of changes in , for example , the material used to construct the tube , the type of press with which the spare paper roll will be used , and the like may be made without departing from the true spirit and scope of the present invention which is accordingly to be limited only by the following claims ."}
|
Is the patent correctly categorized?
| 0.25 |
80dcd2ffc37235c3e85318e8e521b30e8f49d20e8764e0e991f9e17984ba79c6
| 0.233398 | 0.072754 | 0.507813 | 0.087402 | 0.445313 | 0.263672 |
null |
{"category": "Performing Operations; Transporting", "patent": "a spare web roll , for example a spare paper web roll 1 for a web - fed rotary printing press , has a continuous tube 3 in the area of its axis of rotation 2 , on whose tube body a web 4 of material has been wound , all as seen in fig1 and 2 . a star - shaped , for example three - armed holder 7 is arranged in the end area of the tube , and in particular in the area of the interior diameter , or respectively interior 6 of the tube 3 , whose arms 8 , 9 , 10 rest resiliently against the interior circumference 6 . for example , the holder 7 supports a cylinder - shaped or disk - shaped code carrier 12 , which can be activated , i . e . can be coded and decoded , coaxially with the axis of rotation 2 . the code carrier 12 is an electronic component , which can be coded by inductive means . the holder 7 can be inserted , for example into the end area of the tube 3 of a spare paper web 1 roll , i . e . into the interior 6 of the tube 3 , in such a way that it remains in a temporary first position b , i . e . close to the lateral edges 13 of the web of material 4 as \u2014 represented in dashed lines in fig1 . only when the spare paper web 1 roll is placed on the shaft , is the holder 7 pushed into a second position c by means of a clamping cone , not represented . the holder 7 will now be in the position shown in solid lines in fig1 . after the roll is used up , for example , the holder 7 can again be taken out of the tube 3 by removing it . for this purpose the arms 8 , 9 , 10 can have bores 14 , 16 , 17 , for example , behind which a suitable tool can reach . a number of code carriers 18 , 21 , 23 , 26 are represented in respectively different fastening positions in fig3 . however , only one of which is used . a code carrier 18 is arranged in the front face 19 of the tube 3 , i . e . in the tube body , and may for example be pressed into a bore . a code carrier 21 is fastened to the inner surface 22 of the tube 3 , i . e . in the interior 6 of the tube 3 , for example by gluing . a code carrier 23 is arranged in the tube body 3 . in this case , the tube wall 24 can have a bore into which the code carrier 23 is pressed . it is also possible to arrange a code carrier 26 in the webs of material 4 which are close to the tube 3 , i . e . the inner or lower ones , i . e . to introduce it from the direction of the lateral edges 13 of the webs of material 4 into an area close to the edge , for example to press it in . therefore the code carrier 26 is outside of and in the vicinity of the tube 3 . the code carriers 12 , 18 , 21 , 23 , 26 arranged in the interior of the tube 3 or in the vicinity of the tube 3 , for example the code carrier 18 located in the front face 19 of the tube 3 , are connected with one of the recording and reading heads 27 , 28 , 29 . such recording and reading heads 27 , 28 , 29 are , as shown in fig4 stationed along a conveyance path for movable spare paper web transport carriages , for example the recording and reading head 27 in the vicinity of a storage facility for spare paper web rolls 1 , the recording and reading head 28 is locked in the vicinity of a preparation station for spare paper web , and the recording and reading head 29 is located in the vicinity of a roll changer . each recording and reading head 27 to 29 consists of an electronic component , which transmits energy 37 and information 38 in a contactless manner to the code carrier 18 . on the other end , all recording and reading heads 27 to 29 are connected with each other via a data bus 39 , as well as with an evaluation unit 41 and a control device 42 , as well as a master computer 43 , all as shown in fig4 . the master computer 43 has an input station 46 . the code carrier 18 , as well as the other code carriers 12 , 21 , 23 , 26 , are essentially embodied as data memory devices , and respectively consist of an electronic component with a memory and logical control device . the above mentioned code carriers are , for example , designed as eeprom versions , and can be written on and read out , for example , with up to eight kilobytes . if , for example , a spare paper web roll 1 is moved out of the storage facility \u2014 the transport carriages are , for example , pulled by means of under - floor or driverless transport systems \u2014 the code carrier 18 located in the tube 3 of the spare paper web roll 1 passes by the recording / reading head 27 . as soon as the code carrier 18 comes into the active area of the recording / reading head 27 , the energy 37 required for data transmission is built up and the data 38 are transmitted by the recording / reading head 27 to the code carrier 18 , and the data 44 are transmitted by the code carrier 18 to the recording / reading head 27 , each in a frequency range of , for example , 70 kilohertz . such data can be , besides the data entered by the manufacturer : new destination , for example the spare paper web preparation station , remaining length of paper , particularities of the paper , for example a \u201c beating \u201d roll . the data , which are inductively coupled in by the code carrier 18 , are converted into a digital energy signal and conducted to the evaluation unit 41 . the evaluation unit 41 manages the data transfer 37 , 44 between the recording / reading head 28 and the code carrier 18 and is used as an intermediate memory . the evaluation unit 41 is the connecting member between the master computer 43 , or respectively the control device 42 , and the code carrier 18 . data can be erased or new data can be added , such as for example , a new destination , the gluing preparation station and the type of the glue tip . following the gluing preparation of the spare paper web roll 1 , it is possible to enter a further destination , for example a roll changer , by means of the recording / reading head 28 . after a portion of the spare paper web roll 1 has been used , another entry to the roll data takes place by means of the recording / reading head 29 , as well as a new destination , for example a storage facility . it is also possible to arrange only stationary reading devices for the purpose of reading out the destination at appropriate positions on the transport path , for example at switches . in accordance with a further preferred embodiment , the recording / reading head 29 can be arranged in the roll changer , i . e . for example in a roll cone of a support arm of the roll changer , so that a data exchange can take place between the recording / reading head 29 and a code carrier 18 while the spare paper web roll 1 is still on the shaft . alternatively it is also possible to fasten the recording / reading head 29 on the end of the support arm of the roll changer . the data exchange preferably takes place during the run - out of the spare paper web roll 1 , i . e . while the spare paper web roll 1 turns slowly . it is also possible to utilize a code carrier 12 arranged in the area of the imagined axis of rotation 2 of the spare paper web roll 1 in the interior 6 of the tube 3 . in this case the recording / reading head 12 is arranged in the direction of an extended axis of rotation 2 on the exterior of a support arm of the roll changer . here , the clamping cone is designed to be hollow . while preferred embodiments of a spare paper roll in accordance with the present invention have been set forth fully and completely hereinabove , it will be apparent to one of skill in the art that a number of changes in , for example , the material used to construct the tube , the type of press with which the spare paper roll will be used , and the like may be made without departing from the true spirit and scope of the present invention which is accordingly to be limited only by the following claims ."}
|
{"patent": "a spare web roll , for example a spare paper web roll 1 for a web - fed rotary printing press , has a continuous tube 3 in the area of its axis of rotation 2 , on whose tube body a web 4 of material has been wound , all as seen in fig1 and 2 . a star - shaped , for example three - armed holder 7 is arranged in the end area of the tube , and in particular in the area of the interior diameter , or respectively interior 6 of the tube 3 , whose arms 8 , 9 , 10 rest resiliently against the interior circumference 6 . for example , the holder 7 supports a cylinder - shaped or disk - shaped code carrier 12 , which can be activated , i . e . can be coded and decoded , coaxially with the axis of rotation 2 . the code carrier 12 is an electronic component , which can be coded by inductive means . the holder 7 can be inserted , for example into the end area of the tube 3 of a spare paper web 1 roll , i . e . into the interior 6 of the tube 3 , in such a way that it remains in a temporary first position b , i . e . close to the lateral edges 13 of the web of material 4 as \u2014 represented in dashed lines in fig1 . only when the spare paper web 1 roll is placed on the shaft , is the holder 7 pushed into a second position c by means of a clamping cone , not represented . the holder 7 will now be in the position shown in solid lines in fig1 . after the roll is used up , for example , the holder 7 can again be taken out of the tube 3 by removing it . for this purpose the arms 8 , 9 , 10 can have bores 14 , 16 , 17 , for example , behind which a suitable tool can reach . a number of code carriers 18 , 21 , 23 , 26 are represented in respectively different fastening positions in fig3 . however , only one of which is used . a code carrier 18 is arranged in the front face 19 of the tube 3 , i . e . in the tube body , and may for example be pressed into a bore . a code carrier 21 is fastened to the inner surface 22 of the tube 3 , i . e . in the interior 6 of the tube 3 , for example by gluing . a code carrier 23 is arranged in the tube body 3 . in this case , the tube wall 24 can have a bore into which the code carrier 23 is pressed . it is also possible to arrange a code carrier 26 in the webs of material 4 which are close to the tube 3 , i . e . the inner or lower ones , i . e . to introduce it from the direction of the lateral edges 13 of the webs of material 4 into an area close to the edge , for example to press it in . therefore the code carrier 26 is outside of and in the vicinity of the tube 3 . the code carriers 12 , 18 , 21 , 23 , 26 arranged in the interior of the tube 3 or in the vicinity of the tube 3 , for example the code carrier 18 located in the front face 19 of the tube 3 , are connected with one of the recording and reading heads 27 , 28 , 29 . such recording and reading heads 27 , 28 , 29 are , as shown in fig4 stationed along a conveyance path for movable spare paper web transport carriages , for example the recording and reading head 27 in the vicinity of a storage facility for spare paper web rolls 1 , the recording and reading head 28 is locked in the vicinity of a preparation station for spare paper web , and the recording and reading head 29 is located in the vicinity of a roll changer . each recording and reading head 27 to 29 consists of an electronic component , which transmits energy 37 and information 38 in a contactless manner to the code carrier 18 . on the other end , all recording and reading heads 27 to 29 are connected with each other via a data bus 39 , as well as with an evaluation unit 41 and a control device 42 , as well as a master computer 43 , all as shown in fig4 . the master computer 43 has an input station 46 . the code carrier 18 , as well as the other code carriers 12 , 21 , 23 , 26 , are essentially embodied as data memory devices , and respectively consist of an electronic component with a memory and logical control device . the above mentioned code carriers are , for example , designed as eeprom versions , and can be written on and read out , for example , with up to eight kilobytes . if , for example , a spare paper web roll 1 is moved out of the storage facility \u2014 the transport carriages are , for example , pulled by means of under - floor or driverless transport systems \u2014 the code carrier 18 located in the tube 3 of the spare paper web roll 1 passes by the recording / reading head 27 . as soon as the code carrier 18 comes into the active area of the recording / reading head 27 , the energy 37 required for data transmission is built up and the data 38 are transmitted by the recording / reading head 27 to the code carrier 18 , and the data 44 are transmitted by the code carrier 18 to the recording / reading head 27 , each in a frequency range of , for example , 70 kilohertz . such data can be , besides the data entered by the manufacturer : new destination , for example the spare paper web preparation station , remaining length of paper , particularities of the paper , for example a \u201c beating \u201d roll . the data , which are inductively coupled in by the code carrier 18 , are converted into a digital energy signal and conducted to the evaluation unit 41 . the evaluation unit 41 manages the data transfer 37 , 44 between the recording / reading head 28 and the code carrier 18 and is used as an intermediate memory . the evaluation unit 41 is the connecting member between the master computer 43 , or respectively the control device 42 , and the code carrier 18 . data can be erased or new data can be added , such as for example , a new destination , the gluing preparation station and the type of the glue tip . following the gluing preparation of the spare paper web roll 1 , it is possible to enter a further destination , for example a roll changer , by means of the recording / reading head 28 . after a portion of the spare paper web roll 1 has been used , another entry to the roll data takes place by means of the recording / reading head 29 , as well as a new destination , for example a storage facility . it is also possible to arrange only stationary reading devices for the purpose of reading out the destination at appropriate positions on the transport path , for example at switches . in accordance with a further preferred embodiment , the recording / reading head 29 can be arranged in the roll changer , i . e . for example in a roll cone of a support arm of the roll changer , so that a data exchange can take place between the recording / reading head 29 and a code carrier 18 while the spare paper web roll 1 is still on the shaft . alternatively it is also possible to fasten the recording / reading head 29 on the end of the support arm of the roll changer . the data exchange preferably takes place during the run - out of the spare paper web roll 1 , i . e . while the spare paper web roll 1 turns slowly . it is also possible to utilize a code carrier 12 arranged in the area of the imagined axis of rotation 2 of the spare paper web roll 1 in the interior 6 of the tube 3 . in this case the recording / reading head 12 is arranged in the direction of an extended axis of rotation 2 on the exterior of a support arm of the roll changer . here , the clamping cone is designed to be hollow . while preferred embodiments of a spare paper roll in accordance with the present invention have been set forth fully and completely hereinabove , it will be apparent to one of skill in the art that a number of changes in , for example , the material used to construct the tube , the type of press with which the spare paper roll will be used , and the like may be made without departing from the true spirit and scope of the present invention which is accordingly to be limited only by the following claims .", "category": "Chemistry; Metallurgy"}
|
Is the patent correctly categorized?
| 0.25 |
80dcd2ffc37235c3e85318e8e521b30e8f49d20e8764e0e991f9e17984ba79c6
| 0.792969 | 0.000368 | 0.679688 | 0.010986 | 0.941406 | 0.006104 |
null |
{"patent": "a spare web roll , for example a spare paper web roll 1 for a web - fed rotary printing press , has a continuous tube 3 in the area of its axis of rotation 2 , on whose tube body a web 4 of material has been wound , all as seen in fig1 and 2 . a star - shaped , for example three - armed holder 7 is arranged in the end area of the tube , and in particular in the area of the interior diameter , or respectively interior 6 of the tube 3 , whose arms 8 , 9 , 10 rest resiliently against the interior circumference 6 . for example , the holder 7 supports a cylinder - shaped or disk - shaped code carrier 12 , which can be activated , i . e . can be coded and decoded , coaxially with the axis of rotation 2 . the code carrier 12 is an electronic component , which can be coded by inductive means . the holder 7 can be inserted , for example into the end area of the tube 3 of a spare paper web 1 roll , i . e . into the interior 6 of the tube 3 , in such a way that it remains in a temporary first position b , i . e . close to the lateral edges 13 of the web of material 4 as \u2014 represented in dashed lines in fig1 . only when the spare paper web 1 roll is placed on the shaft , is the holder 7 pushed into a second position c by means of a clamping cone , not represented . the holder 7 will now be in the position shown in solid lines in fig1 . after the roll is used up , for example , the holder 7 can again be taken out of the tube 3 by removing it . for this purpose the arms 8 , 9 , 10 can have bores 14 , 16 , 17 , for example , behind which a suitable tool can reach . a number of code carriers 18 , 21 , 23 , 26 are represented in respectively different fastening positions in fig3 . however , only one of which is used . a code carrier 18 is arranged in the front face 19 of the tube 3 , i . e . in the tube body , and may for example be pressed into a bore . a code carrier 21 is fastened to the inner surface 22 of the tube 3 , i . e . in the interior 6 of the tube 3 , for example by gluing . a code carrier 23 is arranged in the tube body 3 . in this case , the tube wall 24 can have a bore into which the code carrier 23 is pressed . it is also possible to arrange a code carrier 26 in the webs of material 4 which are close to the tube 3 , i . e . the inner or lower ones , i . e . to introduce it from the direction of the lateral edges 13 of the webs of material 4 into an area close to the edge , for example to press it in . therefore the code carrier 26 is outside of and in the vicinity of the tube 3 . the code carriers 12 , 18 , 21 , 23 , 26 arranged in the interior of the tube 3 or in the vicinity of the tube 3 , for example the code carrier 18 located in the front face 19 of the tube 3 , are connected with one of the recording and reading heads 27 , 28 , 29 . such recording and reading heads 27 , 28 , 29 are , as shown in fig4 stationed along a conveyance path for movable spare paper web transport carriages , for example the recording and reading head 27 in the vicinity of a storage facility for spare paper web rolls 1 , the recording and reading head 28 is locked in the vicinity of a preparation station for spare paper web , and the recording and reading head 29 is located in the vicinity of a roll changer . each recording and reading head 27 to 29 consists of an electronic component , which transmits energy 37 and information 38 in a contactless manner to the code carrier 18 . on the other end , all recording and reading heads 27 to 29 are connected with each other via a data bus 39 , as well as with an evaluation unit 41 and a control device 42 , as well as a master computer 43 , all as shown in fig4 . the master computer 43 has an input station 46 . the code carrier 18 , as well as the other code carriers 12 , 21 , 23 , 26 , are essentially embodied as data memory devices , and respectively consist of an electronic component with a memory and logical control device . the above mentioned code carriers are , for example , designed as eeprom versions , and can be written on and read out , for example , with up to eight kilobytes . if , for example , a spare paper web roll 1 is moved out of the storage facility \u2014 the transport carriages are , for example , pulled by means of under - floor or driverless transport systems \u2014 the code carrier 18 located in the tube 3 of the spare paper web roll 1 passes by the recording / reading head 27 . as soon as the code carrier 18 comes into the active area of the recording / reading head 27 , the energy 37 required for data transmission is built up and the data 38 are transmitted by the recording / reading head 27 to the code carrier 18 , and the data 44 are transmitted by the code carrier 18 to the recording / reading head 27 , each in a frequency range of , for example , 70 kilohertz . such data can be , besides the data entered by the manufacturer : new destination , for example the spare paper web preparation station , remaining length of paper , particularities of the paper , for example a \u201c beating \u201d roll . the data , which are inductively coupled in by the code carrier 18 , are converted into a digital energy signal and conducted to the evaluation unit 41 . the evaluation unit 41 manages the data transfer 37 , 44 between the recording / reading head 28 and the code carrier 18 and is used as an intermediate memory . the evaluation unit 41 is the connecting member between the master computer 43 , or respectively the control device 42 , and the code carrier 18 . data can be erased or new data can be added , such as for example , a new destination , the gluing preparation station and the type of the glue tip . following the gluing preparation of the spare paper web roll 1 , it is possible to enter a further destination , for example a roll changer , by means of the recording / reading head 28 . after a portion of the spare paper web roll 1 has been used , another entry to the roll data takes place by means of the recording / reading head 29 , as well as a new destination , for example a storage facility . it is also possible to arrange only stationary reading devices for the purpose of reading out the destination at appropriate positions on the transport path , for example at switches . in accordance with a further preferred embodiment , the recording / reading head 29 can be arranged in the roll changer , i . e . for example in a roll cone of a support arm of the roll changer , so that a data exchange can take place between the recording / reading head 29 and a code carrier 18 while the spare paper web roll 1 is still on the shaft . alternatively it is also possible to fasten the recording / reading head 29 on the end of the support arm of the roll changer . the data exchange preferably takes place during the run - out of the spare paper web roll 1 , i . e . while the spare paper web roll 1 turns slowly . it is also possible to utilize a code carrier 12 arranged in the area of the imagined axis of rotation 2 of the spare paper web roll 1 in the interior 6 of the tube 3 . in this case the recording / reading head 12 is arranged in the direction of an extended axis of rotation 2 on the exterior of a support arm of the roll changer . here , the clamping cone is designed to be hollow . while preferred embodiments of a spare paper roll in accordance with the present invention have been set forth fully and completely hereinabove , it will be apparent to one of skill in the art that a number of changes in , for example , the material used to construct the tube , the type of press with which the spare paper roll will be used , and the like may be made without departing from the true spirit and scope of the present invention which is accordingly to be limited only by the following claims .", "category": "Performing Operations; Transporting"}
|
{"patent": "a spare web roll , for example a spare paper web roll 1 for a web - fed rotary printing press , has a continuous tube 3 in the area of its axis of rotation 2 , on whose tube body a web 4 of material has been wound , all as seen in fig1 and 2 . a star - shaped , for example three - armed holder 7 is arranged in the end area of the tube , and in particular in the area of the interior diameter , or respectively interior 6 of the tube 3 , whose arms 8 , 9 , 10 rest resiliently against the interior circumference 6 . for example , the holder 7 supports a cylinder - shaped or disk - shaped code carrier 12 , which can be activated , i . e . can be coded and decoded , coaxially with the axis of rotation 2 . the code carrier 12 is an electronic component , which can be coded by inductive means . the holder 7 can be inserted , for example into the end area of the tube 3 of a spare paper web 1 roll , i . e . into the interior 6 of the tube 3 , in such a way that it remains in a temporary first position b , i . e . close to the lateral edges 13 of the web of material 4 as \u2014 represented in dashed lines in fig1 . only when the spare paper web 1 roll is placed on the shaft , is the holder 7 pushed into a second position c by means of a clamping cone , not represented . the holder 7 will now be in the position shown in solid lines in fig1 . after the roll is used up , for example , the holder 7 can again be taken out of the tube 3 by removing it . for this purpose the arms 8 , 9 , 10 can have bores 14 , 16 , 17 , for example , behind which a suitable tool can reach . a number of code carriers 18 , 21 , 23 , 26 are represented in respectively different fastening positions in fig3 . however , only one of which is used . a code carrier 18 is arranged in the front face 19 of the tube 3 , i . e . in the tube body , and may for example be pressed into a bore . a code carrier 21 is fastened to the inner surface 22 of the tube 3 , i . e . in the interior 6 of the tube 3 , for example by gluing . a code carrier 23 is arranged in the tube body 3 . in this case , the tube wall 24 can have a bore into which the code carrier 23 is pressed . it is also possible to arrange a code carrier 26 in the webs of material 4 which are close to the tube 3 , i . e . the inner or lower ones , i . e . to introduce it from the direction of the lateral edges 13 of the webs of material 4 into an area close to the edge , for example to press it in . therefore the code carrier 26 is outside of and in the vicinity of the tube 3 . the code carriers 12 , 18 , 21 , 23 , 26 arranged in the interior of the tube 3 or in the vicinity of the tube 3 , for example the code carrier 18 located in the front face 19 of the tube 3 , are connected with one of the recording and reading heads 27 , 28 , 29 . such recording and reading heads 27 , 28 , 29 are , as shown in fig4 stationed along a conveyance path for movable spare paper web transport carriages , for example the recording and reading head 27 in the vicinity of a storage facility for spare paper web rolls 1 , the recording and reading head 28 is locked in the vicinity of a preparation station for spare paper web , and the recording and reading head 29 is located in the vicinity of a roll changer . each recording and reading head 27 to 29 consists of an electronic component , which transmits energy 37 and information 38 in a contactless manner to the code carrier 18 . on the other end , all recording and reading heads 27 to 29 are connected with each other via a data bus 39 , as well as with an evaluation unit 41 and a control device 42 , as well as a master computer 43 , all as shown in fig4 . the master computer 43 has an input station 46 . the code carrier 18 , as well as the other code carriers 12 , 21 , 23 , 26 , are essentially embodied as data memory devices , and respectively consist of an electronic component with a memory and logical control device . the above mentioned code carriers are , for example , designed as eeprom versions , and can be written on and read out , for example , with up to eight kilobytes . if , for example , a spare paper web roll 1 is moved out of the storage facility \u2014 the transport carriages are , for example , pulled by means of under - floor or driverless transport systems \u2014 the code carrier 18 located in the tube 3 of the spare paper web roll 1 passes by the recording / reading head 27 . as soon as the code carrier 18 comes into the active area of the recording / reading head 27 , the energy 37 required for data transmission is built up and the data 38 are transmitted by the recording / reading head 27 to the code carrier 18 , and the data 44 are transmitted by the code carrier 18 to the recording / reading head 27 , each in a frequency range of , for example , 70 kilohertz . such data can be , besides the data entered by the manufacturer : new destination , for example the spare paper web preparation station , remaining length of paper , particularities of the paper , for example a \u201c beating \u201d roll . the data , which are inductively coupled in by the code carrier 18 , are converted into a digital energy signal and conducted to the evaluation unit 41 . the evaluation unit 41 manages the data transfer 37 , 44 between the recording / reading head 28 and the code carrier 18 and is used as an intermediate memory . the evaluation unit 41 is the connecting member between the master computer 43 , or respectively the control device 42 , and the code carrier 18 . data can be erased or new data can be added , such as for example , a new destination , the gluing preparation station and the type of the glue tip . following the gluing preparation of the spare paper web roll 1 , it is possible to enter a further destination , for example a roll changer , by means of the recording / reading head 28 . after a portion of the spare paper web roll 1 has been used , another entry to the roll data takes place by means of the recording / reading head 29 , as well as a new destination , for example a storage facility . it is also possible to arrange only stationary reading devices for the purpose of reading out the destination at appropriate positions on the transport path , for example at switches . in accordance with a further preferred embodiment , the recording / reading head 29 can be arranged in the roll changer , i . e . for example in a roll cone of a support arm of the roll changer , so that a data exchange can take place between the recording / reading head 29 and a code carrier 18 while the spare paper web roll 1 is still on the shaft . alternatively it is also possible to fasten the recording / reading head 29 on the end of the support arm of the roll changer . the data exchange preferably takes place during the run - out of the spare paper web roll 1 , i . e . while the spare paper web roll 1 turns slowly . it is also possible to utilize a code carrier 12 arranged in the area of the imagined axis of rotation 2 of the spare paper web roll 1 in the interior 6 of the tube 3 . in this case the recording / reading head 12 is arranged in the direction of an extended axis of rotation 2 on the exterior of a support arm of the roll changer . here , the clamping cone is designed to be hollow . while preferred embodiments of a spare paper roll in accordance with the present invention have been set forth fully and completely hereinabove , it will be apparent to one of skill in the art that a number of changes in , for example , the material used to construct the tube , the type of press with which the spare paper roll will be used , and the like may be made without departing from the true spirit and scope of the present invention which is accordingly to be limited only by the following claims .", "category": "Textiles; Paper"}
|
Does the patent belong in this category?
| 0.25 |
80dcd2ffc37235c3e85318e8e521b30e8f49d20e8764e0e991f9e17984ba79c6
| 0.460938 | 0.246094 | 0.679688 | 0.067383 | 0.597656 | 0.273438 |
null |
{"patent": "a spare web roll , for example a spare paper web roll 1 for a web - fed rotary printing press , has a continuous tube 3 in the area of its axis of rotation 2 , on whose tube body a web 4 of material has been wound , all as seen in fig1 and 2 . a star - shaped , for example three - armed holder 7 is arranged in the end area of the tube , and in particular in the area of the interior diameter , or respectively interior 6 of the tube 3 , whose arms 8 , 9 , 10 rest resiliently against the interior circumference 6 . for example , the holder 7 supports a cylinder - shaped or disk - shaped code carrier 12 , which can be activated , i . e . can be coded and decoded , coaxially with the axis of rotation 2 . the code carrier 12 is an electronic component , which can be coded by inductive means . the holder 7 can be inserted , for example into the end area of the tube 3 of a spare paper web 1 roll , i . e . into the interior 6 of the tube 3 , in such a way that it remains in a temporary first position b , i . e . close to the lateral edges 13 of the web of material 4 as \u2014 represented in dashed lines in fig1 . only when the spare paper web 1 roll is placed on the shaft , is the holder 7 pushed into a second position c by means of a clamping cone , not represented . the holder 7 will now be in the position shown in solid lines in fig1 . after the roll is used up , for example , the holder 7 can again be taken out of the tube 3 by removing it . for this purpose the arms 8 , 9 , 10 can have bores 14 , 16 , 17 , for example , behind which a suitable tool can reach . a number of code carriers 18 , 21 , 23 , 26 are represented in respectively different fastening positions in fig3 . however , only one of which is used . a code carrier 18 is arranged in the front face 19 of the tube 3 , i . e . in the tube body , and may for example be pressed into a bore . a code carrier 21 is fastened to the inner surface 22 of the tube 3 , i . e . in the interior 6 of the tube 3 , for example by gluing . a code carrier 23 is arranged in the tube body 3 . in this case , the tube wall 24 can have a bore into which the code carrier 23 is pressed . it is also possible to arrange a code carrier 26 in the webs of material 4 which are close to the tube 3 , i . e . the inner or lower ones , i . e . to introduce it from the direction of the lateral edges 13 of the webs of material 4 into an area close to the edge , for example to press it in . therefore the code carrier 26 is outside of and in the vicinity of the tube 3 . the code carriers 12 , 18 , 21 , 23 , 26 arranged in the interior of the tube 3 or in the vicinity of the tube 3 , for example the code carrier 18 located in the front face 19 of the tube 3 , are connected with one of the recording and reading heads 27 , 28 , 29 . such recording and reading heads 27 , 28 , 29 are , as shown in fig4 stationed along a conveyance path for movable spare paper web transport carriages , for example the recording and reading head 27 in the vicinity of a storage facility for spare paper web rolls 1 , the recording and reading head 28 is locked in the vicinity of a preparation station for spare paper web , and the recording and reading head 29 is located in the vicinity of a roll changer . each recording and reading head 27 to 29 consists of an electronic component , which transmits energy 37 and information 38 in a contactless manner to the code carrier 18 . on the other end , all recording and reading heads 27 to 29 are connected with each other via a data bus 39 , as well as with an evaluation unit 41 and a control device 42 , as well as a master computer 43 , all as shown in fig4 . the master computer 43 has an input station 46 . the code carrier 18 , as well as the other code carriers 12 , 21 , 23 , 26 , are essentially embodied as data memory devices , and respectively consist of an electronic component with a memory and logical control device . the above mentioned code carriers are , for example , designed as eeprom versions , and can be written on and read out , for example , with up to eight kilobytes . if , for example , a spare paper web roll 1 is moved out of the storage facility \u2014 the transport carriages are , for example , pulled by means of under - floor or driverless transport systems \u2014 the code carrier 18 located in the tube 3 of the spare paper web roll 1 passes by the recording / reading head 27 . as soon as the code carrier 18 comes into the active area of the recording / reading head 27 , the energy 37 required for data transmission is built up and the data 38 are transmitted by the recording / reading head 27 to the code carrier 18 , and the data 44 are transmitted by the code carrier 18 to the recording / reading head 27 , each in a frequency range of , for example , 70 kilohertz . such data can be , besides the data entered by the manufacturer : new destination , for example the spare paper web preparation station , remaining length of paper , particularities of the paper , for example a \u201c beating \u201d roll . the data , which are inductively coupled in by the code carrier 18 , are converted into a digital energy signal and conducted to the evaluation unit 41 . the evaluation unit 41 manages the data transfer 37 , 44 between the recording / reading head 28 and the code carrier 18 and is used as an intermediate memory . the evaluation unit 41 is the connecting member between the master computer 43 , or respectively the control device 42 , and the code carrier 18 . data can be erased or new data can be added , such as for example , a new destination , the gluing preparation station and the type of the glue tip . following the gluing preparation of the spare paper web roll 1 , it is possible to enter a further destination , for example a roll changer , by means of the recording / reading head 28 . after a portion of the spare paper web roll 1 has been used , another entry to the roll data takes place by means of the recording / reading head 29 , as well as a new destination , for example a storage facility . it is also possible to arrange only stationary reading devices for the purpose of reading out the destination at appropriate positions on the transport path , for example at switches . in accordance with a further preferred embodiment , the recording / reading head 29 can be arranged in the roll changer , i . e . for example in a roll cone of a support arm of the roll changer , so that a data exchange can take place between the recording / reading head 29 and a code carrier 18 while the spare paper web roll 1 is still on the shaft . alternatively it is also possible to fasten the recording / reading head 29 on the end of the support arm of the roll changer . the data exchange preferably takes place during the run - out of the spare paper web roll 1 , i . e . while the spare paper web roll 1 turns slowly . it is also possible to utilize a code carrier 12 arranged in the area of the imagined axis of rotation 2 of the spare paper web roll 1 in the interior 6 of the tube 3 . in this case the recording / reading head 12 is arranged in the direction of an extended axis of rotation 2 on the exterior of a support arm of the roll changer . here , the clamping cone is designed to be hollow . while preferred embodiments of a spare paper roll in accordance with the present invention have been set forth fully and completely hereinabove , it will be apparent to one of skill in the art that a number of changes in , for example , the material used to construct the tube , the type of press with which the spare paper roll will be used , and the like may be made without departing from the true spirit and scope of the present invention which is accordingly to be limited only by the following claims .", "category": "Performing Operations; Transporting"}
|
{"category": "Fixed Constructions", "patent": "a spare web roll , for example a spare paper web roll 1 for a web - fed rotary printing press , has a continuous tube 3 in the area of its axis of rotation 2 , on whose tube body a web 4 of material has been wound , all as seen in fig1 and 2 . a star - shaped , for example three - armed holder 7 is arranged in the end area of the tube , and in particular in the area of the interior diameter , or respectively interior 6 of the tube 3 , whose arms 8 , 9 , 10 rest resiliently against the interior circumference 6 . for example , the holder 7 supports a cylinder - shaped or disk - shaped code carrier 12 , which can be activated , i . e . can be coded and decoded , coaxially with the axis of rotation 2 . the code carrier 12 is an electronic component , which can be coded by inductive means . the holder 7 can be inserted , for example into the end area of the tube 3 of a spare paper web 1 roll , i . e . into the interior 6 of the tube 3 , in such a way that it remains in a temporary first position b , i . e . close to the lateral edges 13 of the web of material 4 as \u2014 represented in dashed lines in fig1 . only when the spare paper web 1 roll is placed on the shaft , is the holder 7 pushed into a second position c by means of a clamping cone , not represented . the holder 7 will now be in the position shown in solid lines in fig1 . after the roll is used up , for example , the holder 7 can again be taken out of the tube 3 by removing it . for this purpose the arms 8 , 9 , 10 can have bores 14 , 16 , 17 , for example , behind which a suitable tool can reach . a number of code carriers 18 , 21 , 23 , 26 are represented in respectively different fastening positions in fig3 . however , only one of which is used . a code carrier 18 is arranged in the front face 19 of the tube 3 , i . e . in the tube body , and may for example be pressed into a bore . a code carrier 21 is fastened to the inner surface 22 of the tube 3 , i . e . in the interior 6 of the tube 3 , for example by gluing . a code carrier 23 is arranged in the tube body 3 . in this case , the tube wall 24 can have a bore into which the code carrier 23 is pressed . it is also possible to arrange a code carrier 26 in the webs of material 4 which are close to the tube 3 , i . e . the inner or lower ones , i . e . to introduce it from the direction of the lateral edges 13 of the webs of material 4 into an area close to the edge , for example to press it in . therefore the code carrier 26 is outside of and in the vicinity of the tube 3 . the code carriers 12 , 18 , 21 , 23 , 26 arranged in the interior of the tube 3 or in the vicinity of the tube 3 , for example the code carrier 18 located in the front face 19 of the tube 3 , are connected with one of the recording and reading heads 27 , 28 , 29 . such recording and reading heads 27 , 28 , 29 are , as shown in fig4 stationed along a conveyance path for movable spare paper web transport carriages , for example the recording and reading head 27 in the vicinity of a storage facility for spare paper web rolls 1 , the recording and reading head 28 is locked in the vicinity of a preparation station for spare paper web , and the recording and reading head 29 is located in the vicinity of a roll changer . each recording and reading head 27 to 29 consists of an electronic component , which transmits energy 37 and information 38 in a contactless manner to the code carrier 18 . on the other end , all recording and reading heads 27 to 29 are connected with each other via a data bus 39 , as well as with an evaluation unit 41 and a control device 42 , as well as a master computer 43 , all as shown in fig4 . the master computer 43 has an input station 46 . the code carrier 18 , as well as the other code carriers 12 , 21 , 23 , 26 , are essentially embodied as data memory devices , and respectively consist of an electronic component with a memory and logical control device . the above mentioned code carriers are , for example , designed as eeprom versions , and can be written on and read out , for example , with up to eight kilobytes . if , for example , a spare paper web roll 1 is moved out of the storage facility \u2014 the transport carriages are , for example , pulled by means of under - floor or driverless transport systems \u2014 the code carrier 18 located in the tube 3 of the spare paper web roll 1 passes by the recording / reading head 27 . as soon as the code carrier 18 comes into the active area of the recording / reading head 27 , the energy 37 required for data transmission is built up and the data 38 are transmitted by the recording / reading head 27 to the code carrier 18 , and the data 44 are transmitted by the code carrier 18 to the recording / reading head 27 , each in a frequency range of , for example , 70 kilohertz . such data can be , besides the data entered by the manufacturer : new destination , for example the spare paper web preparation station , remaining length of paper , particularities of the paper , for example a \u201c beating \u201d roll . the data , which are inductively coupled in by the code carrier 18 , are converted into a digital energy signal and conducted to the evaluation unit 41 . the evaluation unit 41 manages the data transfer 37 , 44 between the recording / reading head 28 and the code carrier 18 and is used as an intermediate memory . the evaluation unit 41 is the connecting member between the master computer 43 , or respectively the control device 42 , and the code carrier 18 . data can be erased or new data can be added , such as for example , a new destination , the gluing preparation station and the type of the glue tip . following the gluing preparation of the spare paper web roll 1 , it is possible to enter a further destination , for example a roll changer , by means of the recording / reading head 28 . after a portion of the spare paper web roll 1 has been used , another entry to the roll data takes place by means of the recording / reading head 29 , as well as a new destination , for example a storage facility . it is also possible to arrange only stationary reading devices for the purpose of reading out the destination at appropriate positions on the transport path , for example at switches . in accordance with a further preferred embodiment , the recording / reading head 29 can be arranged in the roll changer , i . e . for example in a roll cone of a support arm of the roll changer , so that a data exchange can take place between the recording / reading head 29 and a code carrier 18 while the spare paper web roll 1 is still on the shaft . alternatively it is also possible to fasten the recording / reading head 29 on the end of the support arm of the roll changer . the data exchange preferably takes place during the run - out of the spare paper web roll 1 , i . e . while the spare paper web roll 1 turns slowly . it is also possible to utilize a code carrier 12 arranged in the area of the imagined axis of rotation 2 of the spare paper web roll 1 in the interior 6 of the tube 3 . in this case the recording / reading head 12 is arranged in the direction of an extended axis of rotation 2 on the exterior of a support arm of the roll changer . here , the clamping cone is designed to be hollow . while preferred embodiments of a spare paper roll in accordance with the present invention have been set forth fully and completely hereinabove , it will be apparent to one of skill in the art that a number of changes in , for example , the material used to construct the tube , the type of press with which the spare paper roll will be used , and the like may be made without departing from the true spirit and scope of the present invention which is accordingly to be limited only by the following claims ."}
|
Does the category match the content of the patent?
| 0.25 |
80dcd2ffc37235c3e85318e8e521b30e8f49d20e8764e0e991f9e17984ba79c6
| 0.679688 | 0.206055 | 0.777344 | 0.785156 | 0.503906 | 0.496094 |
null |
{"patent": "a spare web roll , for example a spare paper web roll 1 for a web - fed rotary printing press , has a continuous tube 3 in the area of its axis of rotation 2 , on whose tube body a web 4 of material has been wound , all as seen in fig1 and 2 . a star - shaped , for example three - armed holder 7 is arranged in the end area of the tube , and in particular in the area of the interior diameter , or respectively interior 6 of the tube 3 , whose arms 8 , 9 , 10 rest resiliently against the interior circumference 6 . for example , the holder 7 supports a cylinder - shaped or disk - shaped code carrier 12 , which can be activated , i . e . can be coded and decoded , coaxially with the axis of rotation 2 . the code carrier 12 is an electronic component , which can be coded by inductive means . the holder 7 can be inserted , for example into the end area of the tube 3 of a spare paper web 1 roll , i . e . into the interior 6 of the tube 3 , in such a way that it remains in a temporary first position b , i . e . close to the lateral edges 13 of the web of material 4 as \u2014 represented in dashed lines in fig1 . only when the spare paper web 1 roll is placed on the shaft , is the holder 7 pushed into a second position c by means of a clamping cone , not represented . the holder 7 will now be in the position shown in solid lines in fig1 . after the roll is used up , for example , the holder 7 can again be taken out of the tube 3 by removing it . for this purpose the arms 8 , 9 , 10 can have bores 14 , 16 , 17 , for example , behind which a suitable tool can reach . a number of code carriers 18 , 21 , 23 , 26 are represented in respectively different fastening positions in fig3 . however , only one of which is used . a code carrier 18 is arranged in the front face 19 of the tube 3 , i . e . in the tube body , and may for example be pressed into a bore . a code carrier 21 is fastened to the inner surface 22 of the tube 3 , i . e . in the interior 6 of the tube 3 , for example by gluing . a code carrier 23 is arranged in the tube body 3 . in this case , the tube wall 24 can have a bore into which the code carrier 23 is pressed . it is also possible to arrange a code carrier 26 in the webs of material 4 which are close to the tube 3 , i . e . the inner or lower ones , i . e . to introduce it from the direction of the lateral edges 13 of the webs of material 4 into an area close to the edge , for example to press it in . therefore the code carrier 26 is outside of and in the vicinity of the tube 3 . the code carriers 12 , 18 , 21 , 23 , 26 arranged in the interior of the tube 3 or in the vicinity of the tube 3 , for example the code carrier 18 located in the front face 19 of the tube 3 , are connected with one of the recording and reading heads 27 , 28 , 29 . such recording and reading heads 27 , 28 , 29 are , as shown in fig4 stationed along a conveyance path for movable spare paper web transport carriages , for example the recording and reading head 27 in the vicinity of a storage facility for spare paper web rolls 1 , the recording and reading head 28 is locked in the vicinity of a preparation station for spare paper web , and the recording and reading head 29 is located in the vicinity of a roll changer . each recording and reading head 27 to 29 consists of an electronic component , which transmits energy 37 and information 38 in a contactless manner to the code carrier 18 . on the other end , all recording and reading heads 27 to 29 are connected with each other via a data bus 39 , as well as with an evaluation unit 41 and a control device 42 , as well as a master computer 43 , all as shown in fig4 . the master computer 43 has an input station 46 . the code carrier 18 , as well as the other code carriers 12 , 21 , 23 , 26 , are essentially embodied as data memory devices , and respectively consist of an electronic component with a memory and logical control device . the above mentioned code carriers are , for example , designed as eeprom versions , and can be written on and read out , for example , with up to eight kilobytes . if , for example , a spare paper web roll 1 is moved out of the storage facility \u2014 the transport carriages are , for example , pulled by means of under - floor or driverless transport systems \u2014 the code carrier 18 located in the tube 3 of the spare paper web roll 1 passes by the recording / reading head 27 . as soon as the code carrier 18 comes into the active area of the recording / reading head 27 , the energy 37 required for data transmission is built up and the data 38 are transmitted by the recording / reading head 27 to the code carrier 18 , and the data 44 are transmitted by the code carrier 18 to the recording / reading head 27 , each in a frequency range of , for example , 70 kilohertz . such data can be , besides the data entered by the manufacturer : new destination , for example the spare paper web preparation station , remaining length of paper , particularities of the paper , for example a \u201c beating \u201d roll . the data , which are inductively coupled in by the code carrier 18 , are converted into a digital energy signal and conducted to the evaluation unit 41 . the evaluation unit 41 manages the data transfer 37 , 44 between the recording / reading head 28 and the code carrier 18 and is used as an intermediate memory . the evaluation unit 41 is the connecting member between the master computer 43 , or respectively the control device 42 , and the code carrier 18 . data can be erased or new data can be added , such as for example , a new destination , the gluing preparation station and the type of the glue tip . following the gluing preparation of the spare paper web roll 1 , it is possible to enter a further destination , for example a roll changer , by means of the recording / reading head 28 . after a portion of the spare paper web roll 1 has been used , another entry to the roll data takes place by means of the recording / reading head 29 , as well as a new destination , for example a storage facility . it is also possible to arrange only stationary reading devices for the purpose of reading out the destination at appropriate positions on the transport path , for example at switches . in accordance with a further preferred embodiment , the recording / reading head 29 can be arranged in the roll changer , i . e . for example in a roll cone of a support arm of the roll changer , so that a data exchange can take place between the recording / reading head 29 and a code carrier 18 while the spare paper web roll 1 is still on the shaft . alternatively it is also possible to fasten the recording / reading head 29 on the end of the support arm of the roll changer . the data exchange preferably takes place during the run - out of the spare paper web roll 1 , i . e . while the spare paper web roll 1 turns slowly . it is also possible to utilize a code carrier 12 arranged in the area of the imagined axis of rotation 2 of the spare paper web roll 1 in the interior 6 of the tube 3 . in this case the recording / reading head 12 is arranged in the direction of an extended axis of rotation 2 on the exterior of a support arm of the roll changer . here , the clamping cone is designed to be hollow . while preferred embodiments of a spare paper roll in accordance with the present invention have been set forth fully and completely hereinabove , it will be apparent to one of skill in the art that a number of changes in , for example , the material used to construct the tube , the type of press with which the spare paper roll will be used , and the like may be made without departing from the true spirit and scope of the present invention which is accordingly to be limited only by the following claims .", "category": "Performing Operations; Transporting"}
|
{"category": "Mechanical Engineering; Lightning; Heating; Weapons; Blasting", "patent": "a spare web roll , for example a spare paper web roll 1 for a web - fed rotary printing press , has a continuous tube 3 in the area of its axis of rotation 2 , on whose tube body a web 4 of material has been wound , all as seen in fig1 and 2 . a star - shaped , for example three - armed holder 7 is arranged in the end area of the tube , and in particular in the area of the interior diameter , or respectively interior 6 of the tube 3 , whose arms 8 , 9 , 10 rest resiliently against the interior circumference 6 . for example , the holder 7 supports a cylinder - shaped or disk - shaped code carrier 12 , which can be activated , i . e . can be coded and decoded , coaxially with the axis of rotation 2 . the code carrier 12 is an electronic component , which can be coded by inductive means . the holder 7 can be inserted , for example into the end area of the tube 3 of a spare paper web 1 roll , i . e . into the interior 6 of the tube 3 , in such a way that it remains in a temporary first position b , i . e . close to the lateral edges 13 of the web of material 4 as \u2014 represented in dashed lines in fig1 . only when the spare paper web 1 roll is placed on the shaft , is the holder 7 pushed into a second position c by means of a clamping cone , not represented . the holder 7 will now be in the position shown in solid lines in fig1 . after the roll is used up , for example , the holder 7 can again be taken out of the tube 3 by removing it . for this purpose the arms 8 , 9 , 10 can have bores 14 , 16 , 17 , for example , behind which a suitable tool can reach . a number of code carriers 18 , 21 , 23 , 26 are represented in respectively different fastening positions in fig3 . however , only one of which is used . a code carrier 18 is arranged in the front face 19 of the tube 3 , i . e . in the tube body , and may for example be pressed into a bore . a code carrier 21 is fastened to the inner surface 22 of the tube 3 , i . e . in the interior 6 of the tube 3 , for example by gluing . a code carrier 23 is arranged in the tube body 3 . in this case , the tube wall 24 can have a bore into which the code carrier 23 is pressed . it is also possible to arrange a code carrier 26 in the webs of material 4 which are close to the tube 3 , i . e . the inner or lower ones , i . e . to introduce it from the direction of the lateral edges 13 of the webs of material 4 into an area close to the edge , for example to press it in . therefore the code carrier 26 is outside of and in the vicinity of the tube 3 . the code carriers 12 , 18 , 21 , 23 , 26 arranged in the interior of the tube 3 or in the vicinity of the tube 3 , for example the code carrier 18 located in the front face 19 of the tube 3 , are connected with one of the recording and reading heads 27 , 28 , 29 . such recording and reading heads 27 , 28 , 29 are , as shown in fig4 stationed along a conveyance path for movable spare paper web transport carriages , for example the recording and reading head 27 in the vicinity of a storage facility for spare paper web rolls 1 , the recording and reading head 28 is locked in the vicinity of a preparation station for spare paper web , and the recording and reading head 29 is located in the vicinity of a roll changer . each recording and reading head 27 to 29 consists of an electronic component , which transmits energy 37 and information 38 in a contactless manner to the code carrier 18 . on the other end , all recording and reading heads 27 to 29 are connected with each other via a data bus 39 , as well as with an evaluation unit 41 and a control device 42 , as well as a master computer 43 , all as shown in fig4 . the master computer 43 has an input station 46 . the code carrier 18 , as well as the other code carriers 12 , 21 , 23 , 26 , are essentially embodied as data memory devices , and respectively consist of an electronic component with a memory and logical control device . the above mentioned code carriers are , for example , designed as eeprom versions , and can be written on and read out , for example , with up to eight kilobytes . if , for example , a spare paper web roll 1 is moved out of the storage facility \u2014 the transport carriages are , for example , pulled by means of under - floor or driverless transport systems \u2014 the code carrier 18 located in the tube 3 of the spare paper web roll 1 passes by the recording / reading head 27 . as soon as the code carrier 18 comes into the active area of the recording / reading head 27 , the energy 37 required for data transmission is built up and the data 38 are transmitted by the recording / reading head 27 to the code carrier 18 , and the data 44 are transmitted by the code carrier 18 to the recording / reading head 27 , each in a frequency range of , for example , 70 kilohertz . such data can be , besides the data entered by the manufacturer : new destination , for example the spare paper web preparation station , remaining length of paper , particularities of the paper , for example a \u201c beating \u201d roll . the data , which are inductively coupled in by the code carrier 18 , are converted into a digital energy signal and conducted to the evaluation unit 41 . the evaluation unit 41 manages the data transfer 37 , 44 between the recording / reading head 28 and the code carrier 18 and is used as an intermediate memory . the evaluation unit 41 is the connecting member between the master computer 43 , or respectively the control device 42 , and the code carrier 18 . data can be erased or new data can be added , such as for example , a new destination , the gluing preparation station and the type of the glue tip . following the gluing preparation of the spare paper web roll 1 , it is possible to enter a further destination , for example a roll changer , by means of the recording / reading head 28 . after a portion of the spare paper web roll 1 has been used , another entry to the roll data takes place by means of the recording / reading head 29 , as well as a new destination , for example a storage facility . it is also possible to arrange only stationary reading devices for the purpose of reading out the destination at appropriate positions on the transport path , for example at switches . in accordance with a further preferred embodiment , the recording / reading head 29 can be arranged in the roll changer , i . e . for example in a roll cone of a support arm of the roll changer , so that a data exchange can take place between the recording / reading head 29 and a code carrier 18 while the spare paper web roll 1 is still on the shaft . alternatively it is also possible to fasten the recording / reading head 29 on the end of the support arm of the roll changer . the data exchange preferably takes place during the run - out of the spare paper web roll 1 , i . e . while the spare paper web roll 1 turns slowly . it is also possible to utilize a code carrier 12 arranged in the area of the imagined axis of rotation 2 of the spare paper web roll 1 in the interior 6 of the tube 3 . in this case the recording / reading head 12 is arranged in the direction of an extended axis of rotation 2 on the exterior of a support arm of the roll changer . here , the clamping cone is designed to be hollow . while preferred embodiments of a spare paper roll in accordance with the present invention have been set forth fully and completely hereinabove , it will be apparent to one of skill in the art that a number of changes in , for example , the material used to construct the tube , the type of press with which the spare paper roll will be used , and the like may be made without departing from the true spirit and scope of the present invention which is accordingly to be limited only by the following claims ."}
|
Is the categorization of this patent accurate?
| 0.25 |
80dcd2ffc37235c3e85318e8e521b30e8f49d20e8764e0e991f9e17984ba79c6
| 0.429688 | 0.012024 | 0.628906 | 0.012451 | 0.527344 | 0.087402 |
null |
{"patent": "a spare web roll , for example a spare paper web roll 1 for a web - fed rotary printing press , has a continuous tube 3 in the area of its axis of rotation 2 , on whose tube body a web 4 of material has been wound , all as seen in fig1 and 2 . a star - shaped , for example three - armed holder 7 is arranged in the end area of the tube , and in particular in the area of the interior diameter , or respectively interior 6 of the tube 3 , whose arms 8 , 9 , 10 rest resiliently against the interior circumference 6 . for example , the holder 7 supports a cylinder - shaped or disk - shaped code carrier 12 , which can be activated , i . e . can be coded and decoded , coaxially with the axis of rotation 2 . the code carrier 12 is an electronic component , which can be coded by inductive means . the holder 7 can be inserted , for example into the end area of the tube 3 of a spare paper web 1 roll , i . e . into the interior 6 of the tube 3 , in such a way that it remains in a temporary first position b , i . e . close to the lateral edges 13 of the web of material 4 as \u2014 represented in dashed lines in fig1 . only when the spare paper web 1 roll is placed on the shaft , is the holder 7 pushed into a second position c by means of a clamping cone , not represented . the holder 7 will now be in the position shown in solid lines in fig1 . after the roll is used up , for example , the holder 7 can again be taken out of the tube 3 by removing it . for this purpose the arms 8 , 9 , 10 can have bores 14 , 16 , 17 , for example , behind which a suitable tool can reach . a number of code carriers 18 , 21 , 23 , 26 are represented in respectively different fastening positions in fig3 . however , only one of which is used . a code carrier 18 is arranged in the front face 19 of the tube 3 , i . e . in the tube body , and may for example be pressed into a bore . a code carrier 21 is fastened to the inner surface 22 of the tube 3 , i . e . in the interior 6 of the tube 3 , for example by gluing . a code carrier 23 is arranged in the tube body 3 . in this case , the tube wall 24 can have a bore into which the code carrier 23 is pressed . it is also possible to arrange a code carrier 26 in the webs of material 4 which are close to the tube 3 , i . e . the inner or lower ones , i . e . to introduce it from the direction of the lateral edges 13 of the webs of material 4 into an area close to the edge , for example to press it in . therefore the code carrier 26 is outside of and in the vicinity of the tube 3 . the code carriers 12 , 18 , 21 , 23 , 26 arranged in the interior of the tube 3 or in the vicinity of the tube 3 , for example the code carrier 18 located in the front face 19 of the tube 3 , are connected with one of the recording and reading heads 27 , 28 , 29 . such recording and reading heads 27 , 28 , 29 are , as shown in fig4 stationed along a conveyance path for movable spare paper web transport carriages , for example the recording and reading head 27 in the vicinity of a storage facility for spare paper web rolls 1 , the recording and reading head 28 is locked in the vicinity of a preparation station for spare paper web , and the recording and reading head 29 is located in the vicinity of a roll changer . each recording and reading head 27 to 29 consists of an electronic component , which transmits energy 37 and information 38 in a contactless manner to the code carrier 18 . on the other end , all recording and reading heads 27 to 29 are connected with each other via a data bus 39 , as well as with an evaluation unit 41 and a control device 42 , as well as a master computer 43 , all as shown in fig4 . the master computer 43 has an input station 46 . the code carrier 18 , as well as the other code carriers 12 , 21 , 23 , 26 , are essentially embodied as data memory devices , and respectively consist of an electronic component with a memory and logical control device . the above mentioned code carriers are , for example , designed as eeprom versions , and can be written on and read out , for example , with up to eight kilobytes . if , for example , a spare paper web roll 1 is moved out of the storage facility \u2014 the transport carriages are , for example , pulled by means of under - floor or driverless transport systems \u2014 the code carrier 18 located in the tube 3 of the spare paper web roll 1 passes by the recording / reading head 27 . as soon as the code carrier 18 comes into the active area of the recording / reading head 27 , the energy 37 required for data transmission is built up and the data 38 are transmitted by the recording / reading head 27 to the code carrier 18 , and the data 44 are transmitted by the code carrier 18 to the recording / reading head 27 , each in a frequency range of , for example , 70 kilohertz . such data can be , besides the data entered by the manufacturer : new destination , for example the spare paper web preparation station , remaining length of paper , particularities of the paper , for example a \u201c beating \u201d roll . the data , which are inductively coupled in by the code carrier 18 , are converted into a digital energy signal and conducted to the evaluation unit 41 . the evaluation unit 41 manages the data transfer 37 , 44 between the recording / reading head 28 and the code carrier 18 and is used as an intermediate memory . the evaluation unit 41 is the connecting member between the master computer 43 , or respectively the control device 42 , and the code carrier 18 . data can be erased or new data can be added , such as for example , a new destination , the gluing preparation station and the type of the glue tip . following the gluing preparation of the spare paper web roll 1 , it is possible to enter a further destination , for example a roll changer , by means of the recording / reading head 28 . after a portion of the spare paper web roll 1 has been used , another entry to the roll data takes place by means of the recording / reading head 29 , as well as a new destination , for example a storage facility . it is also possible to arrange only stationary reading devices for the purpose of reading out the destination at appropriate positions on the transport path , for example at switches . in accordance with a further preferred embodiment , the recording / reading head 29 can be arranged in the roll changer , i . e . for example in a roll cone of a support arm of the roll changer , so that a data exchange can take place between the recording / reading head 29 and a code carrier 18 while the spare paper web roll 1 is still on the shaft . alternatively it is also possible to fasten the recording / reading head 29 on the end of the support arm of the roll changer . the data exchange preferably takes place during the run - out of the spare paper web roll 1 , i . e . while the spare paper web roll 1 turns slowly . it is also possible to utilize a code carrier 12 arranged in the area of the imagined axis of rotation 2 of the spare paper web roll 1 in the interior 6 of the tube 3 . in this case the recording / reading head 12 is arranged in the direction of an extended axis of rotation 2 on the exterior of a support arm of the roll changer . here , the clamping cone is designed to be hollow . while preferred embodiments of a spare paper roll in accordance with the present invention have been set forth fully and completely hereinabove , it will be apparent to one of skill in the art that a number of changes in , for example , the material used to construct the tube , the type of press with which the spare paper roll will be used , and the like may be made without departing from the true spirit and scope of the present invention which is accordingly to be limited only by the following claims .", "category": "Performing Operations; Transporting"}
|
{"category": "Physics", "patent": "a spare web roll , for example a spare paper web roll 1 for a web - fed rotary printing press , has a continuous tube 3 in the area of its axis of rotation 2 , on whose tube body a web 4 of material has been wound , all as seen in fig1 and 2 . a star - shaped , for example three - armed holder 7 is arranged in the end area of the tube , and in particular in the area of the interior diameter , or respectively interior 6 of the tube 3 , whose arms 8 , 9 , 10 rest resiliently against the interior circumference 6 . for example , the holder 7 supports a cylinder - shaped or disk - shaped code carrier 12 , which can be activated , i . e . can be coded and decoded , coaxially with the axis of rotation 2 . the code carrier 12 is an electronic component , which can be coded by inductive means . the holder 7 can be inserted , for example into the end area of the tube 3 of a spare paper web 1 roll , i . e . into the interior 6 of the tube 3 , in such a way that it remains in a temporary first position b , i . e . close to the lateral edges 13 of the web of material 4 as \u2014 represented in dashed lines in fig1 . only when the spare paper web 1 roll is placed on the shaft , is the holder 7 pushed into a second position c by means of a clamping cone , not represented . the holder 7 will now be in the position shown in solid lines in fig1 . after the roll is used up , for example , the holder 7 can again be taken out of the tube 3 by removing it . for this purpose the arms 8 , 9 , 10 can have bores 14 , 16 , 17 , for example , behind which a suitable tool can reach . a number of code carriers 18 , 21 , 23 , 26 are represented in respectively different fastening positions in fig3 . however , only one of which is used . a code carrier 18 is arranged in the front face 19 of the tube 3 , i . e . in the tube body , and may for example be pressed into a bore . a code carrier 21 is fastened to the inner surface 22 of the tube 3 , i . e . in the interior 6 of the tube 3 , for example by gluing . a code carrier 23 is arranged in the tube body 3 . in this case , the tube wall 24 can have a bore into which the code carrier 23 is pressed . it is also possible to arrange a code carrier 26 in the webs of material 4 which are close to the tube 3 , i . e . the inner or lower ones , i . e . to introduce it from the direction of the lateral edges 13 of the webs of material 4 into an area close to the edge , for example to press it in . therefore the code carrier 26 is outside of and in the vicinity of the tube 3 . the code carriers 12 , 18 , 21 , 23 , 26 arranged in the interior of the tube 3 or in the vicinity of the tube 3 , for example the code carrier 18 located in the front face 19 of the tube 3 , are connected with one of the recording and reading heads 27 , 28 , 29 . such recording and reading heads 27 , 28 , 29 are , as shown in fig4 stationed along a conveyance path for movable spare paper web transport carriages , for example the recording and reading head 27 in the vicinity of a storage facility for spare paper web rolls 1 , the recording and reading head 28 is locked in the vicinity of a preparation station for spare paper web , and the recording and reading head 29 is located in the vicinity of a roll changer . each recording and reading head 27 to 29 consists of an electronic component , which transmits energy 37 and information 38 in a contactless manner to the code carrier 18 . on the other end , all recording and reading heads 27 to 29 are connected with each other via a data bus 39 , as well as with an evaluation unit 41 and a control device 42 , as well as a master computer 43 , all as shown in fig4 . the master computer 43 has an input station 46 . the code carrier 18 , as well as the other code carriers 12 , 21 , 23 , 26 , are essentially embodied as data memory devices , and respectively consist of an electronic component with a memory and logical control device . the above mentioned code carriers are , for example , designed as eeprom versions , and can be written on and read out , for example , with up to eight kilobytes . if , for example , a spare paper web roll 1 is moved out of the storage facility \u2014 the transport carriages are , for example , pulled by means of under - floor or driverless transport systems \u2014 the code carrier 18 located in the tube 3 of the spare paper web roll 1 passes by the recording / reading head 27 . as soon as the code carrier 18 comes into the active area of the recording / reading head 27 , the energy 37 required for data transmission is built up and the data 38 are transmitted by the recording / reading head 27 to the code carrier 18 , and the data 44 are transmitted by the code carrier 18 to the recording / reading head 27 , each in a frequency range of , for example , 70 kilohertz . such data can be , besides the data entered by the manufacturer : new destination , for example the spare paper web preparation station , remaining length of paper , particularities of the paper , for example a \u201c beating \u201d roll . the data , which are inductively coupled in by the code carrier 18 , are converted into a digital energy signal and conducted to the evaluation unit 41 . the evaluation unit 41 manages the data transfer 37 , 44 between the recording / reading head 28 and the code carrier 18 and is used as an intermediate memory . the evaluation unit 41 is the connecting member between the master computer 43 , or respectively the control device 42 , and the code carrier 18 . data can be erased or new data can be added , such as for example , a new destination , the gluing preparation station and the type of the glue tip . following the gluing preparation of the spare paper web roll 1 , it is possible to enter a further destination , for example a roll changer , by means of the recording / reading head 28 . after a portion of the spare paper web roll 1 has been used , another entry to the roll data takes place by means of the recording / reading head 29 , as well as a new destination , for example a storage facility . it is also possible to arrange only stationary reading devices for the purpose of reading out the destination at appropriate positions on the transport path , for example at switches . in accordance with a further preferred embodiment , the recording / reading head 29 can be arranged in the roll changer , i . e . for example in a roll cone of a support arm of the roll changer , so that a data exchange can take place between the recording / reading head 29 and a code carrier 18 while the spare paper web roll 1 is still on the shaft . alternatively it is also possible to fasten the recording / reading head 29 on the end of the support arm of the roll changer . the data exchange preferably takes place during the run - out of the spare paper web roll 1 , i . e . while the spare paper web roll 1 turns slowly . it is also possible to utilize a code carrier 12 arranged in the area of the imagined axis of rotation 2 of the spare paper web roll 1 in the interior 6 of the tube 3 . in this case the recording / reading head 12 is arranged in the direction of an extended axis of rotation 2 on the exterior of a support arm of the roll changer . here , the clamping cone is designed to be hollow . while preferred embodiments of a spare paper roll in accordance with the present invention have been set forth fully and completely hereinabove , it will be apparent to one of skill in the art that a number of changes in , for example , the material used to construct the tube , the type of press with which the spare paper roll will be used , and the like may be made without departing from the true spirit and scope of the present invention which is accordingly to be limited only by the following claims ."}
|
Does the category match the content of the patent?
| 0.25 |
80dcd2ffc37235c3e85318e8e521b30e8f49d20e8764e0e991f9e17984ba79c6
| 0.679688 | 0.169922 | 0.777344 | 0.449219 | 0.507813 | 0.439453 |
null |
{"category": "Performing Operations; Transporting", "patent": "a spare web roll , for example a spare paper web roll 1 for a web - fed rotary printing press , has a continuous tube 3 in the area of its axis of rotation 2 , on whose tube body a web 4 of material has been wound , all as seen in fig1 and 2 . a star - shaped , for example three - armed holder 7 is arranged in the end area of the tube , and in particular in the area of the interior diameter , or respectively interior 6 of the tube 3 , whose arms 8 , 9 , 10 rest resiliently against the interior circumference 6 . for example , the holder 7 supports a cylinder - shaped or disk - shaped code carrier 12 , which can be activated , i . e . can be coded and decoded , coaxially with the axis of rotation 2 . the code carrier 12 is an electronic component , which can be coded by inductive means . the holder 7 can be inserted , for example into the end area of the tube 3 of a spare paper web 1 roll , i . e . into the interior 6 of the tube 3 , in such a way that it remains in a temporary first position b , i . e . close to the lateral edges 13 of the web of material 4 as \u2014 represented in dashed lines in fig1 . only when the spare paper web 1 roll is placed on the shaft , is the holder 7 pushed into a second position c by means of a clamping cone , not represented . the holder 7 will now be in the position shown in solid lines in fig1 . after the roll is used up , for example , the holder 7 can again be taken out of the tube 3 by removing it . for this purpose the arms 8 , 9 , 10 can have bores 14 , 16 , 17 , for example , behind which a suitable tool can reach . a number of code carriers 18 , 21 , 23 , 26 are represented in respectively different fastening positions in fig3 . however , only one of which is used . a code carrier 18 is arranged in the front face 19 of the tube 3 , i . e . in the tube body , and may for example be pressed into a bore . a code carrier 21 is fastened to the inner surface 22 of the tube 3 , i . e . in the interior 6 of the tube 3 , for example by gluing . a code carrier 23 is arranged in the tube body 3 . in this case , the tube wall 24 can have a bore into which the code carrier 23 is pressed . it is also possible to arrange a code carrier 26 in the webs of material 4 which are close to the tube 3 , i . e . the inner or lower ones , i . e . to introduce it from the direction of the lateral edges 13 of the webs of material 4 into an area close to the edge , for example to press it in . therefore the code carrier 26 is outside of and in the vicinity of the tube 3 . the code carriers 12 , 18 , 21 , 23 , 26 arranged in the interior of the tube 3 or in the vicinity of the tube 3 , for example the code carrier 18 located in the front face 19 of the tube 3 , are connected with one of the recording and reading heads 27 , 28 , 29 . such recording and reading heads 27 , 28 , 29 are , as shown in fig4 stationed along a conveyance path for movable spare paper web transport carriages , for example the recording and reading head 27 in the vicinity of a storage facility for spare paper web rolls 1 , the recording and reading head 28 is locked in the vicinity of a preparation station for spare paper web , and the recording and reading head 29 is located in the vicinity of a roll changer . each recording and reading head 27 to 29 consists of an electronic component , which transmits energy 37 and information 38 in a contactless manner to the code carrier 18 . on the other end , all recording and reading heads 27 to 29 are connected with each other via a data bus 39 , as well as with an evaluation unit 41 and a control device 42 , as well as a master computer 43 , all as shown in fig4 . the master computer 43 has an input station 46 . the code carrier 18 , as well as the other code carriers 12 , 21 , 23 , 26 , are essentially embodied as data memory devices , and respectively consist of an electronic component with a memory and logical control device . the above mentioned code carriers are , for example , designed as eeprom versions , and can be written on and read out , for example , with up to eight kilobytes . if , for example , a spare paper web roll 1 is moved out of the storage facility \u2014 the transport carriages are , for example , pulled by means of under - floor or driverless transport systems \u2014 the code carrier 18 located in the tube 3 of the spare paper web roll 1 passes by the recording / reading head 27 . as soon as the code carrier 18 comes into the active area of the recording / reading head 27 , the energy 37 required for data transmission is built up and the data 38 are transmitted by the recording / reading head 27 to the code carrier 18 , and the data 44 are transmitted by the code carrier 18 to the recording / reading head 27 , each in a frequency range of , for example , 70 kilohertz . such data can be , besides the data entered by the manufacturer : new destination , for example the spare paper web preparation station , remaining length of paper , particularities of the paper , for example a \u201c beating \u201d roll . the data , which are inductively coupled in by the code carrier 18 , are converted into a digital energy signal and conducted to the evaluation unit 41 . the evaluation unit 41 manages the data transfer 37 , 44 between the recording / reading head 28 and the code carrier 18 and is used as an intermediate memory . the evaluation unit 41 is the connecting member between the master computer 43 , or respectively the control device 42 , and the code carrier 18 . data can be erased or new data can be added , such as for example , a new destination , the gluing preparation station and the type of the glue tip . following the gluing preparation of the spare paper web roll 1 , it is possible to enter a further destination , for example a roll changer , by means of the recording / reading head 28 . after a portion of the spare paper web roll 1 has been used , another entry to the roll data takes place by means of the recording / reading head 29 , as well as a new destination , for example a storage facility . it is also possible to arrange only stationary reading devices for the purpose of reading out the destination at appropriate positions on the transport path , for example at switches . in accordance with a further preferred embodiment , the recording / reading head 29 can be arranged in the roll changer , i . e . for example in a roll cone of a support arm of the roll changer , so that a data exchange can take place between the recording / reading head 29 and a code carrier 18 while the spare paper web roll 1 is still on the shaft . alternatively it is also possible to fasten the recording / reading head 29 on the end of the support arm of the roll changer . the data exchange preferably takes place during the run - out of the spare paper web roll 1 , i . e . while the spare paper web roll 1 turns slowly . it is also possible to utilize a code carrier 12 arranged in the area of the imagined axis of rotation 2 of the spare paper web roll 1 in the interior 6 of the tube 3 . in this case the recording / reading head 12 is arranged in the direction of an extended axis of rotation 2 on the exterior of a support arm of the roll changer . here , the clamping cone is designed to be hollow . while preferred embodiments of a spare paper roll in accordance with the present invention have been set forth fully and completely hereinabove , it will be apparent to one of skill in the art that a number of changes in , for example , the material used to construct the tube , the type of press with which the spare paper roll will be used , and the like may be made without departing from the true spirit and scope of the present invention which is accordingly to be limited only by the following claims ."}
|
{"patent": "a spare web roll , for example a spare paper web roll 1 for a web - fed rotary printing press , has a continuous tube 3 in the area of its axis of rotation 2 , on whose tube body a web 4 of material has been wound , all as seen in fig1 and 2 . a star - shaped , for example three - armed holder 7 is arranged in the end area of the tube , and in particular in the area of the interior diameter , or respectively interior 6 of the tube 3 , whose arms 8 , 9 , 10 rest resiliently against the interior circumference 6 . for example , the holder 7 supports a cylinder - shaped or disk - shaped code carrier 12 , which can be activated , i . e . can be coded and decoded , coaxially with the axis of rotation 2 . the code carrier 12 is an electronic component , which can be coded by inductive means . the holder 7 can be inserted , for example into the end area of the tube 3 of a spare paper web 1 roll , i . e . into the interior 6 of the tube 3 , in such a way that it remains in a temporary first position b , i . e . close to the lateral edges 13 of the web of material 4 as \u2014 represented in dashed lines in fig1 . only when the spare paper web 1 roll is placed on the shaft , is the holder 7 pushed into a second position c by means of a clamping cone , not represented . the holder 7 will now be in the position shown in solid lines in fig1 . after the roll is used up , for example , the holder 7 can again be taken out of the tube 3 by removing it . for this purpose the arms 8 , 9 , 10 can have bores 14 , 16 , 17 , for example , behind which a suitable tool can reach . a number of code carriers 18 , 21 , 23 , 26 are represented in respectively different fastening positions in fig3 . however , only one of which is used . a code carrier 18 is arranged in the front face 19 of the tube 3 , i . e . in the tube body , and may for example be pressed into a bore . a code carrier 21 is fastened to the inner surface 22 of the tube 3 , i . e . in the interior 6 of the tube 3 , for example by gluing . a code carrier 23 is arranged in the tube body 3 . in this case , the tube wall 24 can have a bore into which the code carrier 23 is pressed . it is also possible to arrange a code carrier 26 in the webs of material 4 which are close to the tube 3 , i . e . the inner or lower ones , i . e . to introduce it from the direction of the lateral edges 13 of the webs of material 4 into an area close to the edge , for example to press it in . therefore the code carrier 26 is outside of and in the vicinity of the tube 3 . the code carriers 12 , 18 , 21 , 23 , 26 arranged in the interior of the tube 3 or in the vicinity of the tube 3 , for example the code carrier 18 located in the front face 19 of the tube 3 , are connected with one of the recording and reading heads 27 , 28 , 29 . such recording and reading heads 27 , 28 , 29 are , as shown in fig4 stationed along a conveyance path for movable spare paper web transport carriages , for example the recording and reading head 27 in the vicinity of a storage facility for spare paper web rolls 1 , the recording and reading head 28 is locked in the vicinity of a preparation station for spare paper web , and the recording and reading head 29 is located in the vicinity of a roll changer . each recording and reading head 27 to 29 consists of an electronic component , which transmits energy 37 and information 38 in a contactless manner to the code carrier 18 . on the other end , all recording and reading heads 27 to 29 are connected with each other via a data bus 39 , as well as with an evaluation unit 41 and a control device 42 , as well as a master computer 43 , all as shown in fig4 . the master computer 43 has an input station 46 . the code carrier 18 , as well as the other code carriers 12 , 21 , 23 , 26 , are essentially embodied as data memory devices , and respectively consist of an electronic component with a memory and logical control device . the above mentioned code carriers are , for example , designed as eeprom versions , and can be written on and read out , for example , with up to eight kilobytes . if , for example , a spare paper web roll 1 is moved out of the storage facility \u2014 the transport carriages are , for example , pulled by means of under - floor or driverless transport systems \u2014 the code carrier 18 located in the tube 3 of the spare paper web roll 1 passes by the recording / reading head 27 . as soon as the code carrier 18 comes into the active area of the recording / reading head 27 , the energy 37 required for data transmission is built up and the data 38 are transmitted by the recording / reading head 27 to the code carrier 18 , and the data 44 are transmitted by the code carrier 18 to the recording / reading head 27 , each in a frequency range of , for example , 70 kilohertz . such data can be , besides the data entered by the manufacturer : new destination , for example the spare paper web preparation station , remaining length of paper , particularities of the paper , for example a \u201c beating \u201d roll . the data , which are inductively coupled in by the code carrier 18 , are converted into a digital energy signal and conducted to the evaluation unit 41 . the evaluation unit 41 manages the data transfer 37 , 44 between the recording / reading head 28 and the code carrier 18 and is used as an intermediate memory . the evaluation unit 41 is the connecting member between the master computer 43 , or respectively the control device 42 , and the code carrier 18 . data can be erased or new data can be added , such as for example , a new destination , the gluing preparation station and the type of the glue tip . following the gluing preparation of the spare paper web roll 1 , it is possible to enter a further destination , for example a roll changer , by means of the recording / reading head 28 . after a portion of the spare paper web roll 1 has been used , another entry to the roll data takes place by means of the recording / reading head 29 , as well as a new destination , for example a storage facility . it is also possible to arrange only stationary reading devices for the purpose of reading out the destination at appropriate positions on the transport path , for example at switches . in accordance with a further preferred embodiment , the recording / reading head 29 can be arranged in the roll changer , i . e . for example in a roll cone of a support arm of the roll changer , so that a data exchange can take place between the recording / reading head 29 and a code carrier 18 while the spare paper web roll 1 is still on the shaft . alternatively it is also possible to fasten the recording / reading head 29 on the end of the support arm of the roll changer . the data exchange preferably takes place during the run - out of the spare paper web roll 1 , i . e . while the spare paper web roll 1 turns slowly . it is also possible to utilize a code carrier 12 arranged in the area of the imagined axis of rotation 2 of the spare paper web roll 1 in the interior 6 of the tube 3 . in this case the recording / reading head 12 is arranged in the direction of an extended axis of rotation 2 on the exterior of a support arm of the roll changer . here , the clamping cone is designed to be hollow . while preferred embodiments of a spare paper roll in accordance with the present invention have been set forth fully and completely hereinabove , it will be apparent to one of skill in the art that a number of changes in , for example , the material used to construct the tube , the type of press with which the spare paper roll will be used , and the like may be made without departing from the true spirit and scope of the present invention which is accordingly to be limited only by the following claims .", "category": "Electricity"}
|
Is the patent correctly categorized?
| 0.25 |
80dcd2ffc37235c3e85318e8e521b30e8f49d20e8764e0e991f9e17984ba79c6
| 0.792969 | 0.088867 | 0.679688 | 0.123535 | 0.941406 | 0.088867 |
null |
{"category": "Performing Operations; Transporting", "patent": "a spare web roll , for example a spare paper web roll 1 for a web - fed rotary printing press , has a continuous tube 3 in the area of its axis of rotation 2 , on whose tube body a web 4 of material has been wound , all as seen in fig1 and 2 . a star - shaped , for example three - armed holder 7 is arranged in the end area of the tube , and in particular in the area of the interior diameter , or respectively interior 6 of the tube 3 , whose arms 8 , 9 , 10 rest resiliently against the interior circumference 6 . for example , the holder 7 supports a cylinder - shaped or disk - shaped code carrier 12 , which can be activated , i . e . can be coded and decoded , coaxially with the axis of rotation 2 . the code carrier 12 is an electronic component , which can be coded by inductive means . the holder 7 can be inserted , for example into the end area of the tube 3 of a spare paper web 1 roll , i . e . into the interior 6 of the tube 3 , in such a way that it remains in a temporary first position b , i . e . close to the lateral edges 13 of the web of material 4 as \u2014 represented in dashed lines in fig1 . only when the spare paper web 1 roll is placed on the shaft , is the holder 7 pushed into a second position c by means of a clamping cone , not represented . the holder 7 will now be in the position shown in solid lines in fig1 . after the roll is used up , for example , the holder 7 can again be taken out of the tube 3 by removing it . for this purpose the arms 8 , 9 , 10 can have bores 14 , 16 , 17 , for example , behind which a suitable tool can reach . a number of code carriers 18 , 21 , 23 , 26 are represented in respectively different fastening positions in fig3 . however , only one of which is used . a code carrier 18 is arranged in the front face 19 of the tube 3 , i . e . in the tube body , and may for example be pressed into a bore . a code carrier 21 is fastened to the inner surface 22 of the tube 3 , i . e . in the interior 6 of the tube 3 , for example by gluing . a code carrier 23 is arranged in the tube body 3 . in this case , the tube wall 24 can have a bore into which the code carrier 23 is pressed . it is also possible to arrange a code carrier 26 in the webs of material 4 which are close to the tube 3 , i . e . the inner or lower ones , i . e . to introduce it from the direction of the lateral edges 13 of the webs of material 4 into an area close to the edge , for example to press it in . therefore the code carrier 26 is outside of and in the vicinity of the tube 3 . the code carriers 12 , 18 , 21 , 23 , 26 arranged in the interior of the tube 3 or in the vicinity of the tube 3 , for example the code carrier 18 located in the front face 19 of the tube 3 , are connected with one of the recording and reading heads 27 , 28 , 29 . such recording and reading heads 27 , 28 , 29 are , as shown in fig4 stationed along a conveyance path for movable spare paper web transport carriages , for example the recording and reading head 27 in the vicinity of a storage facility for spare paper web rolls 1 , the recording and reading head 28 is locked in the vicinity of a preparation station for spare paper web , and the recording and reading head 29 is located in the vicinity of a roll changer . each recording and reading head 27 to 29 consists of an electronic component , which transmits energy 37 and information 38 in a contactless manner to the code carrier 18 . on the other end , all recording and reading heads 27 to 29 are connected with each other via a data bus 39 , as well as with an evaluation unit 41 and a control device 42 , as well as a master computer 43 , all as shown in fig4 . the master computer 43 has an input station 46 . the code carrier 18 , as well as the other code carriers 12 , 21 , 23 , 26 , are essentially embodied as data memory devices , and respectively consist of an electronic component with a memory and logical control device . the above mentioned code carriers are , for example , designed as eeprom versions , and can be written on and read out , for example , with up to eight kilobytes . if , for example , a spare paper web roll 1 is moved out of the storage facility \u2014 the transport carriages are , for example , pulled by means of under - floor or driverless transport systems \u2014 the code carrier 18 located in the tube 3 of the spare paper web roll 1 passes by the recording / reading head 27 . as soon as the code carrier 18 comes into the active area of the recording / reading head 27 , the energy 37 required for data transmission is built up and the data 38 are transmitted by the recording / reading head 27 to the code carrier 18 , and the data 44 are transmitted by the code carrier 18 to the recording / reading head 27 , each in a frequency range of , for example , 70 kilohertz . such data can be , besides the data entered by the manufacturer : new destination , for example the spare paper web preparation station , remaining length of paper , particularities of the paper , for example a \u201c beating \u201d roll . the data , which are inductively coupled in by the code carrier 18 , are converted into a digital energy signal and conducted to the evaluation unit 41 . the evaluation unit 41 manages the data transfer 37 , 44 between the recording / reading head 28 and the code carrier 18 and is used as an intermediate memory . the evaluation unit 41 is the connecting member between the master computer 43 , or respectively the control device 42 , and the code carrier 18 . data can be erased or new data can be added , such as for example , a new destination , the gluing preparation station and the type of the glue tip . following the gluing preparation of the spare paper web roll 1 , it is possible to enter a further destination , for example a roll changer , by means of the recording / reading head 28 . after a portion of the spare paper web roll 1 has been used , another entry to the roll data takes place by means of the recording / reading head 29 , as well as a new destination , for example a storage facility . it is also possible to arrange only stationary reading devices for the purpose of reading out the destination at appropriate positions on the transport path , for example at switches . in accordance with a further preferred embodiment , the recording / reading head 29 can be arranged in the roll changer , i . e . for example in a roll cone of a support arm of the roll changer , so that a data exchange can take place between the recording / reading head 29 and a code carrier 18 while the spare paper web roll 1 is still on the shaft . alternatively it is also possible to fasten the recording / reading head 29 on the end of the support arm of the roll changer . the data exchange preferably takes place during the run - out of the spare paper web roll 1 , i . e . while the spare paper web roll 1 turns slowly . it is also possible to utilize a code carrier 12 arranged in the area of the imagined axis of rotation 2 of the spare paper web roll 1 in the interior 6 of the tube 3 . in this case the recording / reading head 12 is arranged in the direction of an extended axis of rotation 2 on the exterior of a support arm of the roll changer . here , the clamping cone is designed to be hollow . while preferred embodiments of a spare paper roll in accordance with the present invention have been set forth fully and completely hereinabove , it will be apparent to one of skill in the art that a number of changes in , for example , the material used to construct the tube , the type of press with which the spare paper roll will be used , and the like may be made without departing from the true spirit and scope of the present invention which is accordingly to be limited only by the following claims ."}
|
{"category": "General tagging of new or cross-sectional technology", "patent": "a spare web roll , for example a spare paper web roll 1 for a web - fed rotary printing press , has a continuous tube 3 in the area of its axis of rotation 2 , on whose tube body a web 4 of material has been wound , all as seen in fig1 and 2 . a star - shaped , for example three - armed holder 7 is arranged in the end area of the tube , and in particular in the area of the interior diameter , or respectively interior 6 of the tube 3 , whose arms 8 , 9 , 10 rest resiliently against the interior circumference 6 . for example , the holder 7 supports a cylinder - shaped or disk - shaped code carrier 12 , which can be activated , i . e . can be coded and decoded , coaxially with the axis of rotation 2 . the code carrier 12 is an electronic component , which can be coded by inductive means . the holder 7 can be inserted , for example into the end area of the tube 3 of a spare paper web 1 roll , i . e . into the interior 6 of the tube 3 , in such a way that it remains in a temporary first position b , i . e . close to the lateral edges 13 of the web of material 4 as \u2014 represented in dashed lines in fig1 . only when the spare paper web 1 roll is placed on the shaft , is the holder 7 pushed into a second position c by means of a clamping cone , not represented . the holder 7 will now be in the position shown in solid lines in fig1 . after the roll is used up , for example , the holder 7 can again be taken out of the tube 3 by removing it . for this purpose the arms 8 , 9 , 10 can have bores 14 , 16 , 17 , for example , behind which a suitable tool can reach . a number of code carriers 18 , 21 , 23 , 26 are represented in respectively different fastening positions in fig3 . however , only one of which is used . a code carrier 18 is arranged in the front face 19 of the tube 3 , i . e . in the tube body , and may for example be pressed into a bore . a code carrier 21 is fastened to the inner surface 22 of the tube 3 , i . e . in the interior 6 of the tube 3 , for example by gluing . a code carrier 23 is arranged in the tube body 3 . in this case , the tube wall 24 can have a bore into which the code carrier 23 is pressed . it is also possible to arrange a code carrier 26 in the webs of material 4 which are close to the tube 3 , i . e . the inner or lower ones , i . e . to introduce it from the direction of the lateral edges 13 of the webs of material 4 into an area close to the edge , for example to press it in . therefore the code carrier 26 is outside of and in the vicinity of the tube 3 . the code carriers 12 , 18 , 21 , 23 , 26 arranged in the interior of the tube 3 or in the vicinity of the tube 3 , for example the code carrier 18 located in the front face 19 of the tube 3 , are connected with one of the recording and reading heads 27 , 28 , 29 . such recording and reading heads 27 , 28 , 29 are , as shown in fig4 stationed along a conveyance path for movable spare paper web transport carriages , for example the recording and reading head 27 in the vicinity of a storage facility for spare paper web rolls 1 , the recording and reading head 28 is locked in the vicinity of a preparation station for spare paper web , and the recording and reading head 29 is located in the vicinity of a roll changer . each recording and reading head 27 to 29 consists of an electronic component , which transmits energy 37 and information 38 in a contactless manner to the code carrier 18 . on the other end , all recording and reading heads 27 to 29 are connected with each other via a data bus 39 , as well as with an evaluation unit 41 and a control device 42 , as well as a master computer 43 , all as shown in fig4 . the master computer 43 has an input station 46 . the code carrier 18 , as well as the other code carriers 12 , 21 , 23 , 26 , are essentially embodied as data memory devices , and respectively consist of an electronic component with a memory and logical control device . the above mentioned code carriers are , for example , designed as eeprom versions , and can be written on and read out , for example , with up to eight kilobytes . if , for example , a spare paper web roll 1 is moved out of the storage facility \u2014 the transport carriages are , for example , pulled by means of under - floor or driverless transport systems \u2014 the code carrier 18 located in the tube 3 of the spare paper web roll 1 passes by the recording / reading head 27 . as soon as the code carrier 18 comes into the active area of the recording / reading head 27 , the energy 37 required for data transmission is built up and the data 38 are transmitted by the recording / reading head 27 to the code carrier 18 , and the data 44 are transmitted by the code carrier 18 to the recording / reading head 27 , each in a frequency range of , for example , 70 kilohertz . such data can be , besides the data entered by the manufacturer : new destination , for example the spare paper web preparation station , remaining length of paper , particularities of the paper , for example a \u201c beating \u201d roll . the data , which are inductively coupled in by the code carrier 18 , are converted into a digital energy signal and conducted to the evaluation unit 41 . the evaluation unit 41 manages the data transfer 37 , 44 between the recording / reading head 28 and the code carrier 18 and is used as an intermediate memory . the evaluation unit 41 is the connecting member between the master computer 43 , or respectively the control device 42 , and the code carrier 18 . data can be erased or new data can be added , such as for example , a new destination , the gluing preparation station and the type of the glue tip . following the gluing preparation of the spare paper web roll 1 , it is possible to enter a further destination , for example a roll changer , by means of the recording / reading head 28 . after a portion of the spare paper web roll 1 has been used , another entry to the roll data takes place by means of the recording / reading head 29 , as well as a new destination , for example a storage facility . it is also possible to arrange only stationary reading devices for the purpose of reading out the destination at appropriate positions on the transport path , for example at switches . in accordance with a further preferred embodiment , the recording / reading head 29 can be arranged in the roll changer , i . e . for example in a roll cone of a support arm of the roll changer , so that a data exchange can take place between the recording / reading head 29 and a code carrier 18 while the spare paper web roll 1 is still on the shaft . alternatively it is also possible to fasten the recording / reading head 29 on the end of the support arm of the roll changer . the data exchange preferably takes place during the run - out of the spare paper web roll 1 , i . e . while the spare paper web roll 1 turns slowly . it is also possible to utilize a code carrier 12 arranged in the area of the imagined axis of rotation 2 of the spare paper web roll 1 in the interior 6 of the tube 3 . in this case the recording / reading head 12 is arranged in the direction of an extended axis of rotation 2 on the exterior of a support arm of the roll changer . here , the clamping cone is designed to be hollow . while preferred embodiments of a spare paper roll in accordance with the present invention have been set forth fully and completely hereinabove , it will be apparent to one of skill in the art that a number of changes in , for example , the material used to construct the tube , the type of press with which the spare paper roll will be used , and the like may be made without departing from the true spirit and scope of the present invention which is accordingly to be limited only by the following claims ."}
|
Does the patent belong in this category?
| 0.25 |
80dcd2ffc37235c3e85318e8e521b30e8f49d20e8764e0e991f9e17984ba79c6
| 0.863281 | 0.451172 | 0.707031 | 0.652344 | 0.957031 | 0.585938 |
null |
{"category": "Electricity", "patent": "as noted above , the object of the present invention is to distribute information . this information is transmitted over , for example , telephone lines and converted on the client site into video signals similar to those for television reception . the information subscribed to takes the form of pages of market data , various portions of which are updated from time to time to reflect changes in the market , and subsequently presented on a video display . in a first embodiment of the invention , the video signals representing the market information are being transmitted asynchronously , that is , there are no set characteristic times ( e . g . television field and frame rates ) to restrict the transmission . as shown in fig1 pages 10 of market data are each given individual page key ( pk ) codes , for example &# 34 ; page 203 &# 34 ; and &# 34 ; page 7677 &# 34 ;. these pages 10 are then applied to an encoder 12 for application to a video transmission line 14 . video displays 16 are shown connected to the transmission line 14 for receiving the encoded pages 10 . as shown , each of the video displays 16 has a unique display identification ( id ) code , for example 217 , 503 , 123 and 417 , as shown . fig2 shows a diagram representing the video signals transmitted over the video transmission line 14 . since these signals are in the form of television signals , it should be understood that the video signals for each page of market information is in the form of a sequence of video lines which , when scanned on the video displays , form the relevant pages . as shown in fig2 a first line 20 of the transmission includes an encoded signal flag indicating to the video displays 16 that the following information is encoded data . the exact form of the flag is unimportant since the information contained is just one bit . hence , the flag may be similar to , but not identical to , the vertical synchronizing signal indicating the beginning of each field of information . the line or lines 22 contain enable reception messages . the lines 24 following the enable reception message lines 22 , contain the various updates 1 , 2 and 3 . fig3 a shows a sample enable reception ( er ) message line 22 in detail . following a horizontal synchronizing pulse 30 , an er synchronizing signal 32 is sent indicating the ensuing transmission of enable reception messages and enabling decoders to synchronize to the transmission . the er sync . signal 32 is followed by id / pk conversion pairs 34 each of which includes one of the unique display id codes and one of the individual page key ( pk ) codes for which the display identified by the display code is authorized to receive . in the example shown , the pairs 417 / 7677 , 503 / 203 and 123 / 7677 indicate that video display 417 is authorized to receive page 7677 , display 503 , page 203 ; and display 123 , page 7677 . it should be noted that in the example , display 417 is not authorized to receive either page 203 or page 7677 . the enable reception message continues for as many lines ( each including an er sync . signal 32 ) and includes as many pairs 34 as are required to associate each of the authorized displays with one of the ( many ) subscribed to pages . the process for updating each page is performed by replacing &# 34 ; tiles &# 34 ; in the relevant page . as shown in fig4 the cross - hatched tile 36 to be updated is located by two row and two column ( or two x - y pixel pair ) coordinates . fig3 b - 3f show samples of the data enable sequences , in which , in fig3 b , the sequence for the update of page 7677 is illustrated . in particular , after a horizontal synchronizing pulse 40 , a data synchronizing signal 42 is present . this sync . signal 42 , which indicates the ensuing transmission of a data enable sequence , is followed by the individual page code 44 for the page 7677 and then the coordinates 46 of the tile 36 to be replaced which , in this example , is 4 rows by 40 columns . the actual data for this tile 36 is presented in a series of lines , corresponding to the number of rows in the tile to be updated , following the data enable sequence line . similar examples are shown for pages 203 and 7677 in fig3 c and 3d , in which in page 203 , a tile of 1 row and 11 columns is updated , and again in page 7677 , a second tile of 6 rows and 40 columns is updated . alternatively , as shown in fig3 e , the update data may appear on the same line as the data enable sequence . in particular , the sync . signal 42 &# 39 ; is followed by the individual page code 44 . however , the coordinates 46 &# 39 ; include the pixel start number and the pixel stop number of a single row of the update data , along with the line number of the particular line . the update data 48 then follows on the same line . each tile 36 is then composed of the update data 48 appearing in , for example , a plurality of consecutive lines . further , as shown in fig3 f , the update data may be presented simultaneously on one line for more than one page at a time . in particular , the sync . signal 42 &# 34 ; is followed by two page codes 44 &# 34 ; , and then the coordinates 46 of the tile 36 to be replaced , which in this example , is 5 rows by 80 columns , toward the bottom of pages 203 and 7677 . additionally , all authorized displays connected to the video transmission may be simultaneously updated at the same coordinates by using a special page key , e . g . pk = 0 . an encoder for the first embodiment of the invention is shown in fig5 . the encoder includes a modem 50 for receiving data from a source of market information . this data may be in the form of entire pages of financial information where portions are updated , or the update data itself along with information for the positioning of the update data on the respective display screen . the output of modem 50 is connected to an interface 51 , which is , in turn , connected to the input of a microcomputer 52 . the microcomputer 52 reassigns the data to appropriate locations in new pages for clients of the provider . a keyboard 53 is connected to the microcomputer 52 for controlling the microcomputer . a memory 54 is connected to the microcomputer 52 and supplies thereto the configuration of the new pages , the individual page key ( pk ) code for each of the new pages , and the display ( id ) codes of the clients authorized to receive each of the new pages . based on this information , the microcomputer 52 generates the first data stream and the sequence of second data streams . the output of the microcomputer 52 is applied through an interface 55 , to a video generation unit 56 which reconfigures the output of the microcomputer into video lines . the video generation unit 56 also generates the encoded signal flag and inserts the various synchronizing signals at the beginning of each of the video lines . a clock signal generator 57 is connected to the video generation unit 56 and the microcomputer 52 for applying timing signals thereto at the line frequency . in the event that the financial information applied to the modem 50 is in the form of entire pages , a memory 58 is connected to the microcomputer 52 into which the pages are entered enabling the microcomputer 52 to compare one page with the update of the page to extract therefrom only the update data . fig6 is a block diagram of a decoder for use with the encoder of fig5 . the decoder includes a receiver 60 for receiving the data transmitted by the video generation unit 56 . the output of the receiver 60 is applied to an analog switch 61 for selective application to an output display in the event that standard non - coded signals are being transmitted . a coded signal detector 62 is coupled to the receiver 60 for receiving the encoded signal flag and for switching the analog switch 61 accordingly . an er detection gate 63 is connected to the receiver 60 for receiving the enable reception messages containing the display id / individual pk code pairs . each of the received display id codes are compared with a unique display id code stored in a rom 64 by a comparator 65 . upon each match of the display id code , the individual pk code for the respective page is stored in a memory 66 . the output of the receiver 60 is further connected to a data detection gate 67 for receiving the data enable sequences . the individual pk codes in the received data enable sequences are compared in a comparator 68 with the individual pk codes stored in the memory 66 . upon a match of one of these pk codes , the accompanying coordinates of the update data are loaded into registers 69 . an analog - to - digital converter 70 digitizes the appropriate update data at the output of the receiver 60 and applies its output to a write buffer 71 , which also receives the output of the registers 69 . the output of the write buffer 71 is applied to a picture store 72 in which the section therein corresponding to the location of the update data is updated . a synchronizing signal detector 73 is connected to the output of the receiver 60 for separating the line synchronizing signals . the output of the synchronizing signal detector 73 is applied to a timing and control signal generator 74 for generating timing signals for the analog - to - digital converter 70 , the data detection gate 67 , the er detection gate 63 and the picture store 72 . the output of the picture store 72 is applied to a digital - to - analog converter 75 controlled by the timing and control signal generator 74 . the output of the digital - to - analog converter 75 is applied through a low - pass filter 76 to another input of the analog switch 61 . in a second embodiment of the invention , the video signals representing the market information include three colors . in addition , standard television signals are included in the video signals for selective viewing of realtime television on the video displays . this transmission is necessarily synchronous to the chosen television standard . assuming that the video signals are being transmitted by cable , resulting in a usable bandwidth of approximately 24 mhz . fig7 shows a pictorial representation of the transmitted video signals . the encoded signal flag line 80 , the enable reception messages lines 81 and the data enable sequence lines 82 are transmitted during the vertical blanking interval 83 between each field of the video signal . during the active video portion of the field , in a first half of each scanning line , the television r , g and b signals 84 , each originally having a bandwidth of 4 mhz . and each time compressed by a factor of six to an expanded bandwidth of 24 mhz ., are sequentially transmitted . in the second half of each scanning line , the update data for individual pages of the market information are transmitted . while the television signals 84 are in color , the update information may be monochromatic , color or a mixture of both . in particular , as shown , the first 8 half - lines contain the update monochrome data 85 for the left half and right halves , respectively , for page 7677 . the update data 85 for page 7677 is followed by the update data 86 for page 203 . the update data 86 is presented in color as the three color signals r , g and b . the remainder of the right half of the first field is shown as being unused in this example . the left half of the second field contains the g , b and r components of the television signals 78 &# 39 ;. the right half of the second field contains the monochromatic update data for the pages 208 , 1234 , 5 , 19154 and 264 . due to the complex ordering of the update data in the first and second fields , the data enable sequences in the lines 82 must necessarily be more complex than those shown in fig3 b - 3f . in addition , the enable reception messages must indicate which of the video displays is authorized to receive the television signals sent with the update data . in particular , as shown in fig8 a , the enable reception messages are similar to those shown in fig3 a , with the exception that in addition to the display id code / identification pk code pairs , the messages include a pair 87 indicating which of the video displays , for example , the display with display id code 297 , is authorized to receive the television signals tv1 . fig8 b shows a sample data enable sequence which includes , in addition to that described with respect to fig3 b - 3f , the coordinates of the update data in the source field . fig8 c shows a sample of the data enable sequence for identifying the television signals , and includes a data synchronizing signal 88 , a television pk code 89 and the starting coordinates 90 of the three color signals -- red , green and blue . fig9 shows a block diagram of an encoder for the second embodiment . the video generation unit 56 &# 39 ; has a second set of inputs for receiving the three color components of the television signals . in particular , a source of video signals is connected to a synchronizing signal separation circuit 91 for detecting the vertical and horizontal synchronizing signals in the video signals . the source of the video signals is also connected to a matrix circuit 92 for providing the three color components . each of these components is subjected to a 6 : 1 compression in compression circuit 93 and the three components are then applied to the video generation unit 56 &# 39 ;. the clock signal generator 57 &# 39 ; applies horizontal and vertical synchronizing signals to both the video generation unit 56 &# 39 ; and the microcomputer 52 &# 39 ;, and receives the synchronizing signals from the separation circuit 91 for synchronization therewith . fig1 shows a block diagram of a decoder for the second embodiment . components the same as those in fig6 are designated with the same reference number . the decoder is substantially similar as the decoder of the first embodiment with the exception that the decoder is now capable of processing color signals and the encoded data selectively includes television signals . in particular , a color decoder 101 is included between the output of the receiver 60 and the input of the analog switch 61 . 1 - 61 . 3 . the register 69 &# 39 ; includes a register element for storing the number of the picture store . the synchronizing signal detector 73 &# 39 ; outputs field synchronizing signals in addition to line synchronizing signals . the write buffer 71 &# 39 ; now accesses three picture stores 72 . 1 - 72 . 3 corresponding to the three color components , red , blue and green . the outputs of these picture stores 72 . 1 - 72 . 3 are applied to three digital - to - analog converters 75 . 1 - 75 . 3 , and then to three low - pass filters 76 . 1 - 76 . 3 for application to the other inputs of the three analog switches 61 . 1 - 61 . 3 . numerous alterations of the structure herein disclosed will suggest themselves to those skilled in the art . however it is to be understood that the embodiments herein disclosed are for purposes of illustration only and not to be construed as a limitation of the invention . all such modifications which do not depart from the spirit of the invention are intended to be included within the scope of the appended claims ."}
|
{"patent": "as noted above , the object of the present invention is to distribute information . this information is transmitted over , for example , telephone lines and converted on the client site into video signals similar to those for television reception . the information subscribed to takes the form of pages of market data , various portions of which are updated from time to time to reflect changes in the market , and subsequently presented on a video display . in a first embodiment of the invention , the video signals representing the market information are being transmitted asynchronously , that is , there are no set characteristic times ( e . g . television field and frame rates ) to restrict the transmission . as shown in fig1 pages 10 of market data are each given individual page key ( pk ) codes , for example &# 34 ; page 203 &# 34 ; and &# 34 ; page 7677 &# 34 ;. these pages 10 are then applied to an encoder 12 for application to a video transmission line 14 . video displays 16 are shown connected to the transmission line 14 for receiving the encoded pages 10 . as shown , each of the video displays 16 has a unique display identification ( id ) code , for example 217 , 503 , 123 and 417 , as shown . fig2 shows a diagram representing the video signals transmitted over the video transmission line 14 . since these signals are in the form of television signals , it should be understood that the video signals for each page of market information is in the form of a sequence of video lines which , when scanned on the video displays , form the relevant pages . as shown in fig2 a first line 20 of the transmission includes an encoded signal flag indicating to the video displays 16 that the following information is encoded data . the exact form of the flag is unimportant since the information contained is just one bit . hence , the flag may be similar to , but not identical to , the vertical synchronizing signal indicating the beginning of each field of information . the line or lines 22 contain enable reception messages . the lines 24 following the enable reception message lines 22 , contain the various updates 1 , 2 and 3 . fig3 a shows a sample enable reception ( er ) message line 22 in detail . following a horizontal synchronizing pulse 30 , an er synchronizing signal 32 is sent indicating the ensuing transmission of enable reception messages and enabling decoders to synchronize to the transmission . the er sync . signal 32 is followed by id / pk conversion pairs 34 each of which includes one of the unique display id codes and one of the individual page key ( pk ) codes for which the display identified by the display code is authorized to receive . in the example shown , the pairs 417 / 7677 , 503 / 203 and 123 / 7677 indicate that video display 417 is authorized to receive page 7677 , display 503 , page 203 ; and display 123 , page 7677 . it should be noted that in the example , display 417 is not authorized to receive either page 203 or page 7677 . the enable reception message continues for as many lines ( each including an er sync . signal 32 ) and includes as many pairs 34 as are required to associate each of the authorized displays with one of the ( many ) subscribed to pages . the process for updating each page is performed by replacing &# 34 ; tiles &# 34 ; in the relevant page . as shown in fig4 the cross - hatched tile 36 to be updated is located by two row and two column ( or two x - y pixel pair ) coordinates . fig3 b - 3f show samples of the data enable sequences , in which , in fig3 b , the sequence for the update of page 7677 is illustrated . in particular , after a horizontal synchronizing pulse 40 , a data synchronizing signal 42 is present . this sync . signal 42 , which indicates the ensuing transmission of a data enable sequence , is followed by the individual page code 44 for the page 7677 and then the coordinates 46 of the tile 36 to be replaced which , in this example , is 4 rows by 40 columns . the actual data for this tile 36 is presented in a series of lines , corresponding to the number of rows in the tile to be updated , following the data enable sequence line . similar examples are shown for pages 203 and 7677 in fig3 c and 3d , in which in page 203 , a tile of 1 row and 11 columns is updated , and again in page 7677 , a second tile of 6 rows and 40 columns is updated . alternatively , as shown in fig3 e , the update data may appear on the same line as the data enable sequence . in particular , the sync . signal 42 &# 39 ; is followed by the individual page code 44 . however , the coordinates 46 &# 39 ; include the pixel start number and the pixel stop number of a single row of the update data , along with the line number of the particular line . the update data 48 then follows on the same line . each tile 36 is then composed of the update data 48 appearing in , for example , a plurality of consecutive lines . further , as shown in fig3 f , the update data may be presented simultaneously on one line for more than one page at a time . in particular , the sync . signal 42 &# 34 ; is followed by two page codes 44 &# 34 ; , and then the coordinates 46 of the tile 36 to be replaced , which in this example , is 5 rows by 80 columns , toward the bottom of pages 203 and 7677 . additionally , all authorized displays connected to the video transmission may be simultaneously updated at the same coordinates by using a special page key , e . g . pk = 0 . an encoder for the first embodiment of the invention is shown in fig5 . the encoder includes a modem 50 for receiving data from a source of market information . this data may be in the form of entire pages of financial information where portions are updated , or the update data itself along with information for the positioning of the update data on the respective display screen . the output of modem 50 is connected to an interface 51 , which is , in turn , connected to the input of a microcomputer 52 . the microcomputer 52 reassigns the data to appropriate locations in new pages for clients of the provider . a keyboard 53 is connected to the microcomputer 52 for controlling the microcomputer . a memory 54 is connected to the microcomputer 52 and supplies thereto the configuration of the new pages , the individual page key ( pk ) code for each of the new pages , and the display ( id ) codes of the clients authorized to receive each of the new pages . based on this information , the microcomputer 52 generates the first data stream and the sequence of second data streams . the output of the microcomputer 52 is applied through an interface 55 , to a video generation unit 56 which reconfigures the output of the microcomputer into video lines . the video generation unit 56 also generates the encoded signal flag and inserts the various synchronizing signals at the beginning of each of the video lines . a clock signal generator 57 is connected to the video generation unit 56 and the microcomputer 52 for applying timing signals thereto at the line frequency . in the event that the financial information applied to the modem 50 is in the form of entire pages , a memory 58 is connected to the microcomputer 52 into which the pages are entered enabling the microcomputer 52 to compare one page with the update of the page to extract therefrom only the update data . fig6 is a block diagram of a decoder for use with the encoder of fig5 . the decoder includes a receiver 60 for receiving the data transmitted by the video generation unit 56 . the output of the receiver 60 is applied to an analog switch 61 for selective application to an output display in the event that standard non - coded signals are being transmitted . a coded signal detector 62 is coupled to the receiver 60 for receiving the encoded signal flag and for switching the analog switch 61 accordingly . an er detection gate 63 is connected to the receiver 60 for receiving the enable reception messages containing the display id / individual pk code pairs . each of the received display id codes are compared with a unique display id code stored in a rom 64 by a comparator 65 . upon each match of the display id code , the individual pk code for the respective page is stored in a memory 66 . the output of the receiver 60 is further connected to a data detection gate 67 for receiving the data enable sequences . the individual pk codes in the received data enable sequences are compared in a comparator 68 with the individual pk codes stored in the memory 66 . upon a match of one of these pk codes , the accompanying coordinates of the update data are loaded into registers 69 . an analog - to - digital converter 70 digitizes the appropriate update data at the output of the receiver 60 and applies its output to a write buffer 71 , which also receives the output of the registers 69 . the output of the write buffer 71 is applied to a picture store 72 in which the section therein corresponding to the location of the update data is updated . a synchronizing signal detector 73 is connected to the output of the receiver 60 for separating the line synchronizing signals . the output of the synchronizing signal detector 73 is applied to a timing and control signal generator 74 for generating timing signals for the analog - to - digital converter 70 , the data detection gate 67 , the er detection gate 63 and the picture store 72 . the output of the picture store 72 is applied to a digital - to - analog converter 75 controlled by the timing and control signal generator 74 . the output of the digital - to - analog converter 75 is applied through a low - pass filter 76 to another input of the analog switch 61 . in a second embodiment of the invention , the video signals representing the market information include three colors . in addition , standard television signals are included in the video signals for selective viewing of realtime television on the video displays . this transmission is necessarily synchronous to the chosen television standard . assuming that the video signals are being transmitted by cable , resulting in a usable bandwidth of approximately 24 mhz . fig7 shows a pictorial representation of the transmitted video signals . the encoded signal flag line 80 , the enable reception messages lines 81 and the data enable sequence lines 82 are transmitted during the vertical blanking interval 83 between each field of the video signal . during the active video portion of the field , in a first half of each scanning line , the television r , g and b signals 84 , each originally having a bandwidth of 4 mhz . and each time compressed by a factor of six to an expanded bandwidth of 24 mhz ., are sequentially transmitted . in the second half of each scanning line , the update data for individual pages of the market information are transmitted . while the television signals 84 are in color , the update information may be monochromatic , color or a mixture of both . in particular , as shown , the first 8 half - lines contain the update monochrome data 85 for the left half and right halves , respectively , for page 7677 . the update data 85 for page 7677 is followed by the update data 86 for page 203 . the update data 86 is presented in color as the three color signals r , g and b . the remainder of the right half of the first field is shown as being unused in this example . the left half of the second field contains the g , b and r components of the television signals 78 &# 39 ;. the right half of the second field contains the monochromatic update data for the pages 208 , 1234 , 5 , 19154 and 264 . due to the complex ordering of the update data in the first and second fields , the data enable sequences in the lines 82 must necessarily be more complex than those shown in fig3 b - 3f . in addition , the enable reception messages must indicate which of the video displays is authorized to receive the television signals sent with the update data . in particular , as shown in fig8 a , the enable reception messages are similar to those shown in fig3 a , with the exception that in addition to the display id code / identification pk code pairs , the messages include a pair 87 indicating which of the video displays , for example , the display with display id code 297 , is authorized to receive the television signals tv1 . fig8 b shows a sample data enable sequence which includes , in addition to that described with respect to fig3 b - 3f , the coordinates of the update data in the source field . fig8 c shows a sample of the data enable sequence for identifying the television signals , and includes a data synchronizing signal 88 , a television pk code 89 and the starting coordinates 90 of the three color signals -- red , green and blue . fig9 shows a block diagram of an encoder for the second embodiment . the video generation unit 56 &# 39 ; has a second set of inputs for receiving the three color components of the television signals . in particular , a source of video signals is connected to a synchronizing signal separation circuit 91 for detecting the vertical and horizontal synchronizing signals in the video signals . the source of the video signals is also connected to a matrix circuit 92 for providing the three color components . each of these components is subjected to a 6 : 1 compression in compression circuit 93 and the three components are then applied to the video generation unit 56 &# 39 ;. the clock signal generator 57 &# 39 ; applies horizontal and vertical synchronizing signals to both the video generation unit 56 &# 39 ; and the microcomputer 52 &# 39 ;, and receives the synchronizing signals from the separation circuit 91 for synchronization therewith . fig1 shows a block diagram of a decoder for the second embodiment . components the same as those in fig6 are designated with the same reference number . the decoder is substantially similar as the decoder of the first embodiment with the exception that the decoder is now capable of processing color signals and the encoded data selectively includes television signals . in particular , a color decoder 101 is included between the output of the receiver 60 and the input of the analog switch 61 . 1 - 61 . 3 . the register 69 &# 39 ; includes a register element for storing the number of the picture store . the synchronizing signal detector 73 &# 39 ; outputs field synchronizing signals in addition to line synchronizing signals . the write buffer 71 &# 39 ; now accesses three picture stores 72 . 1 - 72 . 3 corresponding to the three color components , red , blue and green . the outputs of these picture stores 72 . 1 - 72 . 3 are applied to three digital - to - analog converters 75 . 1 - 75 . 3 , and then to three low - pass filters 76 . 1 - 76 . 3 for application to the other inputs of the three analog switches 61 . 1 - 61 . 3 . numerous alterations of the structure herein disclosed will suggest themselves to those skilled in the art . however it is to be understood that the embodiments herein disclosed are for purposes of illustration only and not to be construed as a limitation of the invention . all such modifications which do not depart from the spirit of the invention are intended to be included within the scope of the appended claims .", "category": "Human Necessities"}
|
Does the patent belong in this category?
| 0.25 |
8790d674f8c57a9e5539268a37bf38022b241b63e3f4f82bc8f6033d6eb96c97
| 0.574219 | 0.000051 | 0.671875 | 0.001549 | 0.851563 | 0.003601 |
null |
{"category": "Electricity", "patent": "as noted above , the object of the present invention is to distribute information . this information is transmitted over , for example , telephone lines and converted on the client site into video signals similar to those for television reception . the information subscribed to takes the form of pages of market data , various portions of which are updated from time to time to reflect changes in the market , and subsequently presented on a video display . in a first embodiment of the invention , the video signals representing the market information are being transmitted asynchronously , that is , there are no set characteristic times ( e . g . television field and frame rates ) to restrict the transmission . as shown in fig1 pages 10 of market data are each given individual page key ( pk ) codes , for example &# 34 ; page 203 &# 34 ; and &# 34 ; page 7677 &# 34 ;. these pages 10 are then applied to an encoder 12 for application to a video transmission line 14 . video displays 16 are shown connected to the transmission line 14 for receiving the encoded pages 10 . as shown , each of the video displays 16 has a unique display identification ( id ) code , for example 217 , 503 , 123 and 417 , as shown . fig2 shows a diagram representing the video signals transmitted over the video transmission line 14 . since these signals are in the form of television signals , it should be understood that the video signals for each page of market information is in the form of a sequence of video lines which , when scanned on the video displays , form the relevant pages . as shown in fig2 a first line 20 of the transmission includes an encoded signal flag indicating to the video displays 16 that the following information is encoded data . the exact form of the flag is unimportant since the information contained is just one bit . hence , the flag may be similar to , but not identical to , the vertical synchronizing signal indicating the beginning of each field of information . the line or lines 22 contain enable reception messages . the lines 24 following the enable reception message lines 22 , contain the various updates 1 , 2 and 3 . fig3 a shows a sample enable reception ( er ) message line 22 in detail . following a horizontal synchronizing pulse 30 , an er synchronizing signal 32 is sent indicating the ensuing transmission of enable reception messages and enabling decoders to synchronize to the transmission . the er sync . signal 32 is followed by id / pk conversion pairs 34 each of which includes one of the unique display id codes and one of the individual page key ( pk ) codes for which the display identified by the display code is authorized to receive . in the example shown , the pairs 417 / 7677 , 503 / 203 and 123 / 7677 indicate that video display 417 is authorized to receive page 7677 , display 503 , page 203 ; and display 123 , page 7677 . it should be noted that in the example , display 417 is not authorized to receive either page 203 or page 7677 . the enable reception message continues for as many lines ( each including an er sync . signal 32 ) and includes as many pairs 34 as are required to associate each of the authorized displays with one of the ( many ) subscribed to pages . the process for updating each page is performed by replacing &# 34 ; tiles &# 34 ; in the relevant page . as shown in fig4 the cross - hatched tile 36 to be updated is located by two row and two column ( or two x - y pixel pair ) coordinates . fig3 b - 3f show samples of the data enable sequences , in which , in fig3 b , the sequence for the update of page 7677 is illustrated . in particular , after a horizontal synchronizing pulse 40 , a data synchronizing signal 42 is present . this sync . signal 42 , which indicates the ensuing transmission of a data enable sequence , is followed by the individual page code 44 for the page 7677 and then the coordinates 46 of the tile 36 to be replaced which , in this example , is 4 rows by 40 columns . the actual data for this tile 36 is presented in a series of lines , corresponding to the number of rows in the tile to be updated , following the data enable sequence line . similar examples are shown for pages 203 and 7677 in fig3 c and 3d , in which in page 203 , a tile of 1 row and 11 columns is updated , and again in page 7677 , a second tile of 6 rows and 40 columns is updated . alternatively , as shown in fig3 e , the update data may appear on the same line as the data enable sequence . in particular , the sync . signal 42 &# 39 ; is followed by the individual page code 44 . however , the coordinates 46 &# 39 ; include the pixel start number and the pixel stop number of a single row of the update data , along with the line number of the particular line . the update data 48 then follows on the same line . each tile 36 is then composed of the update data 48 appearing in , for example , a plurality of consecutive lines . further , as shown in fig3 f , the update data may be presented simultaneously on one line for more than one page at a time . in particular , the sync . signal 42 &# 34 ; is followed by two page codes 44 &# 34 ; , and then the coordinates 46 of the tile 36 to be replaced , which in this example , is 5 rows by 80 columns , toward the bottom of pages 203 and 7677 . additionally , all authorized displays connected to the video transmission may be simultaneously updated at the same coordinates by using a special page key , e . g . pk = 0 . an encoder for the first embodiment of the invention is shown in fig5 . the encoder includes a modem 50 for receiving data from a source of market information . this data may be in the form of entire pages of financial information where portions are updated , or the update data itself along with information for the positioning of the update data on the respective display screen . the output of modem 50 is connected to an interface 51 , which is , in turn , connected to the input of a microcomputer 52 . the microcomputer 52 reassigns the data to appropriate locations in new pages for clients of the provider . a keyboard 53 is connected to the microcomputer 52 for controlling the microcomputer . a memory 54 is connected to the microcomputer 52 and supplies thereto the configuration of the new pages , the individual page key ( pk ) code for each of the new pages , and the display ( id ) codes of the clients authorized to receive each of the new pages . based on this information , the microcomputer 52 generates the first data stream and the sequence of second data streams . the output of the microcomputer 52 is applied through an interface 55 , to a video generation unit 56 which reconfigures the output of the microcomputer into video lines . the video generation unit 56 also generates the encoded signal flag and inserts the various synchronizing signals at the beginning of each of the video lines . a clock signal generator 57 is connected to the video generation unit 56 and the microcomputer 52 for applying timing signals thereto at the line frequency . in the event that the financial information applied to the modem 50 is in the form of entire pages , a memory 58 is connected to the microcomputer 52 into which the pages are entered enabling the microcomputer 52 to compare one page with the update of the page to extract therefrom only the update data . fig6 is a block diagram of a decoder for use with the encoder of fig5 . the decoder includes a receiver 60 for receiving the data transmitted by the video generation unit 56 . the output of the receiver 60 is applied to an analog switch 61 for selective application to an output display in the event that standard non - coded signals are being transmitted . a coded signal detector 62 is coupled to the receiver 60 for receiving the encoded signal flag and for switching the analog switch 61 accordingly . an er detection gate 63 is connected to the receiver 60 for receiving the enable reception messages containing the display id / individual pk code pairs . each of the received display id codes are compared with a unique display id code stored in a rom 64 by a comparator 65 . upon each match of the display id code , the individual pk code for the respective page is stored in a memory 66 . the output of the receiver 60 is further connected to a data detection gate 67 for receiving the data enable sequences . the individual pk codes in the received data enable sequences are compared in a comparator 68 with the individual pk codes stored in the memory 66 . upon a match of one of these pk codes , the accompanying coordinates of the update data are loaded into registers 69 . an analog - to - digital converter 70 digitizes the appropriate update data at the output of the receiver 60 and applies its output to a write buffer 71 , which also receives the output of the registers 69 . the output of the write buffer 71 is applied to a picture store 72 in which the section therein corresponding to the location of the update data is updated . a synchronizing signal detector 73 is connected to the output of the receiver 60 for separating the line synchronizing signals . the output of the synchronizing signal detector 73 is applied to a timing and control signal generator 74 for generating timing signals for the analog - to - digital converter 70 , the data detection gate 67 , the er detection gate 63 and the picture store 72 . the output of the picture store 72 is applied to a digital - to - analog converter 75 controlled by the timing and control signal generator 74 . the output of the digital - to - analog converter 75 is applied through a low - pass filter 76 to another input of the analog switch 61 . in a second embodiment of the invention , the video signals representing the market information include three colors . in addition , standard television signals are included in the video signals for selective viewing of realtime television on the video displays . this transmission is necessarily synchronous to the chosen television standard . assuming that the video signals are being transmitted by cable , resulting in a usable bandwidth of approximately 24 mhz . fig7 shows a pictorial representation of the transmitted video signals . the encoded signal flag line 80 , the enable reception messages lines 81 and the data enable sequence lines 82 are transmitted during the vertical blanking interval 83 between each field of the video signal . during the active video portion of the field , in a first half of each scanning line , the television r , g and b signals 84 , each originally having a bandwidth of 4 mhz . and each time compressed by a factor of six to an expanded bandwidth of 24 mhz ., are sequentially transmitted . in the second half of each scanning line , the update data for individual pages of the market information are transmitted . while the television signals 84 are in color , the update information may be monochromatic , color or a mixture of both . in particular , as shown , the first 8 half - lines contain the update monochrome data 85 for the left half and right halves , respectively , for page 7677 . the update data 85 for page 7677 is followed by the update data 86 for page 203 . the update data 86 is presented in color as the three color signals r , g and b . the remainder of the right half of the first field is shown as being unused in this example . the left half of the second field contains the g , b and r components of the television signals 78 &# 39 ;. the right half of the second field contains the monochromatic update data for the pages 208 , 1234 , 5 , 19154 and 264 . due to the complex ordering of the update data in the first and second fields , the data enable sequences in the lines 82 must necessarily be more complex than those shown in fig3 b - 3f . in addition , the enable reception messages must indicate which of the video displays is authorized to receive the television signals sent with the update data . in particular , as shown in fig8 a , the enable reception messages are similar to those shown in fig3 a , with the exception that in addition to the display id code / identification pk code pairs , the messages include a pair 87 indicating which of the video displays , for example , the display with display id code 297 , is authorized to receive the television signals tv1 . fig8 b shows a sample data enable sequence which includes , in addition to that described with respect to fig3 b - 3f , the coordinates of the update data in the source field . fig8 c shows a sample of the data enable sequence for identifying the television signals , and includes a data synchronizing signal 88 , a television pk code 89 and the starting coordinates 90 of the three color signals -- red , green and blue . fig9 shows a block diagram of an encoder for the second embodiment . the video generation unit 56 &# 39 ; has a second set of inputs for receiving the three color components of the television signals . in particular , a source of video signals is connected to a synchronizing signal separation circuit 91 for detecting the vertical and horizontal synchronizing signals in the video signals . the source of the video signals is also connected to a matrix circuit 92 for providing the three color components . each of these components is subjected to a 6 : 1 compression in compression circuit 93 and the three components are then applied to the video generation unit 56 &# 39 ;. the clock signal generator 57 &# 39 ; applies horizontal and vertical synchronizing signals to both the video generation unit 56 &# 39 ; and the microcomputer 52 &# 39 ;, and receives the synchronizing signals from the separation circuit 91 for synchronization therewith . fig1 shows a block diagram of a decoder for the second embodiment . components the same as those in fig6 are designated with the same reference number . the decoder is substantially similar as the decoder of the first embodiment with the exception that the decoder is now capable of processing color signals and the encoded data selectively includes television signals . in particular , a color decoder 101 is included between the output of the receiver 60 and the input of the analog switch 61 . 1 - 61 . 3 . the register 69 &# 39 ; includes a register element for storing the number of the picture store . the synchronizing signal detector 73 &# 39 ; outputs field synchronizing signals in addition to line synchronizing signals . the write buffer 71 &# 39 ; now accesses three picture stores 72 . 1 - 72 . 3 corresponding to the three color components , red , blue and green . the outputs of these picture stores 72 . 1 - 72 . 3 are applied to three digital - to - analog converters 75 . 1 - 75 . 3 , and then to three low - pass filters 76 . 1 - 76 . 3 for application to the other inputs of the three analog switches 61 . 1 - 61 . 3 . numerous alterations of the structure herein disclosed will suggest themselves to those skilled in the art . however it is to be understood that the embodiments herein disclosed are for purposes of illustration only and not to be construed as a limitation of the invention . all such modifications which do not depart from the spirit of the invention are intended to be included within the scope of the appended claims ."}
|
{"patent": "as noted above , the object of the present invention is to distribute information . this information is transmitted over , for example , telephone lines and converted on the client site into video signals similar to those for television reception . the information subscribed to takes the form of pages of market data , various portions of which are updated from time to time to reflect changes in the market , and subsequently presented on a video display . in a first embodiment of the invention , the video signals representing the market information are being transmitted asynchronously , that is , there are no set characteristic times ( e . g . television field and frame rates ) to restrict the transmission . as shown in fig1 pages 10 of market data are each given individual page key ( pk ) codes , for example &# 34 ; page 203 &# 34 ; and &# 34 ; page 7677 &# 34 ;. these pages 10 are then applied to an encoder 12 for application to a video transmission line 14 . video displays 16 are shown connected to the transmission line 14 for receiving the encoded pages 10 . as shown , each of the video displays 16 has a unique display identification ( id ) code , for example 217 , 503 , 123 and 417 , as shown . fig2 shows a diagram representing the video signals transmitted over the video transmission line 14 . since these signals are in the form of television signals , it should be understood that the video signals for each page of market information is in the form of a sequence of video lines which , when scanned on the video displays , form the relevant pages . as shown in fig2 a first line 20 of the transmission includes an encoded signal flag indicating to the video displays 16 that the following information is encoded data . the exact form of the flag is unimportant since the information contained is just one bit . hence , the flag may be similar to , but not identical to , the vertical synchronizing signal indicating the beginning of each field of information . the line or lines 22 contain enable reception messages . the lines 24 following the enable reception message lines 22 , contain the various updates 1 , 2 and 3 . fig3 a shows a sample enable reception ( er ) message line 22 in detail . following a horizontal synchronizing pulse 30 , an er synchronizing signal 32 is sent indicating the ensuing transmission of enable reception messages and enabling decoders to synchronize to the transmission . the er sync . signal 32 is followed by id / pk conversion pairs 34 each of which includes one of the unique display id codes and one of the individual page key ( pk ) codes for which the display identified by the display code is authorized to receive . in the example shown , the pairs 417 / 7677 , 503 / 203 and 123 / 7677 indicate that video display 417 is authorized to receive page 7677 , display 503 , page 203 ; and display 123 , page 7677 . it should be noted that in the example , display 417 is not authorized to receive either page 203 or page 7677 . the enable reception message continues for as many lines ( each including an er sync . signal 32 ) and includes as many pairs 34 as are required to associate each of the authorized displays with one of the ( many ) subscribed to pages . the process for updating each page is performed by replacing &# 34 ; tiles &# 34 ; in the relevant page . as shown in fig4 the cross - hatched tile 36 to be updated is located by two row and two column ( or two x - y pixel pair ) coordinates . fig3 b - 3f show samples of the data enable sequences , in which , in fig3 b , the sequence for the update of page 7677 is illustrated . in particular , after a horizontal synchronizing pulse 40 , a data synchronizing signal 42 is present . this sync . signal 42 , which indicates the ensuing transmission of a data enable sequence , is followed by the individual page code 44 for the page 7677 and then the coordinates 46 of the tile 36 to be replaced which , in this example , is 4 rows by 40 columns . the actual data for this tile 36 is presented in a series of lines , corresponding to the number of rows in the tile to be updated , following the data enable sequence line . similar examples are shown for pages 203 and 7677 in fig3 c and 3d , in which in page 203 , a tile of 1 row and 11 columns is updated , and again in page 7677 , a second tile of 6 rows and 40 columns is updated . alternatively , as shown in fig3 e , the update data may appear on the same line as the data enable sequence . in particular , the sync . signal 42 &# 39 ; is followed by the individual page code 44 . however , the coordinates 46 &# 39 ; include the pixel start number and the pixel stop number of a single row of the update data , along with the line number of the particular line . the update data 48 then follows on the same line . each tile 36 is then composed of the update data 48 appearing in , for example , a plurality of consecutive lines . further , as shown in fig3 f , the update data may be presented simultaneously on one line for more than one page at a time . in particular , the sync . signal 42 &# 34 ; is followed by two page codes 44 &# 34 ; , and then the coordinates 46 of the tile 36 to be replaced , which in this example , is 5 rows by 80 columns , toward the bottom of pages 203 and 7677 . additionally , all authorized displays connected to the video transmission may be simultaneously updated at the same coordinates by using a special page key , e . g . pk = 0 . an encoder for the first embodiment of the invention is shown in fig5 . the encoder includes a modem 50 for receiving data from a source of market information . this data may be in the form of entire pages of financial information where portions are updated , or the update data itself along with information for the positioning of the update data on the respective display screen . the output of modem 50 is connected to an interface 51 , which is , in turn , connected to the input of a microcomputer 52 . the microcomputer 52 reassigns the data to appropriate locations in new pages for clients of the provider . a keyboard 53 is connected to the microcomputer 52 for controlling the microcomputer . a memory 54 is connected to the microcomputer 52 and supplies thereto the configuration of the new pages , the individual page key ( pk ) code for each of the new pages , and the display ( id ) codes of the clients authorized to receive each of the new pages . based on this information , the microcomputer 52 generates the first data stream and the sequence of second data streams . the output of the microcomputer 52 is applied through an interface 55 , to a video generation unit 56 which reconfigures the output of the microcomputer into video lines . the video generation unit 56 also generates the encoded signal flag and inserts the various synchronizing signals at the beginning of each of the video lines . a clock signal generator 57 is connected to the video generation unit 56 and the microcomputer 52 for applying timing signals thereto at the line frequency . in the event that the financial information applied to the modem 50 is in the form of entire pages , a memory 58 is connected to the microcomputer 52 into which the pages are entered enabling the microcomputer 52 to compare one page with the update of the page to extract therefrom only the update data . fig6 is a block diagram of a decoder for use with the encoder of fig5 . the decoder includes a receiver 60 for receiving the data transmitted by the video generation unit 56 . the output of the receiver 60 is applied to an analog switch 61 for selective application to an output display in the event that standard non - coded signals are being transmitted . a coded signal detector 62 is coupled to the receiver 60 for receiving the encoded signal flag and for switching the analog switch 61 accordingly . an er detection gate 63 is connected to the receiver 60 for receiving the enable reception messages containing the display id / individual pk code pairs . each of the received display id codes are compared with a unique display id code stored in a rom 64 by a comparator 65 . upon each match of the display id code , the individual pk code for the respective page is stored in a memory 66 . the output of the receiver 60 is further connected to a data detection gate 67 for receiving the data enable sequences . the individual pk codes in the received data enable sequences are compared in a comparator 68 with the individual pk codes stored in the memory 66 . upon a match of one of these pk codes , the accompanying coordinates of the update data are loaded into registers 69 . an analog - to - digital converter 70 digitizes the appropriate update data at the output of the receiver 60 and applies its output to a write buffer 71 , which also receives the output of the registers 69 . the output of the write buffer 71 is applied to a picture store 72 in which the section therein corresponding to the location of the update data is updated . a synchronizing signal detector 73 is connected to the output of the receiver 60 for separating the line synchronizing signals . the output of the synchronizing signal detector 73 is applied to a timing and control signal generator 74 for generating timing signals for the analog - to - digital converter 70 , the data detection gate 67 , the er detection gate 63 and the picture store 72 . the output of the picture store 72 is applied to a digital - to - analog converter 75 controlled by the timing and control signal generator 74 . the output of the digital - to - analog converter 75 is applied through a low - pass filter 76 to another input of the analog switch 61 . in a second embodiment of the invention , the video signals representing the market information include three colors . in addition , standard television signals are included in the video signals for selective viewing of realtime television on the video displays . this transmission is necessarily synchronous to the chosen television standard . assuming that the video signals are being transmitted by cable , resulting in a usable bandwidth of approximately 24 mhz . fig7 shows a pictorial representation of the transmitted video signals . the encoded signal flag line 80 , the enable reception messages lines 81 and the data enable sequence lines 82 are transmitted during the vertical blanking interval 83 between each field of the video signal . during the active video portion of the field , in a first half of each scanning line , the television r , g and b signals 84 , each originally having a bandwidth of 4 mhz . and each time compressed by a factor of six to an expanded bandwidth of 24 mhz ., are sequentially transmitted . in the second half of each scanning line , the update data for individual pages of the market information are transmitted . while the television signals 84 are in color , the update information may be monochromatic , color or a mixture of both . in particular , as shown , the first 8 half - lines contain the update monochrome data 85 for the left half and right halves , respectively , for page 7677 . the update data 85 for page 7677 is followed by the update data 86 for page 203 . the update data 86 is presented in color as the three color signals r , g and b . the remainder of the right half of the first field is shown as being unused in this example . the left half of the second field contains the g , b and r components of the television signals 78 &# 39 ;. the right half of the second field contains the monochromatic update data for the pages 208 , 1234 , 5 , 19154 and 264 . due to the complex ordering of the update data in the first and second fields , the data enable sequences in the lines 82 must necessarily be more complex than those shown in fig3 b - 3f . in addition , the enable reception messages must indicate which of the video displays is authorized to receive the television signals sent with the update data . in particular , as shown in fig8 a , the enable reception messages are similar to those shown in fig3 a , with the exception that in addition to the display id code / identification pk code pairs , the messages include a pair 87 indicating which of the video displays , for example , the display with display id code 297 , is authorized to receive the television signals tv1 . fig8 b shows a sample data enable sequence which includes , in addition to that described with respect to fig3 b - 3f , the coordinates of the update data in the source field . fig8 c shows a sample of the data enable sequence for identifying the television signals , and includes a data synchronizing signal 88 , a television pk code 89 and the starting coordinates 90 of the three color signals -- red , green and blue . fig9 shows a block diagram of an encoder for the second embodiment . the video generation unit 56 &# 39 ; has a second set of inputs for receiving the three color components of the television signals . in particular , a source of video signals is connected to a synchronizing signal separation circuit 91 for detecting the vertical and horizontal synchronizing signals in the video signals . the source of the video signals is also connected to a matrix circuit 92 for providing the three color components . each of these components is subjected to a 6 : 1 compression in compression circuit 93 and the three components are then applied to the video generation unit 56 &# 39 ;. the clock signal generator 57 &# 39 ; applies horizontal and vertical synchronizing signals to both the video generation unit 56 &# 39 ; and the microcomputer 52 &# 39 ;, and receives the synchronizing signals from the separation circuit 91 for synchronization therewith . fig1 shows a block diagram of a decoder for the second embodiment . components the same as those in fig6 are designated with the same reference number . the decoder is substantially similar as the decoder of the first embodiment with the exception that the decoder is now capable of processing color signals and the encoded data selectively includes television signals . in particular , a color decoder 101 is included between the output of the receiver 60 and the input of the analog switch 61 . 1 - 61 . 3 . the register 69 &# 39 ; includes a register element for storing the number of the picture store . the synchronizing signal detector 73 &# 39 ; outputs field synchronizing signals in addition to line synchronizing signals . the write buffer 71 &# 39 ; now accesses three picture stores 72 . 1 - 72 . 3 corresponding to the three color components , red , blue and green . the outputs of these picture stores 72 . 1 - 72 . 3 are applied to three digital - to - analog converters 75 . 1 - 75 . 3 , and then to three low - pass filters 76 . 1 - 76 . 3 for application to the other inputs of the three analog switches 61 . 1 - 61 . 3 . numerous alterations of the structure herein disclosed will suggest themselves to those skilled in the art . however it is to be understood that the embodiments herein disclosed are for purposes of illustration only and not to be construed as a limitation of the invention . all such modifications which do not depart from the spirit of the invention are intended to be included within the scope of the appended claims .", "category": "Performing Operations; Transporting"}
|
Does the patent belong in this category?
| 0.25 |
8790d674f8c57a9e5539268a37bf38022b241b63e3f4f82bc8f6033d6eb96c97
| 0.609375 | 0.007813 | 0.671875 | 0.01001 | 0.851563 | 0.241211 |
null |
{"category": "Electricity", "patent": "as noted above , the object of the present invention is to distribute information . this information is transmitted over , for example , telephone lines and converted on the client site into video signals similar to those for television reception . the information subscribed to takes the form of pages of market data , various portions of which are updated from time to time to reflect changes in the market , and subsequently presented on a video display . in a first embodiment of the invention , the video signals representing the market information are being transmitted asynchronously , that is , there are no set characteristic times ( e . g . television field and frame rates ) to restrict the transmission . as shown in fig1 pages 10 of market data are each given individual page key ( pk ) codes , for example &# 34 ; page 203 &# 34 ; and &# 34 ; page 7677 &# 34 ;. these pages 10 are then applied to an encoder 12 for application to a video transmission line 14 . video displays 16 are shown connected to the transmission line 14 for receiving the encoded pages 10 . as shown , each of the video displays 16 has a unique display identification ( id ) code , for example 217 , 503 , 123 and 417 , as shown . fig2 shows a diagram representing the video signals transmitted over the video transmission line 14 . since these signals are in the form of television signals , it should be understood that the video signals for each page of market information is in the form of a sequence of video lines which , when scanned on the video displays , form the relevant pages . as shown in fig2 a first line 20 of the transmission includes an encoded signal flag indicating to the video displays 16 that the following information is encoded data . the exact form of the flag is unimportant since the information contained is just one bit . hence , the flag may be similar to , but not identical to , the vertical synchronizing signal indicating the beginning of each field of information . the line or lines 22 contain enable reception messages . the lines 24 following the enable reception message lines 22 , contain the various updates 1 , 2 and 3 . fig3 a shows a sample enable reception ( er ) message line 22 in detail . following a horizontal synchronizing pulse 30 , an er synchronizing signal 32 is sent indicating the ensuing transmission of enable reception messages and enabling decoders to synchronize to the transmission . the er sync . signal 32 is followed by id / pk conversion pairs 34 each of which includes one of the unique display id codes and one of the individual page key ( pk ) codes for which the display identified by the display code is authorized to receive . in the example shown , the pairs 417 / 7677 , 503 / 203 and 123 / 7677 indicate that video display 417 is authorized to receive page 7677 , display 503 , page 203 ; and display 123 , page 7677 . it should be noted that in the example , display 417 is not authorized to receive either page 203 or page 7677 . the enable reception message continues for as many lines ( each including an er sync . signal 32 ) and includes as many pairs 34 as are required to associate each of the authorized displays with one of the ( many ) subscribed to pages . the process for updating each page is performed by replacing &# 34 ; tiles &# 34 ; in the relevant page . as shown in fig4 the cross - hatched tile 36 to be updated is located by two row and two column ( or two x - y pixel pair ) coordinates . fig3 b - 3f show samples of the data enable sequences , in which , in fig3 b , the sequence for the update of page 7677 is illustrated . in particular , after a horizontal synchronizing pulse 40 , a data synchronizing signal 42 is present . this sync . signal 42 , which indicates the ensuing transmission of a data enable sequence , is followed by the individual page code 44 for the page 7677 and then the coordinates 46 of the tile 36 to be replaced which , in this example , is 4 rows by 40 columns . the actual data for this tile 36 is presented in a series of lines , corresponding to the number of rows in the tile to be updated , following the data enable sequence line . similar examples are shown for pages 203 and 7677 in fig3 c and 3d , in which in page 203 , a tile of 1 row and 11 columns is updated , and again in page 7677 , a second tile of 6 rows and 40 columns is updated . alternatively , as shown in fig3 e , the update data may appear on the same line as the data enable sequence . in particular , the sync . signal 42 &# 39 ; is followed by the individual page code 44 . however , the coordinates 46 &# 39 ; include the pixel start number and the pixel stop number of a single row of the update data , along with the line number of the particular line . the update data 48 then follows on the same line . each tile 36 is then composed of the update data 48 appearing in , for example , a plurality of consecutive lines . further , as shown in fig3 f , the update data may be presented simultaneously on one line for more than one page at a time . in particular , the sync . signal 42 &# 34 ; is followed by two page codes 44 &# 34 ; , and then the coordinates 46 of the tile 36 to be replaced , which in this example , is 5 rows by 80 columns , toward the bottom of pages 203 and 7677 . additionally , all authorized displays connected to the video transmission may be simultaneously updated at the same coordinates by using a special page key , e . g . pk = 0 . an encoder for the first embodiment of the invention is shown in fig5 . the encoder includes a modem 50 for receiving data from a source of market information . this data may be in the form of entire pages of financial information where portions are updated , or the update data itself along with information for the positioning of the update data on the respective display screen . the output of modem 50 is connected to an interface 51 , which is , in turn , connected to the input of a microcomputer 52 . the microcomputer 52 reassigns the data to appropriate locations in new pages for clients of the provider . a keyboard 53 is connected to the microcomputer 52 for controlling the microcomputer . a memory 54 is connected to the microcomputer 52 and supplies thereto the configuration of the new pages , the individual page key ( pk ) code for each of the new pages , and the display ( id ) codes of the clients authorized to receive each of the new pages . based on this information , the microcomputer 52 generates the first data stream and the sequence of second data streams . the output of the microcomputer 52 is applied through an interface 55 , to a video generation unit 56 which reconfigures the output of the microcomputer into video lines . the video generation unit 56 also generates the encoded signal flag and inserts the various synchronizing signals at the beginning of each of the video lines . a clock signal generator 57 is connected to the video generation unit 56 and the microcomputer 52 for applying timing signals thereto at the line frequency . in the event that the financial information applied to the modem 50 is in the form of entire pages , a memory 58 is connected to the microcomputer 52 into which the pages are entered enabling the microcomputer 52 to compare one page with the update of the page to extract therefrom only the update data . fig6 is a block diagram of a decoder for use with the encoder of fig5 . the decoder includes a receiver 60 for receiving the data transmitted by the video generation unit 56 . the output of the receiver 60 is applied to an analog switch 61 for selective application to an output display in the event that standard non - coded signals are being transmitted . a coded signal detector 62 is coupled to the receiver 60 for receiving the encoded signal flag and for switching the analog switch 61 accordingly . an er detection gate 63 is connected to the receiver 60 for receiving the enable reception messages containing the display id / individual pk code pairs . each of the received display id codes are compared with a unique display id code stored in a rom 64 by a comparator 65 . upon each match of the display id code , the individual pk code for the respective page is stored in a memory 66 . the output of the receiver 60 is further connected to a data detection gate 67 for receiving the data enable sequences . the individual pk codes in the received data enable sequences are compared in a comparator 68 with the individual pk codes stored in the memory 66 . upon a match of one of these pk codes , the accompanying coordinates of the update data are loaded into registers 69 . an analog - to - digital converter 70 digitizes the appropriate update data at the output of the receiver 60 and applies its output to a write buffer 71 , which also receives the output of the registers 69 . the output of the write buffer 71 is applied to a picture store 72 in which the section therein corresponding to the location of the update data is updated . a synchronizing signal detector 73 is connected to the output of the receiver 60 for separating the line synchronizing signals . the output of the synchronizing signal detector 73 is applied to a timing and control signal generator 74 for generating timing signals for the analog - to - digital converter 70 , the data detection gate 67 , the er detection gate 63 and the picture store 72 . the output of the picture store 72 is applied to a digital - to - analog converter 75 controlled by the timing and control signal generator 74 . the output of the digital - to - analog converter 75 is applied through a low - pass filter 76 to another input of the analog switch 61 . in a second embodiment of the invention , the video signals representing the market information include three colors . in addition , standard television signals are included in the video signals for selective viewing of realtime television on the video displays . this transmission is necessarily synchronous to the chosen television standard . assuming that the video signals are being transmitted by cable , resulting in a usable bandwidth of approximately 24 mhz . fig7 shows a pictorial representation of the transmitted video signals . the encoded signal flag line 80 , the enable reception messages lines 81 and the data enable sequence lines 82 are transmitted during the vertical blanking interval 83 between each field of the video signal . during the active video portion of the field , in a first half of each scanning line , the television r , g and b signals 84 , each originally having a bandwidth of 4 mhz . and each time compressed by a factor of six to an expanded bandwidth of 24 mhz ., are sequentially transmitted . in the second half of each scanning line , the update data for individual pages of the market information are transmitted . while the television signals 84 are in color , the update information may be monochromatic , color or a mixture of both . in particular , as shown , the first 8 half - lines contain the update monochrome data 85 for the left half and right halves , respectively , for page 7677 . the update data 85 for page 7677 is followed by the update data 86 for page 203 . the update data 86 is presented in color as the three color signals r , g and b . the remainder of the right half of the first field is shown as being unused in this example . the left half of the second field contains the g , b and r components of the television signals 78 &# 39 ;. the right half of the second field contains the monochromatic update data for the pages 208 , 1234 , 5 , 19154 and 264 . due to the complex ordering of the update data in the first and second fields , the data enable sequences in the lines 82 must necessarily be more complex than those shown in fig3 b - 3f . in addition , the enable reception messages must indicate which of the video displays is authorized to receive the television signals sent with the update data . in particular , as shown in fig8 a , the enable reception messages are similar to those shown in fig3 a , with the exception that in addition to the display id code / identification pk code pairs , the messages include a pair 87 indicating which of the video displays , for example , the display with display id code 297 , is authorized to receive the television signals tv1 . fig8 b shows a sample data enable sequence which includes , in addition to that described with respect to fig3 b - 3f , the coordinates of the update data in the source field . fig8 c shows a sample of the data enable sequence for identifying the television signals , and includes a data synchronizing signal 88 , a television pk code 89 and the starting coordinates 90 of the three color signals -- red , green and blue . fig9 shows a block diagram of an encoder for the second embodiment . the video generation unit 56 &# 39 ; has a second set of inputs for receiving the three color components of the television signals . in particular , a source of video signals is connected to a synchronizing signal separation circuit 91 for detecting the vertical and horizontal synchronizing signals in the video signals . the source of the video signals is also connected to a matrix circuit 92 for providing the three color components . each of these components is subjected to a 6 : 1 compression in compression circuit 93 and the three components are then applied to the video generation unit 56 &# 39 ;. the clock signal generator 57 &# 39 ; applies horizontal and vertical synchronizing signals to both the video generation unit 56 &# 39 ; and the microcomputer 52 &# 39 ;, and receives the synchronizing signals from the separation circuit 91 for synchronization therewith . fig1 shows a block diagram of a decoder for the second embodiment . components the same as those in fig6 are designated with the same reference number . the decoder is substantially similar as the decoder of the first embodiment with the exception that the decoder is now capable of processing color signals and the encoded data selectively includes television signals . in particular , a color decoder 101 is included between the output of the receiver 60 and the input of the analog switch 61 . 1 - 61 . 3 . the register 69 &# 39 ; includes a register element for storing the number of the picture store . the synchronizing signal detector 73 &# 39 ; outputs field synchronizing signals in addition to line synchronizing signals . the write buffer 71 &# 39 ; now accesses three picture stores 72 . 1 - 72 . 3 corresponding to the three color components , red , blue and green . the outputs of these picture stores 72 . 1 - 72 . 3 are applied to three digital - to - analog converters 75 . 1 - 75 . 3 , and then to three low - pass filters 76 . 1 - 76 . 3 for application to the other inputs of the three analog switches 61 . 1 - 61 . 3 . numerous alterations of the structure herein disclosed will suggest themselves to those skilled in the art . however it is to be understood that the embodiments herein disclosed are for purposes of illustration only and not to be construed as a limitation of the invention . all such modifications which do not depart from the spirit of the invention are intended to be included within the scope of the appended claims ."}
|
{"category": "Chemistry; Metallurgy", "patent": "as noted above , the object of the present invention is to distribute information . this information is transmitted over , for example , telephone lines and converted on the client site into video signals similar to those for television reception . the information subscribed to takes the form of pages of market data , various portions of which are updated from time to time to reflect changes in the market , and subsequently presented on a video display . in a first embodiment of the invention , the video signals representing the market information are being transmitted asynchronously , that is , there are no set characteristic times ( e . g . television field and frame rates ) to restrict the transmission . as shown in fig1 pages 10 of market data are each given individual page key ( pk ) codes , for example &# 34 ; page 203 &# 34 ; and &# 34 ; page 7677 &# 34 ;. these pages 10 are then applied to an encoder 12 for application to a video transmission line 14 . video displays 16 are shown connected to the transmission line 14 for receiving the encoded pages 10 . as shown , each of the video displays 16 has a unique display identification ( id ) code , for example 217 , 503 , 123 and 417 , as shown . fig2 shows a diagram representing the video signals transmitted over the video transmission line 14 . since these signals are in the form of television signals , it should be understood that the video signals for each page of market information is in the form of a sequence of video lines which , when scanned on the video displays , form the relevant pages . as shown in fig2 a first line 20 of the transmission includes an encoded signal flag indicating to the video displays 16 that the following information is encoded data . the exact form of the flag is unimportant since the information contained is just one bit . hence , the flag may be similar to , but not identical to , the vertical synchronizing signal indicating the beginning of each field of information . the line or lines 22 contain enable reception messages . the lines 24 following the enable reception message lines 22 , contain the various updates 1 , 2 and 3 . fig3 a shows a sample enable reception ( er ) message line 22 in detail . following a horizontal synchronizing pulse 30 , an er synchronizing signal 32 is sent indicating the ensuing transmission of enable reception messages and enabling decoders to synchronize to the transmission . the er sync . signal 32 is followed by id / pk conversion pairs 34 each of which includes one of the unique display id codes and one of the individual page key ( pk ) codes for which the display identified by the display code is authorized to receive . in the example shown , the pairs 417 / 7677 , 503 / 203 and 123 / 7677 indicate that video display 417 is authorized to receive page 7677 , display 503 , page 203 ; and display 123 , page 7677 . it should be noted that in the example , display 417 is not authorized to receive either page 203 or page 7677 . the enable reception message continues for as many lines ( each including an er sync . signal 32 ) and includes as many pairs 34 as are required to associate each of the authorized displays with one of the ( many ) subscribed to pages . the process for updating each page is performed by replacing &# 34 ; tiles &# 34 ; in the relevant page . as shown in fig4 the cross - hatched tile 36 to be updated is located by two row and two column ( or two x - y pixel pair ) coordinates . fig3 b - 3f show samples of the data enable sequences , in which , in fig3 b , the sequence for the update of page 7677 is illustrated . in particular , after a horizontal synchronizing pulse 40 , a data synchronizing signal 42 is present . this sync . signal 42 , which indicates the ensuing transmission of a data enable sequence , is followed by the individual page code 44 for the page 7677 and then the coordinates 46 of the tile 36 to be replaced which , in this example , is 4 rows by 40 columns . the actual data for this tile 36 is presented in a series of lines , corresponding to the number of rows in the tile to be updated , following the data enable sequence line . similar examples are shown for pages 203 and 7677 in fig3 c and 3d , in which in page 203 , a tile of 1 row and 11 columns is updated , and again in page 7677 , a second tile of 6 rows and 40 columns is updated . alternatively , as shown in fig3 e , the update data may appear on the same line as the data enable sequence . in particular , the sync . signal 42 &# 39 ; is followed by the individual page code 44 . however , the coordinates 46 &# 39 ; include the pixel start number and the pixel stop number of a single row of the update data , along with the line number of the particular line . the update data 48 then follows on the same line . each tile 36 is then composed of the update data 48 appearing in , for example , a plurality of consecutive lines . further , as shown in fig3 f , the update data may be presented simultaneously on one line for more than one page at a time . in particular , the sync . signal 42 &# 34 ; is followed by two page codes 44 &# 34 ; , and then the coordinates 46 of the tile 36 to be replaced , which in this example , is 5 rows by 80 columns , toward the bottom of pages 203 and 7677 . additionally , all authorized displays connected to the video transmission may be simultaneously updated at the same coordinates by using a special page key , e . g . pk = 0 . an encoder for the first embodiment of the invention is shown in fig5 . the encoder includes a modem 50 for receiving data from a source of market information . this data may be in the form of entire pages of financial information where portions are updated , or the update data itself along with information for the positioning of the update data on the respective display screen . the output of modem 50 is connected to an interface 51 , which is , in turn , connected to the input of a microcomputer 52 . the microcomputer 52 reassigns the data to appropriate locations in new pages for clients of the provider . a keyboard 53 is connected to the microcomputer 52 for controlling the microcomputer . a memory 54 is connected to the microcomputer 52 and supplies thereto the configuration of the new pages , the individual page key ( pk ) code for each of the new pages , and the display ( id ) codes of the clients authorized to receive each of the new pages . based on this information , the microcomputer 52 generates the first data stream and the sequence of second data streams . the output of the microcomputer 52 is applied through an interface 55 , to a video generation unit 56 which reconfigures the output of the microcomputer into video lines . the video generation unit 56 also generates the encoded signal flag and inserts the various synchronizing signals at the beginning of each of the video lines . a clock signal generator 57 is connected to the video generation unit 56 and the microcomputer 52 for applying timing signals thereto at the line frequency . in the event that the financial information applied to the modem 50 is in the form of entire pages , a memory 58 is connected to the microcomputer 52 into which the pages are entered enabling the microcomputer 52 to compare one page with the update of the page to extract therefrom only the update data . fig6 is a block diagram of a decoder for use with the encoder of fig5 . the decoder includes a receiver 60 for receiving the data transmitted by the video generation unit 56 . the output of the receiver 60 is applied to an analog switch 61 for selective application to an output display in the event that standard non - coded signals are being transmitted . a coded signal detector 62 is coupled to the receiver 60 for receiving the encoded signal flag and for switching the analog switch 61 accordingly . an er detection gate 63 is connected to the receiver 60 for receiving the enable reception messages containing the display id / individual pk code pairs . each of the received display id codes are compared with a unique display id code stored in a rom 64 by a comparator 65 . upon each match of the display id code , the individual pk code for the respective page is stored in a memory 66 . the output of the receiver 60 is further connected to a data detection gate 67 for receiving the data enable sequences . the individual pk codes in the received data enable sequences are compared in a comparator 68 with the individual pk codes stored in the memory 66 . upon a match of one of these pk codes , the accompanying coordinates of the update data are loaded into registers 69 . an analog - to - digital converter 70 digitizes the appropriate update data at the output of the receiver 60 and applies its output to a write buffer 71 , which also receives the output of the registers 69 . the output of the write buffer 71 is applied to a picture store 72 in which the section therein corresponding to the location of the update data is updated . a synchronizing signal detector 73 is connected to the output of the receiver 60 for separating the line synchronizing signals . the output of the synchronizing signal detector 73 is applied to a timing and control signal generator 74 for generating timing signals for the analog - to - digital converter 70 , the data detection gate 67 , the er detection gate 63 and the picture store 72 . the output of the picture store 72 is applied to a digital - to - analog converter 75 controlled by the timing and control signal generator 74 . the output of the digital - to - analog converter 75 is applied through a low - pass filter 76 to another input of the analog switch 61 . in a second embodiment of the invention , the video signals representing the market information include three colors . in addition , standard television signals are included in the video signals for selective viewing of realtime television on the video displays . this transmission is necessarily synchronous to the chosen television standard . assuming that the video signals are being transmitted by cable , resulting in a usable bandwidth of approximately 24 mhz . fig7 shows a pictorial representation of the transmitted video signals . the encoded signal flag line 80 , the enable reception messages lines 81 and the data enable sequence lines 82 are transmitted during the vertical blanking interval 83 between each field of the video signal . during the active video portion of the field , in a first half of each scanning line , the television r , g and b signals 84 , each originally having a bandwidth of 4 mhz . and each time compressed by a factor of six to an expanded bandwidth of 24 mhz ., are sequentially transmitted . in the second half of each scanning line , the update data for individual pages of the market information are transmitted . while the television signals 84 are in color , the update information may be monochromatic , color or a mixture of both . in particular , as shown , the first 8 half - lines contain the update monochrome data 85 for the left half and right halves , respectively , for page 7677 . the update data 85 for page 7677 is followed by the update data 86 for page 203 . the update data 86 is presented in color as the three color signals r , g and b . the remainder of the right half of the first field is shown as being unused in this example . the left half of the second field contains the g , b and r components of the television signals 78 &# 39 ;. the right half of the second field contains the monochromatic update data for the pages 208 , 1234 , 5 , 19154 and 264 . due to the complex ordering of the update data in the first and second fields , the data enable sequences in the lines 82 must necessarily be more complex than those shown in fig3 b - 3f . in addition , the enable reception messages must indicate which of the video displays is authorized to receive the television signals sent with the update data . in particular , as shown in fig8 a , the enable reception messages are similar to those shown in fig3 a , with the exception that in addition to the display id code / identification pk code pairs , the messages include a pair 87 indicating which of the video displays , for example , the display with display id code 297 , is authorized to receive the television signals tv1 . fig8 b shows a sample data enable sequence which includes , in addition to that described with respect to fig3 b - 3f , the coordinates of the update data in the source field . fig8 c shows a sample of the data enable sequence for identifying the television signals , and includes a data synchronizing signal 88 , a television pk code 89 and the starting coordinates 90 of the three color signals -- red , green and blue . fig9 shows a block diagram of an encoder for the second embodiment . the video generation unit 56 &# 39 ; has a second set of inputs for receiving the three color components of the television signals . in particular , a source of video signals is connected to a synchronizing signal separation circuit 91 for detecting the vertical and horizontal synchronizing signals in the video signals . the source of the video signals is also connected to a matrix circuit 92 for providing the three color components . each of these components is subjected to a 6 : 1 compression in compression circuit 93 and the three components are then applied to the video generation unit 56 &# 39 ;. the clock signal generator 57 &# 39 ; applies horizontal and vertical synchronizing signals to both the video generation unit 56 &# 39 ; and the microcomputer 52 &# 39 ;, and receives the synchronizing signals from the separation circuit 91 for synchronization therewith . fig1 shows a block diagram of a decoder for the second embodiment . components the same as those in fig6 are designated with the same reference number . the decoder is substantially similar as the decoder of the first embodiment with the exception that the decoder is now capable of processing color signals and the encoded data selectively includes television signals . in particular , a color decoder 101 is included between the output of the receiver 60 and the input of the analog switch 61 . 1 - 61 . 3 . the register 69 &# 39 ; includes a register element for storing the number of the picture store . the synchronizing signal detector 73 &# 39 ; outputs field synchronizing signals in addition to line synchronizing signals . the write buffer 71 &# 39 ; now accesses three picture stores 72 . 1 - 72 . 3 corresponding to the three color components , red , blue and green . the outputs of these picture stores 72 . 1 - 72 . 3 are applied to three digital - to - analog converters 75 . 1 - 75 . 3 , and then to three low - pass filters 76 . 1 - 76 . 3 for application to the other inputs of the three analog switches 61 . 1 - 61 . 3 . numerous alterations of the structure herein disclosed will suggest themselves to those skilled in the art . however it is to be understood that the embodiments herein disclosed are for purposes of illustration only and not to be construed as a limitation of the invention . all such modifications which do not depart from the spirit of the invention are intended to be included within the scope of the appended claims ."}
|
Is the categorization of this patent accurate?
| 0.25 |
8790d674f8c57a9e5539268a37bf38022b241b63e3f4f82bc8f6033d6eb96c97
| 0.429688 | 0.000999 | 0.289063 | 0.000828 | 0.730469 | 0.000969 |
null |
{"category": "Electricity", "patent": "as noted above , the object of the present invention is to distribute information . this information is transmitted over , for example , telephone lines and converted on the client site into video signals similar to those for television reception . the information subscribed to takes the form of pages of market data , various portions of which are updated from time to time to reflect changes in the market , and subsequently presented on a video display . in a first embodiment of the invention , the video signals representing the market information are being transmitted asynchronously , that is , there are no set characteristic times ( e . g . television field and frame rates ) to restrict the transmission . as shown in fig1 pages 10 of market data are each given individual page key ( pk ) codes , for example &# 34 ; page 203 &# 34 ; and &# 34 ; page 7677 &# 34 ;. these pages 10 are then applied to an encoder 12 for application to a video transmission line 14 . video displays 16 are shown connected to the transmission line 14 for receiving the encoded pages 10 . as shown , each of the video displays 16 has a unique display identification ( id ) code , for example 217 , 503 , 123 and 417 , as shown . fig2 shows a diagram representing the video signals transmitted over the video transmission line 14 . since these signals are in the form of television signals , it should be understood that the video signals for each page of market information is in the form of a sequence of video lines which , when scanned on the video displays , form the relevant pages . as shown in fig2 a first line 20 of the transmission includes an encoded signal flag indicating to the video displays 16 that the following information is encoded data . the exact form of the flag is unimportant since the information contained is just one bit . hence , the flag may be similar to , but not identical to , the vertical synchronizing signal indicating the beginning of each field of information . the line or lines 22 contain enable reception messages . the lines 24 following the enable reception message lines 22 , contain the various updates 1 , 2 and 3 . fig3 a shows a sample enable reception ( er ) message line 22 in detail . following a horizontal synchronizing pulse 30 , an er synchronizing signal 32 is sent indicating the ensuing transmission of enable reception messages and enabling decoders to synchronize to the transmission . the er sync . signal 32 is followed by id / pk conversion pairs 34 each of which includes one of the unique display id codes and one of the individual page key ( pk ) codes for which the display identified by the display code is authorized to receive . in the example shown , the pairs 417 / 7677 , 503 / 203 and 123 / 7677 indicate that video display 417 is authorized to receive page 7677 , display 503 , page 203 ; and display 123 , page 7677 . it should be noted that in the example , display 417 is not authorized to receive either page 203 or page 7677 . the enable reception message continues for as many lines ( each including an er sync . signal 32 ) and includes as many pairs 34 as are required to associate each of the authorized displays with one of the ( many ) subscribed to pages . the process for updating each page is performed by replacing &# 34 ; tiles &# 34 ; in the relevant page . as shown in fig4 the cross - hatched tile 36 to be updated is located by two row and two column ( or two x - y pixel pair ) coordinates . fig3 b - 3f show samples of the data enable sequences , in which , in fig3 b , the sequence for the update of page 7677 is illustrated . in particular , after a horizontal synchronizing pulse 40 , a data synchronizing signal 42 is present . this sync . signal 42 , which indicates the ensuing transmission of a data enable sequence , is followed by the individual page code 44 for the page 7677 and then the coordinates 46 of the tile 36 to be replaced which , in this example , is 4 rows by 40 columns . the actual data for this tile 36 is presented in a series of lines , corresponding to the number of rows in the tile to be updated , following the data enable sequence line . similar examples are shown for pages 203 and 7677 in fig3 c and 3d , in which in page 203 , a tile of 1 row and 11 columns is updated , and again in page 7677 , a second tile of 6 rows and 40 columns is updated . alternatively , as shown in fig3 e , the update data may appear on the same line as the data enable sequence . in particular , the sync . signal 42 &# 39 ; is followed by the individual page code 44 . however , the coordinates 46 &# 39 ; include the pixel start number and the pixel stop number of a single row of the update data , along with the line number of the particular line . the update data 48 then follows on the same line . each tile 36 is then composed of the update data 48 appearing in , for example , a plurality of consecutive lines . further , as shown in fig3 f , the update data may be presented simultaneously on one line for more than one page at a time . in particular , the sync . signal 42 &# 34 ; is followed by two page codes 44 &# 34 ; , and then the coordinates 46 of the tile 36 to be replaced , which in this example , is 5 rows by 80 columns , toward the bottom of pages 203 and 7677 . additionally , all authorized displays connected to the video transmission may be simultaneously updated at the same coordinates by using a special page key , e . g . pk = 0 . an encoder for the first embodiment of the invention is shown in fig5 . the encoder includes a modem 50 for receiving data from a source of market information . this data may be in the form of entire pages of financial information where portions are updated , or the update data itself along with information for the positioning of the update data on the respective display screen . the output of modem 50 is connected to an interface 51 , which is , in turn , connected to the input of a microcomputer 52 . the microcomputer 52 reassigns the data to appropriate locations in new pages for clients of the provider . a keyboard 53 is connected to the microcomputer 52 for controlling the microcomputer . a memory 54 is connected to the microcomputer 52 and supplies thereto the configuration of the new pages , the individual page key ( pk ) code for each of the new pages , and the display ( id ) codes of the clients authorized to receive each of the new pages . based on this information , the microcomputer 52 generates the first data stream and the sequence of second data streams . the output of the microcomputer 52 is applied through an interface 55 , to a video generation unit 56 which reconfigures the output of the microcomputer into video lines . the video generation unit 56 also generates the encoded signal flag and inserts the various synchronizing signals at the beginning of each of the video lines . a clock signal generator 57 is connected to the video generation unit 56 and the microcomputer 52 for applying timing signals thereto at the line frequency . in the event that the financial information applied to the modem 50 is in the form of entire pages , a memory 58 is connected to the microcomputer 52 into which the pages are entered enabling the microcomputer 52 to compare one page with the update of the page to extract therefrom only the update data . fig6 is a block diagram of a decoder for use with the encoder of fig5 . the decoder includes a receiver 60 for receiving the data transmitted by the video generation unit 56 . the output of the receiver 60 is applied to an analog switch 61 for selective application to an output display in the event that standard non - coded signals are being transmitted . a coded signal detector 62 is coupled to the receiver 60 for receiving the encoded signal flag and for switching the analog switch 61 accordingly . an er detection gate 63 is connected to the receiver 60 for receiving the enable reception messages containing the display id / individual pk code pairs . each of the received display id codes are compared with a unique display id code stored in a rom 64 by a comparator 65 . upon each match of the display id code , the individual pk code for the respective page is stored in a memory 66 . the output of the receiver 60 is further connected to a data detection gate 67 for receiving the data enable sequences . the individual pk codes in the received data enable sequences are compared in a comparator 68 with the individual pk codes stored in the memory 66 . upon a match of one of these pk codes , the accompanying coordinates of the update data are loaded into registers 69 . an analog - to - digital converter 70 digitizes the appropriate update data at the output of the receiver 60 and applies its output to a write buffer 71 , which also receives the output of the registers 69 . the output of the write buffer 71 is applied to a picture store 72 in which the section therein corresponding to the location of the update data is updated . a synchronizing signal detector 73 is connected to the output of the receiver 60 for separating the line synchronizing signals . the output of the synchronizing signal detector 73 is applied to a timing and control signal generator 74 for generating timing signals for the analog - to - digital converter 70 , the data detection gate 67 , the er detection gate 63 and the picture store 72 . the output of the picture store 72 is applied to a digital - to - analog converter 75 controlled by the timing and control signal generator 74 . the output of the digital - to - analog converter 75 is applied through a low - pass filter 76 to another input of the analog switch 61 . in a second embodiment of the invention , the video signals representing the market information include three colors . in addition , standard television signals are included in the video signals for selective viewing of realtime television on the video displays . this transmission is necessarily synchronous to the chosen television standard . assuming that the video signals are being transmitted by cable , resulting in a usable bandwidth of approximately 24 mhz . fig7 shows a pictorial representation of the transmitted video signals . the encoded signal flag line 80 , the enable reception messages lines 81 and the data enable sequence lines 82 are transmitted during the vertical blanking interval 83 between each field of the video signal . during the active video portion of the field , in a first half of each scanning line , the television r , g and b signals 84 , each originally having a bandwidth of 4 mhz . and each time compressed by a factor of six to an expanded bandwidth of 24 mhz ., are sequentially transmitted . in the second half of each scanning line , the update data for individual pages of the market information are transmitted . while the television signals 84 are in color , the update information may be monochromatic , color or a mixture of both . in particular , as shown , the first 8 half - lines contain the update monochrome data 85 for the left half and right halves , respectively , for page 7677 . the update data 85 for page 7677 is followed by the update data 86 for page 203 . the update data 86 is presented in color as the three color signals r , g and b . the remainder of the right half of the first field is shown as being unused in this example . the left half of the second field contains the g , b and r components of the television signals 78 &# 39 ;. the right half of the second field contains the monochromatic update data for the pages 208 , 1234 , 5 , 19154 and 264 . due to the complex ordering of the update data in the first and second fields , the data enable sequences in the lines 82 must necessarily be more complex than those shown in fig3 b - 3f . in addition , the enable reception messages must indicate which of the video displays is authorized to receive the television signals sent with the update data . in particular , as shown in fig8 a , the enable reception messages are similar to those shown in fig3 a , with the exception that in addition to the display id code / identification pk code pairs , the messages include a pair 87 indicating which of the video displays , for example , the display with display id code 297 , is authorized to receive the television signals tv1 . fig8 b shows a sample data enable sequence which includes , in addition to that described with respect to fig3 b - 3f , the coordinates of the update data in the source field . fig8 c shows a sample of the data enable sequence for identifying the television signals , and includes a data synchronizing signal 88 , a television pk code 89 and the starting coordinates 90 of the three color signals -- red , green and blue . fig9 shows a block diagram of an encoder for the second embodiment . the video generation unit 56 &# 39 ; has a second set of inputs for receiving the three color components of the television signals . in particular , a source of video signals is connected to a synchronizing signal separation circuit 91 for detecting the vertical and horizontal synchronizing signals in the video signals . the source of the video signals is also connected to a matrix circuit 92 for providing the three color components . each of these components is subjected to a 6 : 1 compression in compression circuit 93 and the three components are then applied to the video generation unit 56 &# 39 ;. the clock signal generator 57 &# 39 ; applies horizontal and vertical synchronizing signals to both the video generation unit 56 &# 39 ; and the microcomputer 52 &# 39 ;, and receives the synchronizing signals from the separation circuit 91 for synchronization therewith . fig1 shows a block diagram of a decoder for the second embodiment . components the same as those in fig6 are designated with the same reference number . the decoder is substantially similar as the decoder of the first embodiment with the exception that the decoder is now capable of processing color signals and the encoded data selectively includes television signals . in particular , a color decoder 101 is included between the output of the receiver 60 and the input of the analog switch 61 . 1 - 61 . 3 . the register 69 &# 39 ; includes a register element for storing the number of the picture store . the synchronizing signal detector 73 &# 39 ; outputs field synchronizing signals in addition to line synchronizing signals . the write buffer 71 &# 39 ; now accesses three picture stores 72 . 1 - 72 . 3 corresponding to the three color components , red , blue and green . the outputs of these picture stores 72 . 1 - 72 . 3 are applied to three digital - to - analog converters 75 . 1 - 75 . 3 , and then to three low - pass filters 76 . 1 - 76 . 3 for application to the other inputs of the three analog switches 61 . 1 - 61 . 3 . numerous alterations of the structure herein disclosed will suggest themselves to those skilled in the art . however it is to be understood that the embodiments herein disclosed are for purposes of illustration only and not to be construed as a limitation of the invention . all such modifications which do not depart from the spirit of the invention are intended to be included within the scope of the appended claims ."}
|
{"patent": "as noted above , the object of the present invention is to distribute information . this information is transmitted over , for example , telephone lines and converted on the client site into video signals similar to those for television reception . the information subscribed to takes the form of pages of market data , various portions of which are updated from time to time to reflect changes in the market , and subsequently presented on a video display . in a first embodiment of the invention , the video signals representing the market information are being transmitted asynchronously , that is , there are no set characteristic times ( e . g . television field and frame rates ) to restrict the transmission . as shown in fig1 pages 10 of market data are each given individual page key ( pk ) codes , for example &# 34 ; page 203 &# 34 ; and &# 34 ; page 7677 &# 34 ;. these pages 10 are then applied to an encoder 12 for application to a video transmission line 14 . video displays 16 are shown connected to the transmission line 14 for receiving the encoded pages 10 . as shown , each of the video displays 16 has a unique display identification ( id ) code , for example 217 , 503 , 123 and 417 , as shown . fig2 shows a diagram representing the video signals transmitted over the video transmission line 14 . since these signals are in the form of television signals , it should be understood that the video signals for each page of market information is in the form of a sequence of video lines which , when scanned on the video displays , form the relevant pages . as shown in fig2 a first line 20 of the transmission includes an encoded signal flag indicating to the video displays 16 that the following information is encoded data . the exact form of the flag is unimportant since the information contained is just one bit . hence , the flag may be similar to , but not identical to , the vertical synchronizing signal indicating the beginning of each field of information . the line or lines 22 contain enable reception messages . the lines 24 following the enable reception message lines 22 , contain the various updates 1 , 2 and 3 . fig3 a shows a sample enable reception ( er ) message line 22 in detail . following a horizontal synchronizing pulse 30 , an er synchronizing signal 32 is sent indicating the ensuing transmission of enable reception messages and enabling decoders to synchronize to the transmission . the er sync . signal 32 is followed by id / pk conversion pairs 34 each of which includes one of the unique display id codes and one of the individual page key ( pk ) codes for which the display identified by the display code is authorized to receive . in the example shown , the pairs 417 / 7677 , 503 / 203 and 123 / 7677 indicate that video display 417 is authorized to receive page 7677 , display 503 , page 203 ; and display 123 , page 7677 . it should be noted that in the example , display 417 is not authorized to receive either page 203 or page 7677 . the enable reception message continues for as many lines ( each including an er sync . signal 32 ) and includes as many pairs 34 as are required to associate each of the authorized displays with one of the ( many ) subscribed to pages . the process for updating each page is performed by replacing &# 34 ; tiles &# 34 ; in the relevant page . as shown in fig4 the cross - hatched tile 36 to be updated is located by two row and two column ( or two x - y pixel pair ) coordinates . fig3 b - 3f show samples of the data enable sequences , in which , in fig3 b , the sequence for the update of page 7677 is illustrated . in particular , after a horizontal synchronizing pulse 40 , a data synchronizing signal 42 is present . this sync . signal 42 , which indicates the ensuing transmission of a data enable sequence , is followed by the individual page code 44 for the page 7677 and then the coordinates 46 of the tile 36 to be replaced which , in this example , is 4 rows by 40 columns . the actual data for this tile 36 is presented in a series of lines , corresponding to the number of rows in the tile to be updated , following the data enable sequence line . similar examples are shown for pages 203 and 7677 in fig3 c and 3d , in which in page 203 , a tile of 1 row and 11 columns is updated , and again in page 7677 , a second tile of 6 rows and 40 columns is updated . alternatively , as shown in fig3 e , the update data may appear on the same line as the data enable sequence . in particular , the sync . signal 42 &# 39 ; is followed by the individual page code 44 . however , the coordinates 46 &# 39 ; include the pixel start number and the pixel stop number of a single row of the update data , along with the line number of the particular line . the update data 48 then follows on the same line . each tile 36 is then composed of the update data 48 appearing in , for example , a plurality of consecutive lines . further , as shown in fig3 f , the update data may be presented simultaneously on one line for more than one page at a time . in particular , the sync . signal 42 &# 34 ; is followed by two page codes 44 &# 34 ; , and then the coordinates 46 of the tile 36 to be replaced , which in this example , is 5 rows by 80 columns , toward the bottom of pages 203 and 7677 . additionally , all authorized displays connected to the video transmission may be simultaneously updated at the same coordinates by using a special page key , e . g . pk = 0 . an encoder for the first embodiment of the invention is shown in fig5 . the encoder includes a modem 50 for receiving data from a source of market information . this data may be in the form of entire pages of financial information where portions are updated , or the update data itself along with information for the positioning of the update data on the respective display screen . the output of modem 50 is connected to an interface 51 , which is , in turn , connected to the input of a microcomputer 52 . the microcomputer 52 reassigns the data to appropriate locations in new pages for clients of the provider . a keyboard 53 is connected to the microcomputer 52 for controlling the microcomputer . a memory 54 is connected to the microcomputer 52 and supplies thereto the configuration of the new pages , the individual page key ( pk ) code for each of the new pages , and the display ( id ) codes of the clients authorized to receive each of the new pages . based on this information , the microcomputer 52 generates the first data stream and the sequence of second data streams . the output of the microcomputer 52 is applied through an interface 55 , to a video generation unit 56 which reconfigures the output of the microcomputer into video lines . the video generation unit 56 also generates the encoded signal flag and inserts the various synchronizing signals at the beginning of each of the video lines . a clock signal generator 57 is connected to the video generation unit 56 and the microcomputer 52 for applying timing signals thereto at the line frequency . in the event that the financial information applied to the modem 50 is in the form of entire pages , a memory 58 is connected to the microcomputer 52 into which the pages are entered enabling the microcomputer 52 to compare one page with the update of the page to extract therefrom only the update data . fig6 is a block diagram of a decoder for use with the encoder of fig5 . the decoder includes a receiver 60 for receiving the data transmitted by the video generation unit 56 . the output of the receiver 60 is applied to an analog switch 61 for selective application to an output display in the event that standard non - coded signals are being transmitted . a coded signal detector 62 is coupled to the receiver 60 for receiving the encoded signal flag and for switching the analog switch 61 accordingly . an er detection gate 63 is connected to the receiver 60 for receiving the enable reception messages containing the display id / individual pk code pairs . each of the received display id codes are compared with a unique display id code stored in a rom 64 by a comparator 65 . upon each match of the display id code , the individual pk code for the respective page is stored in a memory 66 . the output of the receiver 60 is further connected to a data detection gate 67 for receiving the data enable sequences . the individual pk codes in the received data enable sequences are compared in a comparator 68 with the individual pk codes stored in the memory 66 . upon a match of one of these pk codes , the accompanying coordinates of the update data are loaded into registers 69 . an analog - to - digital converter 70 digitizes the appropriate update data at the output of the receiver 60 and applies its output to a write buffer 71 , which also receives the output of the registers 69 . the output of the write buffer 71 is applied to a picture store 72 in which the section therein corresponding to the location of the update data is updated . a synchronizing signal detector 73 is connected to the output of the receiver 60 for separating the line synchronizing signals . the output of the synchronizing signal detector 73 is applied to a timing and control signal generator 74 for generating timing signals for the analog - to - digital converter 70 , the data detection gate 67 , the er detection gate 63 and the picture store 72 . the output of the picture store 72 is applied to a digital - to - analog converter 75 controlled by the timing and control signal generator 74 . the output of the digital - to - analog converter 75 is applied through a low - pass filter 76 to another input of the analog switch 61 . in a second embodiment of the invention , the video signals representing the market information include three colors . in addition , standard television signals are included in the video signals for selective viewing of realtime television on the video displays . this transmission is necessarily synchronous to the chosen television standard . assuming that the video signals are being transmitted by cable , resulting in a usable bandwidth of approximately 24 mhz . fig7 shows a pictorial representation of the transmitted video signals . the encoded signal flag line 80 , the enable reception messages lines 81 and the data enable sequence lines 82 are transmitted during the vertical blanking interval 83 between each field of the video signal . during the active video portion of the field , in a first half of each scanning line , the television r , g and b signals 84 , each originally having a bandwidth of 4 mhz . and each time compressed by a factor of six to an expanded bandwidth of 24 mhz ., are sequentially transmitted . in the second half of each scanning line , the update data for individual pages of the market information are transmitted . while the television signals 84 are in color , the update information may be monochromatic , color or a mixture of both . in particular , as shown , the first 8 half - lines contain the update monochrome data 85 for the left half and right halves , respectively , for page 7677 . the update data 85 for page 7677 is followed by the update data 86 for page 203 . the update data 86 is presented in color as the three color signals r , g and b . the remainder of the right half of the first field is shown as being unused in this example . the left half of the second field contains the g , b and r components of the television signals 78 &# 39 ;. the right half of the second field contains the monochromatic update data for the pages 208 , 1234 , 5 , 19154 and 264 . due to the complex ordering of the update data in the first and second fields , the data enable sequences in the lines 82 must necessarily be more complex than those shown in fig3 b - 3f . in addition , the enable reception messages must indicate which of the video displays is authorized to receive the television signals sent with the update data . in particular , as shown in fig8 a , the enable reception messages are similar to those shown in fig3 a , with the exception that in addition to the display id code / identification pk code pairs , the messages include a pair 87 indicating which of the video displays , for example , the display with display id code 297 , is authorized to receive the television signals tv1 . fig8 b shows a sample data enable sequence which includes , in addition to that described with respect to fig3 b - 3f , the coordinates of the update data in the source field . fig8 c shows a sample of the data enable sequence for identifying the television signals , and includes a data synchronizing signal 88 , a television pk code 89 and the starting coordinates 90 of the three color signals -- red , green and blue . fig9 shows a block diagram of an encoder for the second embodiment . the video generation unit 56 &# 39 ; has a second set of inputs for receiving the three color components of the television signals . in particular , a source of video signals is connected to a synchronizing signal separation circuit 91 for detecting the vertical and horizontal synchronizing signals in the video signals . the source of the video signals is also connected to a matrix circuit 92 for providing the three color components . each of these components is subjected to a 6 : 1 compression in compression circuit 93 and the three components are then applied to the video generation unit 56 &# 39 ;. the clock signal generator 57 &# 39 ; applies horizontal and vertical synchronizing signals to both the video generation unit 56 &# 39 ; and the microcomputer 52 &# 39 ;, and receives the synchronizing signals from the separation circuit 91 for synchronization therewith . fig1 shows a block diagram of a decoder for the second embodiment . components the same as those in fig6 are designated with the same reference number . the decoder is substantially similar as the decoder of the first embodiment with the exception that the decoder is now capable of processing color signals and the encoded data selectively includes television signals . in particular , a color decoder 101 is included between the output of the receiver 60 and the input of the analog switch 61 . 1 - 61 . 3 . the register 69 &# 39 ; includes a register element for storing the number of the picture store . the synchronizing signal detector 73 &# 39 ; outputs field synchronizing signals in addition to line synchronizing signals . the write buffer 71 &# 39 ; now accesses three picture stores 72 . 1 - 72 . 3 corresponding to the three color components , red , blue and green . the outputs of these picture stores 72 . 1 - 72 . 3 are applied to three digital - to - analog converters 75 . 1 - 75 . 3 , and then to three low - pass filters 76 . 1 - 76 . 3 for application to the other inputs of the three analog switches 61 . 1 - 61 . 3 . numerous alterations of the structure herein disclosed will suggest themselves to those skilled in the art . however it is to be understood that the embodiments herein disclosed are for purposes of illustration only and not to be construed as a limitation of the invention . all such modifications which do not depart from the spirit of the invention are intended to be included within the scope of the appended claims .", "category": "Textiles; Paper"}
|
Does the patent belong in this category?
| 0.25 |
8790d674f8c57a9e5539268a37bf38022b241b63e3f4f82bc8f6033d6eb96c97
| 0.574219 | 0.000043 | 0.652344 | 0.000645 | 0.851563 | 0.001808 |
null |
{"category": "Electricity", "patent": "as noted above , the object of the present invention is to distribute information . this information is transmitted over , for example , telephone lines and converted on the client site into video signals similar to those for television reception . the information subscribed to takes the form of pages of market data , various portions of which are updated from time to time to reflect changes in the market , and subsequently presented on a video display . in a first embodiment of the invention , the video signals representing the market information are being transmitted asynchronously , that is , there are no set characteristic times ( e . g . television field and frame rates ) to restrict the transmission . as shown in fig1 pages 10 of market data are each given individual page key ( pk ) codes , for example &# 34 ; page 203 &# 34 ; and &# 34 ; page 7677 &# 34 ;. these pages 10 are then applied to an encoder 12 for application to a video transmission line 14 . video displays 16 are shown connected to the transmission line 14 for receiving the encoded pages 10 . as shown , each of the video displays 16 has a unique display identification ( id ) code , for example 217 , 503 , 123 and 417 , as shown . fig2 shows a diagram representing the video signals transmitted over the video transmission line 14 . since these signals are in the form of television signals , it should be understood that the video signals for each page of market information is in the form of a sequence of video lines which , when scanned on the video displays , form the relevant pages . as shown in fig2 a first line 20 of the transmission includes an encoded signal flag indicating to the video displays 16 that the following information is encoded data . the exact form of the flag is unimportant since the information contained is just one bit . hence , the flag may be similar to , but not identical to , the vertical synchronizing signal indicating the beginning of each field of information . the line or lines 22 contain enable reception messages . the lines 24 following the enable reception message lines 22 , contain the various updates 1 , 2 and 3 . fig3 a shows a sample enable reception ( er ) message line 22 in detail . following a horizontal synchronizing pulse 30 , an er synchronizing signal 32 is sent indicating the ensuing transmission of enable reception messages and enabling decoders to synchronize to the transmission . the er sync . signal 32 is followed by id / pk conversion pairs 34 each of which includes one of the unique display id codes and one of the individual page key ( pk ) codes for which the display identified by the display code is authorized to receive . in the example shown , the pairs 417 / 7677 , 503 / 203 and 123 / 7677 indicate that video display 417 is authorized to receive page 7677 , display 503 , page 203 ; and display 123 , page 7677 . it should be noted that in the example , display 417 is not authorized to receive either page 203 or page 7677 . the enable reception message continues for as many lines ( each including an er sync . signal 32 ) and includes as many pairs 34 as are required to associate each of the authorized displays with one of the ( many ) subscribed to pages . the process for updating each page is performed by replacing &# 34 ; tiles &# 34 ; in the relevant page . as shown in fig4 the cross - hatched tile 36 to be updated is located by two row and two column ( or two x - y pixel pair ) coordinates . fig3 b - 3f show samples of the data enable sequences , in which , in fig3 b , the sequence for the update of page 7677 is illustrated . in particular , after a horizontal synchronizing pulse 40 , a data synchronizing signal 42 is present . this sync . signal 42 , which indicates the ensuing transmission of a data enable sequence , is followed by the individual page code 44 for the page 7677 and then the coordinates 46 of the tile 36 to be replaced which , in this example , is 4 rows by 40 columns . the actual data for this tile 36 is presented in a series of lines , corresponding to the number of rows in the tile to be updated , following the data enable sequence line . similar examples are shown for pages 203 and 7677 in fig3 c and 3d , in which in page 203 , a tile of 1 row and 11 columns is updated , and again in page 7677 , a second tile of 6 rows and 40 columns is updated . alternatively , as shown in fig3 e , the update data may appear on the same line as the data enable sequence . in particular , the sync . signal 42 &# 39 ; is followed by the individual page code 44 . however , the coordinates 46 &# 39 ; include the pixel start number and the pixel stop number of a single row of the update data , along with the line number of the particular line . the update data 48 then follows on the same line . each tile 36 is then composed of the update data 48 appearing in , for example , a plurality of consecutive lines . further , as shown in fig3 f , the update data may be presented simultaneously on one line for more than one page at a time . in particular , the sync . signal 42 &# 34 ; is followed by two page codes 44 &# 34 ; , and then the coordinates 46 of the tile 36 to be replaced , which in this example , is 5 rows by 80 columns , toward the bottom of pages 203 and 7677 . additionally , all authorized displays connected to the video transmission may be simultaneously updated at the same coordinates by using a special page key , e . g . pk = 0 . an encoder for the first embodiment of the invention is shown in fig5 . the encoder includes a modem 50 for receiving data from a source of market information . this data may be in the form of entire pages of financial information where portions are updated , or the update data itself along with information for the positioning of the update data on the respective display screen . the output of modem 50 is connected to an interface 51 , which is , in turn , connected to the input of a microcomputer 52 . the microcomputer 52 reassigns the data to appropriate locations in new pages for clients of the provider . a keyboard 53 is connected to the microcomputer 52 for controlling the microcomputer . a memory 54 is connected to the microcomputer 52 and supplies thereto the configuration of the new pages , the individual page key ( pk ) code for each of the new pages , and the display ( id ) codes of the clients authorized to receive each of the new pages . based on this information , the microcomputer 52 generates the first data stream and the sequence of second data streams . the output of the microcomputer 52 is applied through an interface 55 , to a video generation unit 56 which reconfigures the output of the microcomputer into video lines . the video generation unit 56 also generates the encoded signal flag and inserts the various synchronizing signals at the beginning of each of the video lines . a clock signal generator 57 is connected to the video generation unit 56 and the microcomputer 52 for applying timing signals thereto at the line frequency . in the event that the financial information applied to the modem 50 is in the form of entire pages , a memory 58 is connected to the microcomputer 52 into which the pages are entered enabling the microcomputer 52 to compare one page with the update of the page to extract therefrom only the update data . fig6 is a block diagram of a decoder for use with the encoder of fig5 . the decoder includes a receiver 60 for receiving the data transmitted by the video generation unit 56 . the output of the receiver 60 is applied to an analog switch 61 for selective application to an output display in the event that standard non - coded signals are being transmitted . a coded signal detector 62 is coupled to the receiver 60 for receiving the encoded signal flag and for switching the analog switch 61 accordingly . an er detection gate 63 is connected to the receiver 60 for receiving the enable reception messages containing the display id / individual pk code pairs . each of the received display id codes are compared with a unique display id code stored in a rom 64 by a comparator 65 . upon each match of the display id code , the individual pk code for the respective page is stored in a memory 66 . the output of the receiver 60 is further connected to a data detection gate 67 for receiving the data enable sequences . the individual pk codes in the received data enable sequences are compared in a comparator 68 with the individual pk codes stored in the memory 66 . upon a match of one of these pk codes , the accompanying coordinates of the update data are loaded into registers 69 . an analog - to - digital converter 70 digitizes the appropriate update data at the output of the receiver 60 and applies its output to a write buffer 71 , which also receives the output of the registers 69 . the output of the write buffer 71 is applied to a picture store 72 in which the section therein corresponding to the location of the update data is updated . a synchronizing signal detector 73 is connected to the output of the receiver 60 for separating the line synchronizing signals . the output of the synchronizing signal detector 73 is applied to a timing and control signal generator 74 for generating timing signals for the analog - to - digital converter 70 , the data detection gate 67 , the er detection gate 63 and the picture store 72 . the output of the picture store 72 is applied to a digital - to - analog converter 75 controlled by the timing and control signal generator 74 . the output of the digital - to - analog converter 75 is applied through a low - pass filter 76 to another input of the analog switch 61 . in a second embodiment of the invention , the video signals representing the market information include three colors . in addition , standard television signals are included in the video signals for selective viewing of realtime television on the video displays . this transmission is necessarily synchronous to the chosen television standard . assuming that the video signals are being transmitted by cable , resulting in a usable bandwidth of approximately 24 mhz . fig7 shows a pictorial representation of the transmitted video signals . the encoded signal flag line 80 , the enable reception messages lines 81 and the data enable sequence lines 82 are transmitted during the vertical blanking interval 83 between each field of the video signal . during the active video portion of the field , in a first half of each scanning line , the television r , g and b signals 84 , each originally having a bandwidth of 4 mhz . and each time compressed by a factor of six to an expanded bandwidth of 24 mhz ., are sequentially transmitted . in the second half of each scanning line , the update data for individual pages of the market information are transmitted . while the television signals 84 are in color , the update information may be monochromatic , color or a mixture of both . in particular , as shown , the first 8 half - lines contain the update monochrome data 85 for the left half and right halves , respectively , for page 7677 . the update data 85 for page 7677 is followed by the update data 86 for page 203 . the update data 86 is presented in color as the three color signals r , g and b . the remainder of the right half of the first field is shown as being unused in this example . the left half of the second field contains the g , b and r components of the television signals 78 &# 39 ;. the right half of the second field contains the monochromatic update data for the pages 208 , 1234 , 5 , 19154 and 264 . due to the complex ordering of the update data in the first and second fields , the data enable sequences in the lines 82 must necessarily be more complex than those shown in fig3 b - 3f . in addition , the enable reception messages must indicate which of the video displays is authorized to receive the television signals sent with the update data . in particular , as shown in fig8 a , the enable reception messages are similar to those shown in fig3 a , with the exception that in addition to the display id code / identification pk code pairs , the messages include a pair 87 indicating which of the video displays , for example , the display with display id code 297 , is authorized to receive the television signals tv1 . fig8 b shows a sample data enable sequence which includes , in addition to that described with respect to fig3 b - 3f , the coordinates of the update data in the source field . fig8 c shows a sample of the data enable sequence for identifying the television signals , and includes a data synchronizing signal 88 , a television pk code 89 and the starting coordinates 90 of the three color signals -- red , green and blue . fig9 shows a block diagram of an encoder for the second embodiment . the video generation unit 56 &# 39 ; has a second set of inputs for receiving the three color components of the television signals . in particular , a source of video signals is connected to a synchronizing signal separation circuit 91 for detecting the vertical and horizontal synchronizing signals in the video signals . the source of the video signals is also connected to a matrix circuit 92 for providing the three color components . each of these components is subjected to a 6 : 1 compression in compression circuit 93 and the three components are then applied to the video generation unit 56 &# 39 ;. the clock signal generator 57 &# 39 ; applies horizontal and vertical synchronizing signals to both the video generation unit 56 &# 39 ; and the microcomputer 52 &# 39 ;, and receives the synchronizing signals from the separation circuit 91 for synchronization therewith . fig1 shows a block diagram of a decoder for the second embodiment . components the same as those in fig6 are designated with the same reference number . the decoder is substantially similar as the decoder of the first embodiment with the exception that the decoder is now capable of processing color signals and the encoded data selectively includes television signals . in particular , a color decoder 101 is included between the output of the receiver 60 and the input of the analog switch 61 . 1 - 61 . 3 . the register 69 &# 39 ; includes a register element for storing the number of the picture store . the synchronizing signal detector 73 &# 39 ; outputs field synchronizing signals in addition to line synchronizing signals . the write buffer 71 &# 39 ; now accesses three picture stores 72 . 1 - 72 . 3 corresponding to the three color components , red , blue and green . the outputs of these picture stores 72 . 1 - 72 . 3 are applied to three digital - to - analog converters 75 . 1 - 75 . 3 , and then to three low - pass filters 76 . 1 - 76 . 3 for application to the other inputs of the three analog switches 61 . 1 - 61 . 3 . numerous alterations of the structure herein disclosed will suggest themselves to those skilled in the art . however it is to be understood that the embodiments herein disclosed are for purposes of illustration only and not to be construed as a limitation of the invention . all such modifications which do not depart from the spirit of the invention are intended to be included within the scope of the appended claims ."}
|
{"patent": "as noted above , the object of the present invention is to distribute information . this information is transmitted over , for example , telephone lines and converted on the client site into video signals similar to those for television reception . the information subscribed to takes the form of pages of market data , various portions of which are updated from time to time to reflect changes in the market , and subsequently presented on a video display . in a first embodiment of the invention , the video signals representing the market information are being transmitted asynchronously , that is , there are no set characteristic times ( e . g . television field and frame rates ) to restrict the transmission . as shown in fig1 pages 10 of market data are each given individual page key ( pk ) codes , for example &# 34 ; page 203 &# 34 ; and &# 34 ; page 7677 &# 34 ;. these pages 10 are then applied to an encoder 12 for application to a video transmission line 14 . video displays 16 are shown connected to the transmission line 14 for receiving the encoded pages 10 . as shown , each of the video displays 16 has a unique display identification ( id ) code , for example 217 , 503 , 123 and 417 , as shown . fig2 shows a diagram representing the video signals transmitted over the video transmission line 14 . since these signals are in the form of television signals , it should be understood that the video signals for each page of market information is in the form of a sequence of video lines which , when scanned on the video displays , form the relevant pages . as shown in fig2 a first line 20 of the transmission includes an encoded signal flag indicating to the video displays 16 that the following information is encoded data . the exact form of the flag is unimportant since the information contained is just one bit . hence , the flag may be similar to , but not identical to , the vertical synchronizing signal indicating the beginning of each field of information . the line or lines 22 contain enable reception messages . the lines 24 following the enable reception message lines 22 , contain the various updates 1 , 2 and 3 . fig3 a shows a sample enable reception ( er ) message line 22 in detail . following a horizontal synchronizing pulse 30 , an er synchronizing signal 32 is sent indicating the ensuing transmission of enable reception messages and enabling decoders to synchronize to the transmission . the er sync . signal 32 is followed by id / pk conversion pairs 34 each of which includes one of the unique display id codes and one of the individual page key ( pk ) codes for which the display identified by the display code is authorized to receive . in the example shown , the pairs 417 / 7677 , 503 / 203 and 123 / 7677 indicate that video display 417 is authorized to receive page 7677 , display 503 , page 203 ; and display 123 , page 7677 . it should be noted that in the example , display 417 is not authorized to receive either page 203 or page 7677 . the enable reception message continues for as many lines ( each including an er sync . signal 32 ) and includes as many pairs 34 as are required to associate each of the authorized displays with one of the ( many ) subscribed to pages . the process for updating each page is performed by replacing &# 34 ; tiles &# 34 ; in the relevant page . as shown in fig4 the cross - hatched tile 36 to be updated is located by two row and two column ( or two x - y pixel pair ) coordinates . fig3 b - 3f show samples of the data enable sequences , in which , in fig3 b , the sequence for the update of page 7677 is illustrated . in particular , after a horizontal synchronizing pulse 40 , a data synchronizing signal 42 is present . this sync . signal 42 , which indicates the ensuing transmission of a data enable sequence , is followed by the individual page code 44 for the page 7677 and then the coordinates 46 of the tile 36 to be replaced which , in this example , is 4 rows by 40 columns . the actual data for this tile 36 is presented in a series of lines , corresponding to the number of rows in the tile to be updated , following the data enable sequence line . similar examples are shown for pages 203 and 7677 in fig3 c and 3d , in which in page 203 , a tile of 1 row and 11 columns is updated , and again in page 7677 , a second tile of 6 rows and 40 columns is updated . alternatively , as shown in fig3 e , the update data may appear on the same line as the data enable sequence . in particular , the sync . signal 42 &# 39 ; is followed by the individual page code 44 . however , the coordinates 46 &# 39 ; include the pixel start number and the pixel stop number of a single row of the update data , along with the line number of the particular line . the update data 48 then follows on the same line . each tile 36 is then composed of the update data 48 appearing in , for example , a plurality of consecutive lines . further , as shown in fig3 f , the update data may be presented simultaneously on one line for more than one page at a time . in particular , the sync . signal 42 &# 34 ; is followed by two page codes 44 &# 34 ; , and then the coordinates 46 of the tile 36 to be replaced , which in this example , is 5 rows by 80 columns , toward the bottom of pages 203 and 7677 . additionally , all authorized displays connected to the video transmission may be simultaneously updated at the same coordinates by using a special page key , e . g . pk = 0 . an encoder for the first embodiment of the invention is shown in fig5 . the encoder includes a modem 50 for receiving data from a source of market information . this data may be in the form of entire pages of financial information where portions are updated , or the update data itself along with information for the positioning of the update data on the respective display screen . the output of modem 50 is connected to an interface 51 , which is , in turn , connected to the input of a microcomputer 52 . the microcomputer 52 reassigns the data to appropriate locations in new pages for clients of the provider . a keyboard 53 is connected to the microcomputer 52 for controlling the microcomputer . a memory 54 is connected to the microcomputer 52 and supplies thereto the configuration of the new pages , the individual page key ( pk ) code for each of the new pages , and the display ( id ) codes of the clients authorized to receive each of the new pages . based on this information , the microcomputer 52 generates the first data stream and the sequence of second data streams . the output of the microcomputer 52 is applied through an interface 55 , to a video generation unit 56 which reconfigures the output of the microcomputer into video lines . the video generation unit 56 also generates the encoded signal flag and inserts the various synchronizing signals at the beginning of each of the video lines . a clock signal generator 57 is connected to the video generation unit 56 and the microcomputer 52 for applying timing signals thereto at the line frequency . in the event that the financial information applied to the modem 50 is in the form of entire pages , a memory 58 is connected to the microcomputer 52 into which the pages are entered enabling the microcomputer 52 to compare one page with the update of the page to extract therefrom only the update data . fig6 is a block diagram of a decoder for use with the encoder of fig5 . the decoder includes a receiver 60 for receiving the data transmitted by the video generation unit 56 . the output of the receiver 60 is applied to an analog switch 61 for selective application to an output display in the event that standard non - coded signals are being transmitted . a coded signal detector 62 is coupled to the receiver 60 for receiving the encoded signal flag and for switching the analog switch 61 accordingly . an er detection gate 63 is connected to the receiver 60 for receiving the enable reception messages containing the display id / individual pk code pairs . each of the received display id codes are compared with a unique display id code stored in a rom 64 by a comparator 65 . upon each match of the display id code , the individual pk code for the respective page is stored in a memory 66 . the output of the receiver 60 is further connected to a data detection gate 67 for receiving the data enable sequences . the individual pk codes in the received data enable sequences are compared in a comparator 68 with the individual pk codes stored in the memory 66 . upon a match of one of these pk codes , the accompanying coordinates of the update data are loaded into registers 69 . an analog - to - digital converter 70 digitizes the appropriate update data at the output of the receiver 60 and applies its output to a write buffer 71 , which also receives the output of the registers 69 . the output of the write buffer 71 is applied to a picture store 72 in which the section therein corresponding to the location of the update data is updated . a synchronizing signal detector 73 is connected to the output of the receiver 60 for separating the line synchronizing signals . the output of the synchronizing signal detector 73 is applied to a timing and control signal generator 74 for generating timing signals for the analog - to - digital converter 70 , the data detection gate 67 , the er detection gate 63 and the picture store 72 . the output of the picture store 72 is applied to a digital - to - analog converter 75 controlled by the timing and control signal generator 74 . the output of the digital - to - analog converter 75 is applied through a low - pass filter 76 to another input of the analog switch 61 . in a second embodiment of the invention , the video signals representing the market information include three colors . in addition , standard television signals are included in the video signals for selective viewing of realtime television on the video displays . this transmission is necessarily synchronous to the chosen television standard . assuming that the video signals are being transmitted by cable , resulting in a usable bandwidth of approximately 24 mhz . fig7 shows a pictorial representation of the transmitted video signals . the encoded signal flag line 80 , the enable reception messages lines 81 and the data enable sequence lines 82 are transmitted during the vertical blanking interval 83 between each field of the video signal . during the active video portion of the field , in a first half of each scanning line , the television r , g and b signals 84 , each originally having a bandwidth of 4 mhz . and each time compressed by a factor of six to an expanded bandwidth of 24 mhz ., are sequentially transmitted . in the second half of each scanning line , the update data for individual pages of the market information are transmitted . while the television signals 84 are in color , the update information may be monochromatic , color or a mixture of both . in particular , as shown , the first 8 half - lines contain the update monochrome data 85 for the left half and right halves , respectively , for page 7677 . the update data 85 for page 7677 is followed by the update data 86 for page 203 . the update data 86 is presented in color as the three color signals r , g and b . the remainder of the right half of the first field is shown as being unused in this example . the left half of the second field contains the g , b and r components of the television signals 78 &# 39 ;. the right half of the second field contains the monochromatic update data for the pages 208 , 1234 , 5 , 19154 and 264 . due to the complex ordering of the update data in the first and second fields , the data enable sequences in the lines 82 must necessarily be more complex than those shown in fig3 b - 3f . in addition , the enable reception messages must indicate which of the video displays is authorized to receive the television signals sent with the update data . in particular , as shown in fig8 a , the enable reception messages are similar to those shown in fig3 a , with the exception that in addition to the display id code / identification pk code pairs , the messages include a pair 87 indicating which of the video displays , for example , the display with display id code 297 , is authorized to receive the television signals tv1 . fig8 b shows a sample data enable sequence which includes , in addition to that described with respect to fig3 b - 3f , the coordinates of the update data in the source field . fig8 c shows a sample of the data enable sequence for identifying the television signals , and includes a data synchronizing signal 88 , a television pk code 89 and the starting coordinates 90 of the three color signals -- red , green and blue . fig9 shows a block diagram of an encoder for the second embodiment . the video generation unit 56 &# 39 ; has a second set of inputs for receiving the three color components of the television signals . in particular , a source of video signals is connected to a synchronizing signal separation circuit 91 for detecting the vertical and horizontal synchronizing signals in the video signals . the source of the video signals is also connected to a matrix circuit 92 for providing the three color components . each of these components is subjected to a 6 : 1 compression in compression circuit 93 and the three components are then applied to the video generation unit 56 &# 39 ;. the clock signal generator 57 &# 39 ; applies horizontal and vertical synchronizing signals to both the video generation unit 56 &# 39 ; and the microcomputer 52 &# 39 ;, and receives the synchronizing signals from the separation circuit 91 for synchronization therewith . fig1 shows a block diagram of a decoder for the second embodiment . components the same as those in fig6 are designated with the same reference number . the decoder is substantially similar as the decoder of the first embodiment with the exception that the decoder is now capable of processing color signals and the encoded data selectively includes television signals . in particular , a color decoder 101 is included between the output of the receiver 60 and the input of the analog switch 61 . 1 - 61 . 3 . the register 69 &# 39 ; includes a register element for storing the number of the picture store . the synchronizing signal detector 73 &# 39 ; outputs field synchronizing signals in addition to line synchronizing signals . the write buffer 71 &# 39 ; now accesses three picture stores 72 . 1 - 72 . 3 corresponding to the three color components , red , blue and green . the outputs of these picture stores 72 . 1 - 72 . 3 are applied to three digital - to - analog converters 75 . 1 - 75 . 3 , and then to three low - pass filters 76 . 1 - 76 . 3 for application to the other inputs of the three analog switches 61 . 1 - 61 . 3 . numerous alterations of the structure herein disclosed will suggest themselves to those skilled in the art . however it is to be understood that the embodiments herein disclosed are for purposes of illustration only and not to be construed as a limitation of the invention . all such modifications which do not depart from the spirit of the invention are intended to be included within the scope of the appended claims .", "category": "Fixed Constructions"}
|
Is the patent correctly categorized?
| 0.25 |
8790d674f8c57a9e5539268a37bf38022b241b63e3f4f82bc8f6033d6eb96c97
| 0.535156 | 0.036865 | 0.660156 | 0.108398 | 0.894531 | 0.474609 |
null |
{"patent": "as noted above , the object of the present invention is to distribute information . this information is transmitted over , for example , telephone lines and converted on the client site into video signals similar to those for television reception . the information subscribed to takes the form of pages of market data , various portions of which are updated from time to time to reflect changes in the market , and subsequently presented on a video display . in a first embodiment of the invention , the video signals representing the market information are being transmitted asynchronously , that is , there are no set characteristic times ( e . g . television field and frame rates ) to restrict the transmission . as shown in fig1 pages 10 of market data are each given individual page key ( pk ) codes , for example &# 34 ; page 203 &# 34 ; and &# 34 ; page 7677 &# 34 ;. these pages 10 are then applied to an encoder 12 for application to a video transmission line 14 . video displays 16 are shown connected to the transmission line 14 for receiving the encoded pages 10 . as shown , each of the video displays 16 has a unique display identification ( id ) code , for example 217 , 503 , 123 and 417 , as shown . fig2 shows a diagram representing the video signals transmitted over the video transmission line 14 . since these signals are in the form of television signals , it should be understood that the video signals for each page of market information is in the form of a sequence of video lines which , when scanned on the video displays , form the relevant pages . as shown in fig2 a first line 20 of the transmission includes an encoded signal flag indicating to the video displays 16 that the following information is encoded data . the exact form of the flag is unimportant since the information contained is just one bit . hence , the flag may be similar to , but not identical to , the vertical synchronizing signal indicating the beginning of each field of information . the line or lines 22 contain enable reception messages . the lines 24 following the enable reception message lines 22 , contain the various updates 1 , 2 and 3 . fig3 a shows a sample enable reception ( er ) message line 22 in detail . following a horizontal synchronizing pulse 30 , an er synchronizing signal 32 is sent indicating the ensuing transmission of enable reception messages and enabling decoders to synchronize to the transmission . the er sync . signal 32 is followed by id / pk conversion pairs 34 each of which includes one of the unique display id codes and one of the individual page key ( pk ) codes for which the display identified by the display code is authorized to receive . in the example shown , the pairs 417 / 7677 , 503 / 203 and 123 / 7677 indicate that video display 417 is authorized to receive page 7677 , display 503 , page 203 ; and display 123 , page 7677 . it should be noted that in the example , display 417 is not authorized to receive either page 203 or page 7677 . the enable reception message continues for as many lines ( each including an er sync . signal 32 ) and includes as many pairs 34 as are required to associate each of the authorized displays with one of the ( many ) subscribed to pages . the process for updating each page is performed by replacing &# 34 ; tiles &# 34 ; in the relevant page . as shown in fig4 the cross - hatched tile 36 to be updated is located by two row and two column ( or two x - y pixel pair ) coordinates . fig3 b - 3f show samples of the data enable sequences , in which , in fig3 b , the sequence for the update of page 7677 is illustrated . in particular , after a horizontal synchronizing pulse 40 , a data synchronizing signal 42 is present . this sync . signal 42 , which indicates the ensuing transmission of a data enable sequence , is followed by the individual page code 44 for the page 7677 and then the coordinates 46 of the tile 36 to be replaced which , in this example , is 4 rows by 40 columns . the actual data for this tile 36 is presented in a series of lines , corresponding to the number of rows in the tile to be updated , following the data enable sequence line . similar examples are shown for pages 203 and 7677 in fig3 c and 3d , in which in page 203 , a tile of 1 row and 11 columns is updated , and again in page 7677 , a second tile of 6 rows and 40 columns is updated . alternatively , as shown in fig3 e , the update data may appear on the same line as the data enable sequence . in particular , the sync . signal 42 &# 39 ; is followed by the individual page code 44 . however , the coordinates 46 &# 39 ; include the pixel start number and the pixel stop number of a single row of the update data , along with the line number of the particular line . the update data 48 then follows on the same line . each tile 36 is then composed of the update data 48 appearing in , for example , a plurality of consecutive lines . further , as shown in fig3 f , the update data may be presented simultaneously on one line for more than one page at a time . in particular , the sync . signal 42 &# 34 ; is followed by two page codes 44 &# 34 ; , and then the coordinates 46 of the tile 36 to be replaced , which in this example , is 5 rows by 80 columns , toward the bottom of pages 203 and 7677 . additionally , all authorized displays connected to the video transmission may be simultaneously updated at the same coordinates by using a special page key , e . g . pk = 0 . an encoder for the first embodiment of the invention is shown in fig5 . the encoder includes a modem 50 for receiving data from a source of market information . this data may be in the form of entire pages of financial information where portions are updated , or the update data itself along with information for the positioning of the update data on the respective display screen . the output of modem 50 is connected to an interface 51 , which is , in turn , connected to the input of a microcomputer 52 . the microcomputer 52 reassigns the data to appropriate locations in new pages for clients of the provider . a keyboard 53 is connected to the microcomputer 52 for controlling the microcomputer . a memory 54 is connected to the microcomputer 52 and supplies thereto the configuration of the new pages , the individual page key ( pk ) code for each of the new pages , and the display ( id ) codes of the clients authorized to receive each of the new pages . based on this information , the microcomputer 52 generates the first data stream and the sequence of second data streams . the output of the microcomputer 52 is applied through an interface 55 , to a video generation unit 56 which reconfigures the output of the microcomputer into video lines . the video generation unit 56 also generates the encoded signal flag and inserts the various synchronizing signals at the beginning of each of the video lines . a clock signal generator 57 is connected to the video generation unit 56 and the microcomputer 52 for applying timing signals thereto at the line frequency . in the event that the financial information applied to the modem 50 is in the form of entire pages , a memory 58 is connected to the microcomputer 52 into which the pages are entered enabling the microcomputer 52 to compare one page with the update of the page to extract therefrom only the update data . fig6 is a block diagram of a decoder for use with the encoder of fig5 . the decoder includes a receiver 60 for receiving the data transmitted by the video generation unit 56 . the output of the receiver 60 is applied to an analog switch 61 for selective application to an output display in the event that standard non - coded signals are being transmitted . a coded signal detector 62 is coupled to the receiver 60 for receiving the encoded signal flag and for switching the analog switch 61 accordingly . an er detection gate 63 is connected to the receiver 60 for receiving the enable reception messages containing the display id / individual pk code pairs . each of the received display id codes are compared with a unique display id code stored in a rom 64 by a comparator 65 . upon each match of the display id code , the individual pk code for the respective page is stored in a memory 66 . the output of the receiver 60 is further connected to a data detection gate 67 for receiving the data enable sequences . the individual pk codes in the received data enable sequences are compared in a comparator 68 with the individual pk codes stored in the memory 66 . upon a match of one of these pk codes , the accompanying coordinates of the update data are loaded into registers 69 . an analog - to - digital converter 70 digitizes the appropriate update data at the output of the receiver 60 and applies its output to a write buffer 71 , which also receives the output of the registers 69 . the output of the write buffer 71 is applied to a picture store 72 in which the section therein corresponding to the location of the update data is updated . a synchronizing signal detector 73 is connected to the output of the receiver 60 for separating the line synchronizing signals . the output of the synchronizing signal detector 73 is applied to a timing and control signal generator 74 for generating timing signals for the analog - to - digital converter 70 , the data detection gate 67 , the er detection gate 63 and the picture store 72 . the output of the picture store 72 is applied to a digital - to - analog converter 75 controlled by the timing and control signal generator 74 . the output of the digital - to - analog converter 75 is applied through a low - pass filter 76 to another input of the analog switch 61 . in a second embodiment of the invention , the video signals representing the market information include three colors . in addition , standard television signals are included in the video signals for selective viewing of realtime television on the video displays . this transmission is necessarily synchronous to the chosen television standard . assuming that the video signals are being transmitted by cable , resulting in a usable bandwidth of approximately 24 mhz . fig7 shows a pictorial representation of the transmitted video signals . the encoded signal flag line 80 , the enable reception messages lines 81 and the data enable sequence lines 82 are transmitted during the vertical blanking interval 83 between each field of the video signal . during the active video portion of the field , in a first half of each scanning line , the television r , g and b signals 84 , each originally having a bandwidth of 4 mhz . and each time compressed by a factor of six to an expanded bandwidth of 24 mhz ., are sequentially transmitted . in the second half of each scanning line , the update data for individual pages of the market information are transmitted . while the television signals 84 are in color , the update information may be monochromatic , color or a mixture of both . in particular , as shown , the first 8 half - lines contain the update monochrome data 85 for the left half and right halves , respectively , for page 7677 . the update data 85 for page 7677 is followed by the update data 86 for page 203 . the update data 86 is presented in color as the three color signals r , g and b . the remainder of the right half of the first field is shown as being unused in this example . the left half of the second field contains the g , b and r components of the television signals 78 &# 39 ;. the right half of the second field contains the monochromatic update data for the pages 208 , 1234 , 5 , 19154 and 264 . due to the complex ordering of the update data in the first and second fields , the data enable sequences in the lines 82 must necessarily be more complex than those shown in fig3 b - 3f . in addition , the enable reception messages must indicate which of the video displays is authorized to receive the television signals sent with the update data . in particular , as shown in fig8 a , the enable reception messages are similar to those shown in fig3 a , with the exception that in addition to the display id code / identification pk code pairs , the messages include a pair 87 indicating which of the video displays , for example , the display with display id code 297 , is authorized to receive the television signals tv1 . fig8 b shows a sample data enable sequence which includes , in addition to that described with respect to fig3 b - 3f , the coordinates of the update data in the source field . fig8 c shows a sample of the data enable sequence for identifying the television signals , and includes a data synchronizing signal 88 , a television pk code 89 and the starting coordinates 90 of the three color signals -- red , green and blue . fig9 shows a block diagram of an encoder for the second embodiment . the video generation unit 56 &# 39 ; has a second set of inputs for receiving the three color components of the television signals . in particular , a source of video signals is connected to a synchronizing signal separation circuit 91 for detecting the vertical and horizontal synchronizing signals in the video signals . the source of the video signals is also connected to a matrix circuit 92 for providing the three color components . each of these components is subjected to a 6 : 1 compression in compression circuit 93 and the three components are then applied to the video generation unit 56 &# 39 ;. the clock signal generator 57 &# 39 ; applies horizontal and vertical synchronizing signals to both the video generation unit 56 &# 39 ; and the microcomputer 52 &# 39 ;, and receives the synchronizing signals from the separation circuit 91 for synchronization therewith . fig1 shows a block diagram of a decoder for the second embodiment . components the same as those in fig6 are designated with the same reference number . the decoder is substantially similar as the decoder of the first embodiment with the exception that the decoder is now capable of processing color signals and the encoded data selectively includes television signals . in particular , a color decoder 101 is included between the output of the receiver 60 and the input of the analog switch 61 . 1 - 61 . 3 . the register 69 &# 39 ; includes a register element for storing the number of the picture store . the synchronizing signal detector 73 &# 39 ; outputs field synchronizing signals in addition to line synchronizing signals . the write buffer 71 &# 39 ; now accesses three picture stores 72 . 1 - 72 . 3 corresponding to the three color components , red , blue and green . the outputs of these picture stores 72 . 1 - 72 . 3 are applied to three digital - to - analog converters 75 . 1 - 75 . 3 , and then to three low - pass filters 76 . 1 - 76 . 3 for application to the other inputs of the three analog switches 61 . 1 - 61 . 3 . numerous alterations of the structure herein disclosed will suggest themselves to those skilled in the art . however it is to be understood that the embodiments herein disclosed are for purposes of illustration only and not to be construed as a limitation of the invention . all such modifications which do not depart from the spirit of the invention are intended to be included within the scope of the appended claims .", "category": "Electricity"}
|
{"patent": "as noted above , the object of the present invention is to distribute information . this information is transmitted over , for example , telephone lines and converted on the client site into video signals similar to those for television reception . the information subscribed to takes the form of pages of market data , various portions of which are updated from time to time to reflect changes in the market , and subsequently presented on a video display . in a first embodiment of the invention , the video signals representing the market information are being transmitted asynchronously , that is , there are no set characteristic times ( e . g . television field and frame rates ) to restrict the transmission . as shown in fig1 pages 10 of market data are each given individual page key ( pk ) codes , for example &# 34 ; page 203 &# 34 ; and &# 34 ; page 7677 &# 34 ;. these pages 10 are then applied to an encoder 12 for application to a video transmission line 14 . video displays 16 are shown connected to the transmission line 14 for receiving the encoded pages 10 . as shown , each of the video displays 16 has a unique display identification ( id ) code , for example 217 , 503 , 123 and 417 , as shown . fig2 shows a diagram representing the video signals transmitted over the video transmission line 14 . since these signals are in the form of television signals , it should be understood that the video signals for each page of market information is in the form of a sequence of video lines which , when scanned on the video displays , form the relevant pages . as shown in fig2 a first line 20 of the transmission includes an encoded signal flag indicating to the video displays 16 that the following information is encoded data . the exact form of the flag is unimportant since the information contained is just one bit . hence , the flag may be similar to , but not identical to , the vertical synchronizing signal indicating the beginning of each field of information . the line or lines 22 contain enable reception messages . the lines 24 following the enable reception message lines 22 , contain the various updates 1 , 2 and 3 . fig3 a shows a sample enable reception ( er ) message line 22 in detail . following a horizontal synchronizing pulse 30 , an er synchronizing signal 32 is sent indicating the ensuing transmission of enable reception messages and enabling decoders to synchronize to the transmission . the er sync . signal 32 is followed by id / pk conversion pairs 34 each of which includes one of the unique display id codes and one of the individual page key ( pk ) codes for which the display identified by the display code is authorized to receive . in the example shown , the pairs 417 / 7677 , 503 / 203 and 123 / 7677 indicate that video display 417 is authorized to receive page 7677 , display 503 , page 203 ; and display 123 , page 7677 . it should be noted that in the example , display 417 is not authorized to receive either page 203 or page 7677 . the enable reception message continues for as many lines ( each including an er sync . signal 32 ) and includes as many pairs 34 as are required to associate each of the authorized displays with one of the ( many ) subscribed to pages . the process for updating each page is performed by replacing &# 34 ; tiles &# 34 ; in the relevant page . as shown in fig4 the cross - hatched tile 36 to be updated is located by two row and two column ( or two x - y pixel pair ) coordinates . fig3 b - 3f show samples of the data enable sequences , in which , in fig3 b , the sequence for the update of page 7677 is illustrated . in particular , after a horizontal synchronizing pulse 40 , a data synchronizing signal 42 is present . this sync . signal 42 , which indicates the ensuing transmission of a data enable sequence , is followed by the individual page code 44 for the page 7677 and then the coordinates 46 of the tile 36 to be replaced which , in this example , is 4 rows by 40 columns . the actual data for this tile 36 is presented in a series of lines , corresponding to the number of rows in the tile to be updated , following the data enable sequence line . similar examples are shown for pages 203 and 7677 in fig3 c and 3d , in which in page 203 , a tile of 1 row and 11 columns is updated , and again in page 7677 , a second tile of 6 rows and 40 columns is updated . alternatively , as shown in fig3 e , the update data may appear on the same line as the data enable sequence . in particular , the sync . signal 42 &# 39 ; is followed by the individual page code 44 . however , the coordinates 46 &# 39 ; include the pixel start number and the pixel stop number of a single row of the update data , along with the line number of the particular line . the update data 48 then follows on the same line . each tile 36 is then composed of the update data 48 appearing in , for example , a plurality of consecutive lines . further , as shown in fig3 f , the update data may be presented simultaneously on one line for more than one page at a time . in particular , the sync . signal 42 &# 34 ; is followed by two page codes 44 &# 34 ; , and then the coordinates 46 of the tile 36 to be replaced , which in this example , is 5 rows by 80 columns , toward the bottom of pages 203 and 7677 . additionally , all authorized displays connected to the video transmission may be simultaneously updated at the same coordinates by using a special page key , e . g . pk = 0 . an encoder for the first embodiment of the invention is shown in fig5 . the encoder includes a modem 50 for receiving data from a source of market information . this data may be in the form of entire pages of financial information where portions are updated , or the update data itself along with information for the positioning of the update data on the respective display screen . the output of modem 50 is connected to an interface 51 , which is , in turn , connected to the input of a microcomputer 52 . the microcomputer 52 reassigns the data to appropriate locations in new pages for clients of the provider . a keyboard 53 is connected to the microcomputer 52 for controlling the microcomputer . a memory 54 is connected to the microcomputer 52 and supplies thereto the configuration of the new pages , the individual page key ( pk ) code for each of the new pages , and the display ( id ) codes of the clients authorized to receive each of the new pages . based on this information , the microcomputer 52 generates the first data stream and the sequence of second data streams . the output of the microcomputer 52 is applied through an interface 55 , to a video generation unit 56 which reconfigures the output of the microcomputer into video lines . the video generation unit 56 also generates the encoded signal flag and inserts the various synchronizing signals at the beginning of each of the video lines . a clock signal generator 57 is connected to the video generation unit 56 and the microcomputer 52 for applying timing signals thereto at the line frequency . in the event that the financial information applied to the modem 50 is in the form of entire pages , a memory 58 is connected to the microcomputer 52 into which the pages are entered enabling the microcomputer 52 to compare one page with the update of the page to extract therefrom only the update data . fig6 is a block diagram of a decoder for use with the encoder of fig5 . the decoder includes a receiver 60 for receiving the data transmitted by the video generation unit 56 . the output of the receiver 60 is applied to an analog switch 61 for selective application to an output display in the event that standard non - coded signals are being transmitted . a coded signal detector 62 is coupled to the receiver 60 for receiving the encoded signal flag and for switching the analog switch 61 accordingly . an er detection gate 63 is connected to the receiver 60 for receiving the enable reception messages containing the display id / individual pk code pairs . each of the received display id codes are compared with a unique display id code stored in a rom 64 by a comparator 65 . upon each match of the display id code , the individual pk code for the respective page is stored in a memory 66 . the output of the receiver 60 is further connected to a data detection gate 67 for receiving the data enable sequences . the individual pk codes in the received data enable sequences are compared in a comparator 68 with the individual pk codes stored in the memory 66 . upon a match of one of these pk codes , the accompanying coordinates of the update data are loaded into registers 69 . an analog - to - digital converter 70 digitizes the appropriate update data at the output of the receiver 60 and applies its output to a write buffer 71 , which also receives the output of the registers 69 . the output of the write buffer 71 is applied to a picture store 72 in which the section therein corresponding to the location of the update data is updated . a synchronizing signal detector 73 is connected to the output of the receiver 60 for separating the line synchronizing signals . the output of the synchronizing signal detector 73 is applied to a timing and control signal generator 74 for generating timing signals for the analog - to - digital converter 70 , the data detection gate 67 , the er detection gate 63 and the picture store 72 . the output of the picture store 72 is applied to a digital - to - analog converter 75 controlled by the timing and control signal generator 74 . the output of the digital - to - analog converter 75 is applied through a low - pass filter 76 to another input of the analog switch 61 . in a second embodiment of the invention , the video signals representing the market information include three colors . in addition , standard television signals are included in the video signals for selective viewing of realtime television on the video displays . this transmission is necessarily synchronous to the chosen television standard . assuming that the video signals are being transmitted by cable , resulting in a usable bandwidth of approximately 24 mhz . fig7 shows a pictorial representation of the transmitted video signals . the encoded signal flag line 80 , the enable reception messages lines 81 and the data enable sequence lines 82 are transmitted during the vertical blanking interval 83 between each field of the video signal . during the active video portion of the field , in a first half of each scanning line , the television r , g and b signals 84 , each originally having a bandwidth of 4 mhz . and each time compressed by a factor of six to an expanded bandwidth of 24 mhz ., are sequentially transmitted . in the second half of each scanning line , the update data for individual pages of the market information are transmitted . while the television signals 84 are in color , the update information may be monochromatic , color or a mixture of both . in particular , as shown , the first 8 half - lines contain the update monochrome data 85 for the left half and right halves , respectively , for page 7677 . the update data 85 for page 7677 is followed by the update data 86 for page 203 . the update data 86 is presented in color as the three color signals r , g and b . the remainder of the right half of the first field is shown as being unused in this example . the left half of the second field contains the g , b and r components of the television signals 78 &# 39 ;. the right half of the second field contains the monochromatic update data for the pages 208 , 1234 , 5 , 19154 and 264 . due to the complex ordering of the update data in the first and second fields , the data enable sequences in the lines 82 must necessarily be more complex than those shown in fig3 b - 3f . in addition , the enable reception messages must indicate which of the video displays is authorized to receive the television signals sent with the update data . in particular , as shown in fig8 a , the enable reception messages are similar to those shown in fig3 a , with the exception that in addition to the display id code / identification pk code pairs , the messages include a pair 87 indicating which of the video displays , for example , the display with display id code 297 , is authorized to receive the television signals tv1 . fig8 b shows a sample data enable sequence which includes , in addition to that described with respect to fig3 b - 3f , the coordinates of the update data in the source field . fig8 c shows a sample of the data enable sequence for identifying the television signals , and includes a data synchronizing signal 88 , a television pk code 89 and the starting coordinates 90 of the three color signals -- red , green and blue . fig9 shows a block diagram of an encoder for the second embodiment . the video generation unit 56 &# 39 ; has a second set of inputs for receiving the three color components of the television signals . in particular , a source of video signals is connected to a synchronizing signal separation circuit 91 for detecting the vertical and horizontal synchronizing signals in the video signals . the source of the video signals is also connected to a matrix circuit 92 for providing the three color components . each of these components is subjected to a 6 : 1 compression in compression circuit 93 and the three components are then applied to the video generation unit 56 &# 39 ;. the clock signal generator 57 &# 39 ; applies horizontal and vertical synchronizing signals to both the video generation unit 56 &# 39 ; and the microcomputer 52 &# 39 ;, and receives the synchronizing signals from the separation circuit 91 for synchronization therewith . fig1 shows a block diagram of a decoder for the second embodiment . components the same as those in fig6 are designated with the same reference number . the decoder is substantially similar as the decoder of the first embodiment with the exception that the decoder is now capable of processing color signals and the encoded data selectively includes television signals . in particular , a color decoder 101 is included between the output of the receiver 60 and the input of the analog switch 61 . 1 - 61 . 3 . the register 69 &# 39 ; includes a register element for storing the number of the picture store . the synchronizing signal detector 73 &# 39 ; outputs field synchronizing signals in addition to line synchronizing signals . the write buffer 71 &# 39 ; now accesses three picture stores 72 . 1 - 72 . 3 corresponding to the three color components , red , blue and green . the outputs of these picture stores 72 . 1 - 72 . 3 are applied to three digital - to - analog converters 75 . 1 - 75 . 3 , and then to three low - pass filters 76 . 1 - 76 . 3 for application to the other inputs of the three analog switches 61 . 1 - 61 . 3 . numerous alterations of the structure herein disclosed will suggest themselves to those skilled in the art . however it is to be understood that the embodiments herein disclosed are for purposes of illustration only and not to be construed as a limitation of the invention . all such modifications which do not depart from the spirit of the invention are intended to be included within the scope of the appended claims .", "category": "Mechanical Engineering; Lightning; Heating; Weapons; Blasting"}
|
Does the patent belong in this category?
| 0.25 |
8790d674f8c57a9e5539268a37bf38022b241b63e3f4f82bc8f6033d6eb96c97
| 0.00014 | 0.000246 | 0.008057 | 0.001701 | 0.043457 | 0.002716 |
null |
{"patent": "as noted above , the object of the present invention is to distribute information . this information is transmitted over , for example , telephone lines and converted on the client site into video signals similar to those for television reception . the information subscribed to takes the form of pages of market data , various portions of which are updated from time to time to reflect changes in the market , and subsequently presented on a video display . in a first embodiment of the invention , the video signals representing the market information are being transmitted asynchronously , that is , there are no set characteristic times ( e . g . television field and frame rates ) to restrict the transmission . as shown in fig1 pages 10 of market data are each given individual page key ( pk ) codes , for example &# 34 ; page 203 &# 34 ; and &# 34 ; page 7677 &# 34 ;. these pages 10 are then applied to an encoder 12 for application to a video transmission line 14 . video displays 16 are shown connected to the transmission line 14 for receiving the encoded pages 10 . as shown , each of the video displays 16 has a unique display identification ( id ) code , for example 217 , 503 , 123 and 417 , as shown . fig2 shows a diagram representing the video signals transmitted over the video transmission line 14 . since these signals are in the form of television signals , it should be understood that the video signals for each page of market information is in the form of a sequence of video lines which , when scanned on the video displays , form the relevant pages . as shown in fig2 a first line 20 of the transmission includes an encoded signal flag indicating to the video displays 16 that the following information is encoded data . the exact form of the flag is unimportant since the information contained is just one bit . hence , the flag may be similar to , but not identical to , the vertical synchronizing signal indicating the beginning of each field of information . the line or lines 22 contain enable reception messages . the lines 24 following the enable reception message lines 22 , contain the various updates 1 , 2 and 3 . fig3 a shows a sample enable reception ( er ) message line 22 in detail . following a horizontal synchronizing pulse 30 , an er synchronizing signal 32 is sent indicating the ensuing transmission of enable reception messages and enabling decoders to synchronize to the transmission . the er sync . signal 32 is followed by id / pk conversion pairs 34 each of which includes one of the unique display id codes and one of the individual page key ( pk ) codes for which the display identified by the display code is authorized to receive . in the example shown , the pairs 417 / 7677 , 503 / 203 and 123 / 7677 indicate that video display 417 is authorized to receive page 7677 , display 503 , page 203 ; and display 123 , page 7677 . it should be noted that in the example , display 417 is not authorized to receive either page 203 or page 7677 . the enable reception message continues for as many lines ( each including an er sync . signal 32 ) and includes as many pairs 34 as are required to associate each of the authorized displays with one of the ( many ) subscribed to pages . the process for updating each page is performed by replacing &# 34 ; tiles &# 34 ; in the relevant page . as shown in fig4 the cross - hatched tile 36 to be updated is located by two row and two column ( or two x - y pixel pair ) coordinates . fig3 b - 3f show samples of the data enable sequences , in which , in fig3 b , the sequence for the update of page 7677 is illustrated . in particular , after a horizontal synchronizing pulse 40 , a data synchronizing signal 42 is present . this sync . signal 42 , which indicates the ensuing transmission of a data enable sequence , is followed by the individual page code 44 for the page 7677 and then the coordinates 46 of the tile 36 to be replaced which , in this example , is 4 rows by 40 columns . the actual data for this tile 36 is presented in a series of lines , corresponding to the number of rows in the tile to be updated , following the data enable sequence line . similar examples are shown for pages 203 and 7677 in fig3 c and 3d , in which in page 203 , a tile of 1 row and 11 columns is updated , and again in page 7677 , a second tile of 6 rows and 40 columns is updated . alternatively , as shown in fig3 e , the update data may appear on the same line as the data enable sequence . in particular , the sync . signal 42 &# 39 ; is followed by the individual page code 44 . however , the coordinates 46 &# 39 ; include the pixel start number and the pixel stop number of a single row of the update data , along with the line number of the particular line . the update data 48 then follows on the same line . each tile 36 is then composed of the update data 48 appearing in , for example , a plurality of consecutive lines . further , as shown in fig3 f , the update data may be presented simultaneously on one line for more than one page at a time . in particular , the sync . signal 42 &# 34 ; is followed by two page codes 44 &# 34 ; , and then the coordinates 46 of the tile 36 to be replaced , which in this example , is 5 rows by 80 columns , toward the bottom of pages 203 and 7677 . additionally , all authorized displays connected to the video transmission may be simultaneously updated at the same coordinates by using a special page key , e . g . pk = 0 . an encoder for the first embodiment of the invention is shown in fig5 . the encoder includes a modem 50 for receiving data from a source of market information . this data may be in the form of entire pages of financial information where portions are updated , or the update data itself along with information for the positioning of the update data on the respective display screen . the output of modem 50 is connected to an interface 51 , which is , in turn , connected to the input of a microcomputer 52 . the microcomputer 52 reassigns the data to appropriate locations in new pages for clients of the provider . a keyboard 53 is connected to the microcomputer 52 for controlling the microcomputer . a memory 54 is connected to the microcomputer 52 and supplies thereto the configuration of the new pages , the individual page key ( pk ) code for each of the new pages , and the display ( id ) codes of the clients authorized to receive each of the new pages . based on this information , the microcomputer 52 generates the first data stream and the sequence of second data streams . the output of the microcomputer 52 is applied through an interface 55 , to a video generation unit 56 which reconfigures the output of the microcomputer into video lines . the video generation unit 56 also generates the encoded signal flag and inserts the various synchronizing signals at the beginning of each of the video lines . a clock signal generator 57 is connected to the video generation unit 56 and the microcomputer 52 for applying timing signals thereto at the line frequency . in the event that the financial information applied to the modem 50 is in the form of entire pages , a memory 58 is connected to the microcomputer 52 into which the pages are entered enabling the microcomputer 52 to compare one page with the update of the page to extract therefrom only the update data . fig6 is a block diagram of a decoder for use with the encoder of fig5 . the decoder includes a receiver 60 for receiving the data transmitted by the video generation unit 56 . the output of the receiver 60 is applied to an analog switch 61 for selective application to an output display in the event that standard non - coded signals are being transmitted . a coded signal detector 62 is coupled to the receiver 60 for receiving the encoded signal flag and for switching the analog switch 61 accordingly . an er detection gate 63 is connected to the receiver 60 for receiving the enable reception messages containing the display id / individual pk code pairs . each of the received display id codes are compared with a unique display id code stored in a rom 64 by a comparator 65 . upon each match of the display id code , the individual pk code for the respective page is stored in a memory 66 . the output of the receiver 60 is further connected to a data detection gate 67 for receiving the data enable sequences . the individual pk codes in the received data enable sequences are compared in a comparator 68 with the individual pk codes stored in the memory 66 . upon a match of one of these pk codes , the accompanying coordinates of the update data are loaded into registers 69 . an analog - to - digital converter 70 digitizes the appropriate update data at the output of the receiver 60 and applies its output to a write buffer 71 , which also receives the output of the registers 69 . the output of the write buffer 71 is applied to a picture store 72 in which the section therein corresponding to the location of the update data is updated . a synchronizing signal detector 73 is connected to the output of the receiver 60 for separating the line synchronizing signals . the output of the synchronizing signal detector 73 is applied to a timing and control signal generator 74 for generating timing signals for the analog - to - digital converter 70 , the data detection gate 67 , the er detection gate 63 and the picture store 72 . the output of the picture store 72 is applied to a digital - to - analog converter 75 controlled by the timing and control signal generator 74 . the output of the digital - to - analog converter 75 is applied through a low - pass filter 76 to another input of the analog switch 61 . in a second embodiment of the invention , the video signals representing the market information include three colors . in addition , standard television signals are included in the video signals for selective viewing of realtime television on the video displays . this transmission is necessarily synchronous to the chosen television standard . assuming that the video signals are being transmitted by cable , resulting in a usable bandwidth of approximately 24 mhz . fig7 shows a pictorial representation of the transmitted video signals . the encoded signal flag line 80 , the enable reception messages lines 81 and the data enable sequence lines 82 are transmitted during the vertical blanking interval 83 between each field of the video signal . during the active video portion of the field , in a first half of each scanning line , the television r , g and b signals 84 , each originally having a bandwidth of 4 mhz . and each time compressed by a factor of six to an expanded bandwidth of 24 mhz ., are sequentially transmitted . in the second half of each scanning line , the update data for individual pages of the market information are transmitted . while the television signals 84 are in color , the update information may be monochromatic , color or a mixture of both . in particular , as shown , the first 8 half - lines contain the update monochrome data 85 for the left half and right halves , respectively , for page 7677 . the update data 85 for page 7677 is followed by the update data 86 for page 203 . the update data 86 is presented in color as the three color signals r , g and b . the remainder of the right half of the first field is shown as being unused in this example . the left half of the second field contains the g , b and r components of the television signals 78 &# 39 ;. the right half of the second field contains the monochromatic update data for the pages 208 , 1234 , 5 , 19154 and 264 . due to the complex ordering of the update data in the first and second fields , the data enable sequences in the lines 82 must necessarily be more complex than those shown in fig3 b - 3f . in addition , the enable reception messages must indicate which of the video displays is authorized to receive the television signals sent with the update data . in particular , as shown in fig8 a , the enable reception messages are similar to those shown in fig3 a , with the exception that in addition to the display id code / identification pk code pairs , the messages include a pair 87 indicating which of the video displays , for example , the display with display id code 297 , is authorized to receive the television signals tv1 . fig8 b shows a sample data enable sequence which includes , in addition to that described with respect to fig3 b - 3f , the coordinates of the update data in the source field . fig8 c shows a sample of the data enable sequence for identifying the television signals , and includes a data synchronizing signal 88 , a television pk code 89 and the starting coordinates 90 of the three color signals -- red , green and blue . fig9 shows a block diagram of an encoder for the second embodiment . the video generation unit 56 &# 39 ; has a second set of inputs for receiving the three color components of the television signals . in particular , a source of video signals is connected to a synchronizing signal separation circuit 91 for detecting the vertical and horizontal synchronizing signals in the video signals . the source of the video signals is also connected to a matrix circuit 92 for providing the three color components . each of these components is subjected to a 6 : 1 compression in compression circuit 93 and the three components are then applied to the video generation unit 56 &# 39 ;. the clock signal generator 57 &# 39 ; applies horizontal and vertical synchronizing signals to both the video generation unit 56 &# 39 ; and the microcomputer 52 &# 39 ;, and receives the synchronizing signals from the separation circuit 91 for synchronization therewith . fig1 shows a block diagram of a decoder for the second embodiment . components the same as those in fig6 are designated with the same reference number . the decoder is substantially similar as the decoder of the first embodiment with the exception that the decoder is now capable of processing color signals and the encoded data selectively includes television signals . in particular , a color decoder 101 is included between the output of the receiver 60 and the input of the analog switch 61 . 1 - 61 . 3 . the register 69 &# 39 ; includes a register element for storing the number of the picture store . the synchronizing signal detector 73 &# 39 ; outputs field synchronizing signals in addition to line synchronizing signals . the write buffer 71 &# 39 ; now accesses three picture stores 72 . 1 - 72 . 3 corresponding to the three color components , red , blue and green . the outputs of these picture stores 72 . 1 - 72 . 3 are applied to three digital - to - analog converters 75 . 1 - 75 . 3 , and then to three low - pass filters 76 . 1 - 76 . 3 for application to the other inputs of the three analog switches 61 . 1 - 61 . 3 . numerous alterations of the structure herein disclosed will suggest themselves to those skilled in the art . however it is to be understood that the embodiments herein disclosed are for purposes of illustration only and not to be construed as a limitation of the invention . all such modifications which do not depart from the spirit of the invention are intended to be included within the scope of the appended claims .", "category": "Electricity"}
|
{"patent": "as noted above , the object of the present invention is to distribute information . this information is transmitted over , for example , telephone lines and converted on the client site into video signals similar to those for television reception . the information subscribed to takes the form of pages of market data , various portions of which are updated from time to time to reflect changes in the market , and subsequently presented on a video display . in a first embodiment of the invention , the video signals representing the market information are being transmitted asynchronously , that is , there are no set characteristic times ( e . g . television field and frame rates ) to restrict the transmission . as shown in fig1 pages 10 of market data are each given individual page key ( pk ) codes , for example &# 34 ; page 203 &# 34 ; and &# 34 ; page 7677 &# 34 ;. these pages 10 are then applied to an encoder 12 for application to a video transmission line 14 . video displays 16 are shown connected to the transmission line 14 for receiving the encoded pages 10 . as shown , each of the video displays 16 has a unique display identification ( id ) code , for example 217 , 503 , 123 and 417 , as shown . fig2 shows a diagram representing the video signals transmitted over the video transmission line 14 . since these signals are in the form of television signals , it should be understood that the video signals for each page of market information is in the form of a sequence of video lines which , when scanned on the video displays , form the relevant pages . as shown in fig2 a first line 20 of the transmission includes an encoded signal flag indicating to the video displays 16 that the following information is encoded data . the exact form of the flag is unimportant since the information contained is just one bit . hence , the flag may be similar to , but not identical to , the vertical synchronizing signal indicating the beginning of each field of information . the line or lines 22 contain enable reception messages . the lines 24 following the enable reception message lines 22 , contain the various updates 1 , 2 and 3 . fig3 a shows a sample enable reception ( er ) message line 22 in detail . following a horizontal synchronizing pulse 30 , an er synchronizing signal 32 is sent indicating the ensuing transmission of enable reception messages and enabling decoders to synchronize to the transmission . the er sync . signal 32 is followed by id / pk conversion pairs 34 each of which includes one of the unique display id codes and one of the individual page key ( pk ) codes for which the display identified by the display code is authorized to receive . in the example shown , the pairs 417 / 7677 , 503 / 203 and 123 / 7677 indicate that video display 417 is authorized to receive page 7677 , display 503 , page 203 ; and display 123 , page 7677 . it should be noted that in the example , display 417 is not authorized to receive either page 203 or page 7677 . the enable reception message continues for as many lines ( each including an er sync . signal 32 ) and includes as many pairs 34 as are required to associate each of the authorized displays with one of the ( many ) subscribed to pages . the process for updating each page is performed by replacing &# 34 ; tiles &# 34 ; in the relevant page . as shown in fig4 the cross - hatched tile 36 to be updated is located by two row and two column ( or two x - y pixel pair ) coordinates . fig3 b - 3f show samples of the data enable sequences , in which , in fig3 b , the sequence for the update of page 7677 is illustrated . in particular , after a horizontal synchronizing pulse 40 , a data synchronizing signal 42 is present . this sync . signal 42 , which indicates the ensuing transmission of a data enable sequence , is followed by the individual page code 44 for the page 7677 and then the coordinates 46 of the tile 36 to be replaced which , in this example , is 4 rows by 40 columns . the actual data for this tile 36 is presented in a series of lines , corresponding to the number of rows in the tile to be updated , following the data enable sequence line . similar examples are shown for pages 203 and 7677 in fig3 c and 3d , in which in page 203 , a tile of 1 row and 11 columns is updated , and again in page 7677 , a second tile of 6 rows and 40 columns is updated . alternatively , as shown in fig3 e , the update data may appear on the same line as the data enable sequence . in particular , the sync . signal 42 &# 39 ; is followed by the individual page code 44 . however , the coordinates 46 &# 39 ; include the pixel start number and the pixel stop number of a single row of the update data , along with the line number of the particular line . the update data 48 then follows on the same line . each tile 36 is then composed of the update data 48 appearing in , for example , a plurality of consecutive lines . further , as shown in fig3 f , the update data may be presented simultaneously on one line for more than one page at a time . in particular , the sync . signal 42 &# 34 ; is followed by two page codes 44 &# 34 ; , and then the coordinates 46 of the tile 36 to be replaced , which in this example , is 5 rows by 80 columns , toward the bottom of pages 203 and 7677 . additionally , all authorized displays connected to the video transmission may be simultaneously updated at the same coordinates by using a special page key , e . g . pk = 0 . an encoder for the first embodiment of the invention is shown in fig5 . the encoder includes a modem 50 for receiving data from a source of market information . this data may be in the form of entire pages of financial information where portions are updated , or the update data itself along with information for the positioning of the update data on the respective display screen . the output of modem 50 is connected to an interface 51 , which is , in turn , connected to the input of a microcomputer 52 . the microcomputer 52 reassigns the data to appropriate locations in new pages for clients of the provider . a keyboard 53 is connected to the microcomputer 52 for controlling the microcomputer . a memory 54 is connected to the microcomputer 52 and supplies thereto the configuration of the new pages , the individual page key ( pk ) code for each of the new pages , and the display ( id ) codes of the clients authorized to receive each of the new pages . based on this information , the microcomputer 52 generates the first data stream and the sequence of second data streams . the output of the microcomputer 52 is applied through an interface 55 , to a video generation unit 56 which reconfigures the output of the microcomputer into video lines . the video generation unit 56 also generates the encoded signal flag and inserts the various synchronizing signals at the beginning of each of the video lines . a clock signal generator 57 is connected to the video generation unit 56 and the microcomputer 52 for applying timing signals thereto at the line frequency . in the event that the financial information applied to the modem 50 is in the form of entire pages , a memory 58 is connected to the microcomputer 52 into which the pages are entered enabling the microcomputer 52 to compare one page with the update of the page to extract therefrom only the update data . fig6 is a block diagram of a decoder for use with the encoder of fig5 . the decoder includes a receiver 60 for receiving the data transmitted by the video generation unit 56 . the output of the receiver 60 is applied to an analog switch 61 for selective application to an output display in the event that standard non - coded signals are being transmitted . a coded signal detector 62 is coupled to the receiver 60 for receiving the encoded signal flag and for switching the analog switch 61 accordingly . an er detection gate 63 is connected to the receiver 60 for receiving the enable reception messages containing the display id / individual pk code pairs . each of the received display id codes are compared with a unique display id code stored in a rom 64 by a comparator 65 . upon each match of the display id code , the individual pk code for the respective page is stored in a memory 66 . the output of the receiver 60 is further connected to a data detection gate 67 for receiving the data enable sequences . the individual pk codes in the received data enable sequences are compared in a comparator 68 with the individual pk codes stored in the memory 66 . upon a match of one of these pk codes , the accompanying coordinates of the update data are loaded into registers 69 . an analog - to - digital converter 70 digitizes the appropriate update data at the output of the receiver 60 and applies its output to a write buffer 71 , which also receives the output of the registers 69 . the output of the write buffer 71 is applied to a picture store 72 in which the section therein corresponding to the location of the update data is updated . a synchronizing signal detector 73 is connected to the output of the receiver 60 for separating the line synchronizing signals . the output of the synchronizing signal detector 73 is applied to a timing and control signal generator 74 for generating timing signals for the analog - to - digital converter 70 , the data detection gate 67 , the er detection gate 63 and the picture store 72 . the output of the picture store 72 is applied to a digital - to - analog converter 75 controlled by the timing and control signal generator 74 . the output of the digital - to - analog converter 75 is applied through a low - pass filter 76 to another input of the analog switch 61 . in a second embodiment of the invention , the video signals representing the market information include three colors . in addition , standard television signals are included in the video signals for selective viewing of realtime television on the video displays . this transmission is necessarily synchronous to the chosen television standard . assuming that the video signals are being transmitted by cable , resulting in a usable bandwidth of approximately 24 mhz . fig7 shows a pictorial representation of the transmitted video signals . the encoded signal flag line 80 , the enable reception messages lines 81 and the data enable sequence lines 82 are transmitted during the vertical blanking interval 83 between each field of the video signal . during the active video portion of the field , in a first half of each scanning line , the television r , g and b signals 84 , each originally having a bandwidth of 4 mhz . and each time compressed by a factor of six to an expanded bandwidth of 24 mhz ., are sequentially transmitted . in the second half of each scanning line , the update data for individual pages of the market information are transmitted . while the television signals 84 are in color , the update information may be monochromatic , color or a mixture of both . in particular , as shown , the first 8 half - lines contain the update monochrome data 85 for the left half and right halves , respectively , for page 7677 . the update data 85 for page 7677 is followed by the update data 86 for page 203 . the update data 86 is presented in color as the three color signals r , g and b . the remainder of the right half of the first field is shown as being unused in this example . the left half of the second field contains the g , b and r components of the television signals 78 &# 39 ;. the right half of the second field contains the monochromatic update data for the pages 208 , 1234 , 5 , 19154 and 264 . due to the complex ordering of the update data in the first and second fields , the data enable sequences in the lines 82 must necessarily be more complex than those shown in fig3 b - 3f . in addition , the enable reception messages must indicate which of the video displays is authorized to receive the television signals sent with the update data . in particular , as shown in fig8 a , the enable reception messages are similar to those shown in fig3 a , with the exception that in addition to the display id code / identification pk code pairs , the messages include a pair 87 indicating which of the video displays , for example , the display with display id code 297 , is authorized to receive the television signals tv1 . fig8 b shows a sample data enable sequence which includes , in addition to that described with respect to fig3 b - 3f , the coordinates of the update data in the source field . fig8 c shows a sample of the data enable sequence for identifying the television signals , and includes a data synchronizing signal 88 , a television pk code 89 and the starting coordinates 90 of the three color signals -- red , green and blue . fig9 shows a block diagram of an encoder for the second embodiment . the video generation unit 56 &# 39 ; has a second set of inputs for receiving the three color components of the television signals . in particular , a source of video signals is connected to a synchronizing signal separation circuit 91 for detecting the vertical and horizontal synchronizing signals in the video signals . the source of the video signals is also connected to a matrix circuit 92 for providing the three color components . each of these components is subjected to a 6 : 1 compression in compression circuit 93 and the three components are then applied to the video generation unit 56 &# 39 ;. the clock signal generator 57 &# 39 ; applies horizontal and vertical synchronizing signals to both the video generation unit 56 &# 39 ; and the microcomputer 52 &# 39 ;, and receives the synchronizing signals from the separation circuit 91 for synchronization therewith . fig1 shows a block diagram of a decoder for the second embodiment . components the same as those in fig6 are designated with the same reference number . the decoder is substantially similar as the decoder of the first embodiment with the exception that the decoder is now capable of processing color signals and the encoded data selectively includes television signals . in particular , a color decoder 101 is included between the output of the receiver 60 and the input of the analog switch 61 . 1 - 61 . 3 . the register 69 &# 39 ; includes a register element for storing the number of the picture store . the synchronizing signal detector 73 &# 39 ; outputs field synchronizing signals in addition to line synchronizing signals . the write buffer 71 &# 39 ; now accesses three picture stores 72 . 1 - 72 . 3 corresponding to the three color components , red , blue and green . the outputs of these picture stores 72 . 1 - 72 . 3 are applied to three digital - to - analog converters 75 . 1 - 75 . 3 , and then to three low - pass filters 76 . 1 - 76 . 3 for application to the other inputs of the three analog switches 61 . 1 - 61 . 3 . numerous alterations of the structure herein disclosed will suggest themselves to those skilled in the art . however it is to be understood that the embodiments herein disclosed are for purposes of illustration only and not to be construed as a limitation of the invention . all such modifications which do not depart from the spirit of the invention are intended to be included within the scope of the appended claims .", "category": "Physics"}
|
Is the category the most suitable category for the given patent?
| 0.25 |
8790d674f8c57a9e5539268a37bf38022b241b63e3f4f82bc8f6033d6eb96c97
| 0.000149 | 0.002045 | 0.014954 | 0.02063 | 0.066406 | 0.141602 |
null |
{"patent": "as noted above , the object of the present invention is to distribute information . this information is transmitted over , for example , telephone lines and converted on the client site into video signals similar to those for television reception . the information subscribed to takes the form of pages of market data , various portions of which are updated from time to time to reflect changes in the market , and subsequently presented on a video display . in a first embodiment of the invention , the video signals representing the market information are being transmitted asynchronously , that is , there are no set characteristic times ( e . g . television field and frame rates ) to restrict the transmission . as shown in fig1 pages 10 of market data are each given individual page key ( pk ) codes , for example &# 34 ; page 203 &# 34 ; and &# 34 ; page 7677 &# 34 ;. these pages 10 are then applied to an encoder 12 for application to a video transmission line 14 . video displays 16 are shown connected to the transmission line 14 for receiving the encoded pages 10 . as shown , each of the video displays 16 has a unique display identification ( id ) code , for example 217 , 503 , 123 and 417 , as shown . fig2 shows a diagram representing the video signals transmitted over the video transmission line 14 . since these signals are in the form of television signals , it should be understood that the video signals for each page of market information is in the form of a sequence of video lines which , when scanned on the video displays , form the relevant pages . as shown in fig2 a first line 20 of the transmission includes an encoded signal flag indicating to the video displays 16 that the following information is encoded data . the exact form of the flag is unimportant since the information contained is just one bit . hence , the flag may be similar to , but not identical to , the vertical synchronizing signal indicating the beginning of each field of information . the line or lines 22 contain enable reception messages . the lines 24 following the enable reception message lines 22 , contain the various updates 1 , 2 and 3 . fig3 a shows a sample enable reception ( er ) message line 22 in detail . following a horizontal synchronizing pulse 30 , an er synchronizing signal 32 is sent indicating the ensuing transmission of enable reception messages and enabling decoders to synchronize to the transmission . the er sync . signal 32 is followed by id / pk conversion pairs 34 each of which includes one of the unique display id codes and one of the individual page key ( pk ) codes for which the display identified by the display code is authorized to receive . in the example shown , the pairs 417 / 7677 , 503 / 203 and 123 / 7677 indicate that video display 417 is authorized to receive page 7677 , display 503 , page 203 ; and display 123 , page 7677 . it should be noted that in the example , display 417 is not authorized to receive either page 203 or page 7677 . the enable reception message continues for as many lines ( each including an er sync . signal 32 ) and includes as many pairs 34 as are required to associate each of the authorized displays with one of the ( many ) subscribed to pages . the process for updating each page is performed by replacing &# 34 ; tiles &# 34 ; in the relevant page . as shown in fig4 the cross - hatched tile 36 to be updated is located by two row and two column ( or two x - y pixel pair ) coordinates . fig3 b - 3f show samples of the data enable sequences , in which , in fig3 b , the sequence for the update of page 7677 is illustrated . in particular , after a horizontal synchronizing pulse 40 , a data synchronizing signal 42 is present . this sync . signal 42 , which indicates the ensuing transmission of a data enable sequence , is followed by the individual page code 44 for the page 7677 and then the coordinates 46 of the tile 36 to be replaced which , in this example , is 4 rows by 40 columns . the actual data for this tile 36 is presented in a series of lines , corresponding to the number of rows in the tile to be updated , following the data enable sequence line . similar examples are shown for pages 203 and 7677 in fig3 c and 3d , in which in page 203 , a tile of 1 row and 11 columns is updated , and again in page 7677 , a second tile of 6 rows and 40 columns is updated . alternatively , as shown in fig3 e , the update data may appear on the same line as the data enable sequence . in particular , the sync . signal 42 &# 39 ; is followed by the individual page code 44 . however , the coordinates 46 &# 39 ; include the pixel start number and the pixel stop number of a single row of the update data , along with the line number of the particular line . the update data 48 then follows on the same line . each tile 36 is then composed of the update data 48 appearing in , for example , a plurality of consecutive lines . further , as shown in fig3 f , the update data may be presented simultaneously on one line for more than one page at a time . in particular , the sync . signal 42 &# 34 ; is followed by two page codes 44 &# 34 ; , and then the coordinates 46 of the tile 36 to be replaced , which in this example , is 5 rows by 80 columns , toward the bottom of pages 203 and 7677 . additionally , all authorized displays connected to the video transmission may be simultaneously updated at the same coordinates by using a special page key , e . g . pk = 0 . an encoder for the first embodiment of the invention is shown in fig5 . the encoder includes a modem 50 for receiving data from a source of market information . this data may be in the form of entire pages of financial information where portions are updated , or the update data itself along with information for the positioning of the update data on the respective display screen . the output of modem 50 is connected to an interface 51 , which is , in turn , connected to the input of a microcomputer 52 . the microcomputer 52 reassigns the data to appropriate locations in new pages for clients of the provider . a keyboard 53 is connected to the microcomputer 52 for controlling the microcomputer . a memory 54 is connected to the microcomputer 52 and supplies thereto the configuration of the new pages , the individual page key ( pk ) code for each of the new pages , and the display ( id ) codes of the clients authorized to receive each of the new pages . based on this information , the microcomputer 52 generates the first data stream and the sequence of second data streams . the output of the microcomputer 52 is applied through an interface 55 , to a video generation unit 56 which reconfigures the output of the microcomputer into video lines . the video generation unit 56 also generates the encoded signal flag and inserts the various synchronizing signals at the beginning of each of the video lines . a clock signal generator 57 is connected to the video generation unit 56 and the microcomputer 52 for applying timing signals thereto at the line frequency . in the event that the financial information applied to the modem 50 is in the form of entire pages , a memory 58 is connected to the microcomputer 52 into which the pages are entered enabling the microcomputer 52 to compare one page with the update of the page to extract therefrom only the update data . fig6 is a block diagram of a decoder for use with the encoder of fig5 . the decoder includes a receiver 60 for receiving the data transmitted by the video generation unit 56 . the output of the receiver 60 is applied to an analog switch 61 for selective application to an output display in the event that standard non - coded signals are being transmitted . a coded signal detector 62 is coupled to the receiver 60 for receiving the encoded signal flag and for switching the analog switch 61 accordingly . an er detection gate 63 is connected to the receiver 60 for receiving the enable reception messages containing the display id / individual pk code pairs . each of the received display id codes are compared with a unique display id code stored in a rom 64 by a comparator 65 . upon each match of the display id code , the individual pk code for the respective page is stored in a memory 66 . the output of the receiver 60 is further connected to a data detection gate 67 for receiving the data enable sequences . the individual pk codes in the received data enable sequences are compared in a comparator 68 with the individual pk codes stored in the memory 66 . upon a match of one of these pk codes , the accompanying coordinates of the update data are loaded into registers 69 . an analog - to - digital converter 70 digitizes the appropriate update data at the output of the receiver 60 and applies its output to a write buffer 71 , which also receives the output of the registers 69 . the output of the write buffer 71 is applied to a picture store 72 in which the section therein corresponding to the location of the update data is updated . a synchronizing signal detector 73 is connected to the output of the receiver 60 for separating the line synchronizing signals . the output of the synchronizing signal detector 73 is applied to a timing and control signal generator 74 for generating timing signals for the analog - to - digital converter 70 , the data detection gate 67 , the er detection gate 63 and the picture store 72 . the output of the picture store 72 is applied to a digital - to - analog converter 75 controlled by the timing and control signal generator 74 . the output of the digital - to - analog converter 75 is applied through a low - pass filter 76 to another input of the analog switch 61 . in a second embodiment of the invention , the video signals representing the market information include three colors . in addition , standard television signals are included in the video signals for selective viewing of realtime television on the video displays . this transmission is necessarily synchronous to the chosen television standard . assuming that the video signals are being transmitted by cable , resulting in a usable bandwidth of approximately 24 mhz . fig7 shows a pictorial representation of the transmitted video signals . the encoded signal flag line 80 , the enable reception messages lines 81 and the data enable sequence lines 82 are transmitted during the vertical blanking interval 83 between each field of the video signal . during the active video portion of the field , in a first half of each scanning line , the television r , g and b signals 84 , each originally having a bandwidth of 4 mhz . and each time compressed by a factor of six to an expanded bandwidth of 24 mhz ., are sequentially transmitted . in the second half of each scanning line , the update data for individual pages of the market information are transmitted . while the television signals 84 are in color , the update information may be monochromatic , color or a mixture of both . in particular , as shown , the first 8 half - lines contain the update monochrome data 85 for the left half and right halves , respectively , for page 7677 . the update data 85 for page 7677 is followed by the update data 86 for page 203 . the update data 86 is presented in color as the three color signals r , g and b . the remainder of the right half of the first field is shown as being unused in this example . the left half of the second field contains the g , b and r components of the television signals 78 &# 39 ;. the right half of the second field contains the monochromatic update data for the pages 208 , 1234 , 5 , 19154 and 264 . due to the complex ordering of the update data in the first and second fields , the data enable sequences in the lines 82 must necessarily be more complex than those shown in fig3 b - 3f . in addition , the enable reception messages must indicate which of the video displays is authorized to receive the television signals sent with the update data . in particular , as shown in fig8 a , the enable reception messages are similar to those shown in fig3 a , with the exception that in addition to the display id code / identification pk code pairs , the messages include a pair 87 indicating which of the video displays , for example , the display with display id code 297 , is authorized to receive the television signals tv1 . fig8 b shows a sample data enable sequence which includes , in addition to that described with respect to fig3 b - 3f , the coordinates of the update data in the source field . fig8 c shows a sample of the data enable sequence for identifying the television signals , and includes a data synchronizing signal 88 , a television pk code 89 and the starting coordinates 90 of the three color signals -- red , green and blue . fig9 shows a block diagram of an encoder for the second embodiment . the video generation unit 56 &# 39 ; has a second set of inputs for receiving the three color components of the television signals . in particular , a source of video signals is connected to a synchronizing signal separation circuit 91 for detecting the vertical and horizontal synchronizing signals in the video signals . the source of the video signals is also connected to a matrix circuit 92 for providing the three color components . each of these components is subjected to a 6 : 1 compression in compression circuit 93 and the three components are then applied to the video generation unit 56 &# 39 ;. the clock signal generator 57 &# 39 ; applies horizontal and vertical synchronizing signals to both the video generation unit 56 &# 39 ; and the microcomputer 52 &# 39 ;, and receives the synchronizing signals from the separation circuit 91 for synchronization therewith . fig1 shows a block diagram of a decoder for the second embodiment . components the same as those in fig6 are designated with the same reference number . the decoder is substantially similar as the decoder of the first embodiment with the exception that the decoder is now capable of processing color signals and the encoded data selectively includes television signals . in particular , a color decoder 101 is included between the output of the receiver 60 and the input of the analog switch 61 . 1 - 61 . 3 . the register 69 &# 39 ; includes a register element for storing the number of the picture store . the synchronizing signal detector 73 &# 39 ; outputs field synchronizing signals in addition to line synchronizing signals . the write buffer 71 &# 39 ; now accesses three picture stores 72 . 1 - 72 . 3 corresponding to the three color components , red , blue and green . the outputs of these picture stores 72 . 1 - 72 . 3 are applied to three digital - to - analog converters 75 . 1 - 75 . 3 , and then to three low - pass filters 76 . 1 - 76 . 3 for application to the other inputs of the three analog switches 61 . 1 - 61 . 3 . numerous alterations of the structure herein disclosed will suggest themselves to those skilled in the art . however it is to be understood that the embodiments herein disclosed are for purposes of illustration only and not to be construed as a limitation of the invention . all such modifications which do not depart from the spirit of the invention are intended to be included within the scope of the appended claims .", "category": "Electricity"}
|
{"category": "General tagging of new or cross-sectional technology", "patent": "as noted above , the object of the present invention is to distribute information . this information is transmitted over , for example , telephone lines and converted on the client site into video signals similar to those for television reception . the information subscribed to takes the form of pages of market data , various portions of which are updated from time to time to reflect changes in the market , and subsequently presented on a video display . in a first embodiment of the invention , the video signals representing the market information are being transmitted asynchronously , that is , there are no set characteristic times ( e . g . television field and frame rates ) to restrict the transmission . as shown in fig1 pages 10 of market data are each given individual page key ( pk ) codes , for example &# 34 ; page 203 &# 34 ; and &# 34 ; page 7677 &# 34 ;. these pages 10 are then applied to an encoder 12 for application to a video transmission line 14 . video displays 16 are shown connected to the transmission line 14 for receiving the encoded pages 10 . as shown , each of the video displays 16 has a unique display identification ( id ) code , for example 217 , 503 , 123 and 417 , as shown . fig2 shows a diagram representing the video signals transmitted over the video transmission line 14 . since these signals are in the form of television signals , it should be understood that the video signals for each page of market information is in the form of a sequence of video lines which , when scanned on the video displays , form the relevant pages . as shown in fig2 a first line 20 of the transmission includes an encoded signal flag indicating to the video displays 16 that the following information is encoded data . the exact form of the flag is unimportant since the information contained is just one bit . hence , the flag may be similar to , but not identical to , the vertical synchronizing signal indicating the beginning of each field of information . the line or lines 22 contain enable reception messages . the lines 24 following the enable reception message lines 22 , contain the various updates 1 , 2 and 3 . fig3 a shows a sample enable reception ( er ) message line 22 in detail . following a horizontal synchronizing pulse 30 , an er synchronizing signal 32 is sent indicating the ensuing transmission of enable reception messages and enabling decoders to synchronize to the transmission . the er sync . signal 32 is followed by id / pk conversion pairs 34 each of which includes one of the unique display id codes and one of the individual page key ( pk ) codes for which the display identified by the display code is authorized to receive . in the example shown , the pairs 417 / 7677 , 503 / 203 and 123 / 7677 indicate that video display 417 is authorized to receive page 7677 , display 503 , page 203 ; and display 123 , page 7677 . it should be noted that in the example , display 417 is not authorized to receive either page 203 or page 7677 . the enable reception message continues for as many lines ( each including an er sync . signal 32 ) and includes as many pairs 34 as are required to associate each of the authorized displays with one of the ( many ) subscribed to pages . the process for updating each page is performed by replacing &# 34 ; tiles &# 34 ; in the relevant page . as shown in fig4 the cross - hatched tile 36 to be updated is located by two row and two column ( or two x - y pixel pair ) coordinates . fig3 b - 3f show samples of the data enable sequences , in which , in fig3 b , the sequence for the update of page 7677 is illustrated . in particular , after a horizontal synchronizing pulse 40 , a data synchronizing signal 42 is present . this sync . signal 42 , which indicates the ensuing transmission of a data enable sequence , is followed by the individual page code 44 for the page 7677 and then the coordinates 46 of the tile 36 to be replaced which , in this example , is 4 rows by 40 columns . the actual data for this tile 36 is presented in a series of lines , corresponding to the number of rows in the tile to be updated , following the data enable sequence line . similar examples are shown for pages 203 and 7677 in fig3 c and 3d , in which in page 203 , a tile of 1 row and 11 columns is updated , and again in page 7677 , a second tile of 6 rows and 40 columns is updated . alternatively , as shown in fig3 e , the update data may appear on the same line as the data enable sequence . in particular , the sync . signal 42 &# 39 ; is followed by the individual page code 44 . however , the coordinates 46 &# 39 ; include the pixel start number and the pixel stop number of a single row of the update data , along with the line number of the particular line . the update data 48 then follows on the same line . each tile 36 is then composed of the update data 48 appearing in , for example , a plurality of consecutive lines . further , as shown in fig3 f , the update data may be presented simultaneously on one line for more than one page at a time . in particular , the sync . signal 42 &# 34 ; is followed by two page codes 44 &# 34 ; , and then the coordinates 46 of the tile 36 to be replaced , which in this example , is 5 rows by 80 columns , toward the bottom of pages 203 and 7677 . additionally , all authorized displays connected to the video transmission may be simultaneously updated at the same coordinates by using a special page key , e . g . pk = 0 . an encoder for the first embodiment of the invention is shown in fig5 . the encoder includes a modem 50 for receiving data from a source of market information . this data may be in the form of entire pages of financial information where portions are updated , or the update data itself along with information for the positioning of the update data on the respective display screen . the output of modem 50 is connected to an interface 51 , which is , in turn , connected to the input of a microcomputer 52 . the microcomputer 52 reassigns the data to appropriate locations in new pages for clients of the provider . a keyboard 53 is connected to the microcomputer 52 for controlling the microcomputer . a memory 54 is connected to the microcomputer 52 and supplies thereto the configuration of the new pages , the individual page key ( pk ) code for each of the new pages , and the display ( id ) codes of the clients authorized to receive each of the new pages . based on this information , the microcomputer 52 generates the first data stream and the sequence of second data streams . the output of the microcomputer 52 is applied through an interface 55 , to a video generation unit 56 which reconfigures the output of the microcomputer into video lines . the video generation unit 56 also generates the encoded signal flag and inserts the various synchronizing signals at the beginning of each of the video lines . a clock signal generator 57 is connected to the video generation unit 56 and the microcomputer 52 for applying timing signals thereto at the line frequency . in the event that the financial information applied to the modem 50 is in the form of entire pages , a memory 58 is connected to the microcomputer 52 into which the pages are entered enabling the microcomputer 52 to compare one page with the update of the page to extract therefrom only the update data . fig6 is a block diagram of a decoder for use with the encoder of fig5 . the decoder includes a receiver 60 for receiving the data transmitted by the video generation unit 56 . the output of the receiver 60 is applied to an analog switch 61 for selective application to an output display in the event that standard non - coded signals are being transmitted . a coded signal detector 62 is coupled to the receiver 60 for receiving the encoded signal flag and for switching the analog switch 61 accordingly . an er detection gate 63 is connected to the receiver 60 for receiving the enable reception messages containing the display id / individual pk code pairs . each of the received display id codes are compared with a unique display id code stored in a rom 64 by a comparator 65 . upon each match of the display id code , the individual pk code for the respective page is stored in a memory 66 . the output of the receiver 60 is further connected to a data detection gate 67 for receiving the data enable sequences . the individual pk codes in the received data enable sequences are compared in a comparator 68 with the individual pk codes stored in the memory 66 . upon a match of one of these pk codes , the accompanying coordinates of the update data are loaded into registers 69 . an analog - to - digital converter 70 digitizes the appropriate update data at the output of the receiver 60 and applies its output to a write buffer 71 , which also receives the output of the registers 69 . the output of the write buffer 71 is applied to a picture store 72 in which the section therein corresponding to the location of the update data is updated . a synchronizing signal detector 73 is connected to the output of the receiver 60 for separating the line synchronizing signals . the output of the synchronizing signal detector 73 is applied to a timing and control signal generator 74 for generating timing signals for the analog - to - digital converter 70 , the data detection gate 67 , the er detection gate 63 and the picture store 72 . the output of the picture store 72 is applied to a digital - to - analog converter 75 controlled by the timing and control signal generator 74 . the output of the digital - to - analog converter 75 is applied through a low - pass filter 76 to another input of the analog switch 61 . in a second embodiment of the invention , the video signals representing the market information include three colors . in addition , standard television signals are included in the video signals for selective viewing of realtime television on the video displays . this transmission is necessarily synchronous to the chosen television standard . assuming that the video signals are being transmitted by cable , resulting in a usable bandwidth of approximately 24 mhz . fig7 shows a pictorial representation of the transmitted video signals . the encoded signal flag line 80 , the enable reception messages lines 81 and the data enable sequence lines 82 are transmitted during the vertical blanking interval 83 between each field of the video signal . during the active video portion of the field , in a first half of each scanning line , the television r , g and b signals 84 , each originally having a bandwidth of 4 mhz . and each time compressed by a factor of six to an expanded bandwidth of 24 mhz ., are sequentially transmitted . in the second half of each scanning line , the update data for individual pages of the market information are transmitted . while the television signals 84 are in color , the update information may be monochromatic , color or a mixture of both . in particular , as shown , the first 8 half - lines contain the update monochrome data 85 for the left half and right halves , respectively , for page 7677 . the update data 85 for page 7677 is followed by the update data 86 for page 203 . the update data 86 is presented in color as the three color signals r , g and b . the remainder of the right half of the first field is shown as being unused in this example . the left half of the second field contains the g , b and r components of the television signals 78 &# 39 ;. the right half of the second field contains the monochromatic update data for the pages 208 , 1234 , 5 , 19154 and 264 . due to the complex ordering of the update data in the first and second fields , the data enable sequences in the lines 82 must necessarily be more complex than those shown in fig3 b - 3f . in addition , the enable reception messages must indicate which of the video displays is authorized to receive the television signals sent with the update data . in particular , as shown in fig8 a , the enable reception messages are similar to those shown in fig3 a , with the exception that in addition to the display id code / identification pk code pairs , the messages include a pair 87 indicating which of the video displays , for example , the display with display id code 297 , is authorized to receive the television signals tv1 . fig8 b shows a sample data enable sequence which includes , in addition to that described with respect to fig3 b - 3f , the coordinates of the update data in the source field . fig8 c shows a sample of the data enable sequence for identifying the television signals , and includes a data synchronizing signal 88 , a television pk code 89 and the starting coordinates 90 of the three color signals -- red , green and blue . fig9 shows a block diagram of an encoder for the second embodiment . the video generation unit 56 &# 39 ; has a second set of inputs for receiving the three color components of the television signals . in particular , a source of video signals is connected to a synchronizing signal separation circuit 91 for detecting the vertical and horizontal synchronizing signals in the video signals . the source of the video signals is also connected to a matrix circuit 92 for providing the three color components . each of these components is subjected to a 6 : 1 compression in compression circuit 93 and the three components are then applied to the video generation unit 56 &# 39 ;. the clock signal generator 57 &# 39 ; applies horizontal and vertical synchronizing signals to both the video generation unit 56 &# 39 ; and the microcomputer 52 &# 39 ;, and receives the synchronizing signals from the separation circuit 91 for synchronization therewith . fig1 shows a block diagram of a decoder for the second embodiment . components the same as those in fig6 are designated with the same reference number . the decoder is substantially similar as the decoder of the first embodiment with the exception that the decoder is now capable of processing color signals and the encoded data selectively includes television signals . in particular , a color decoder 101 is included between the output of the receiver 60 and the input of the analog switch 61 . 1 - 61 . 3 . the register 69 &# 39 ; includes a register element for storing the number of the picture store . the synchronizing signal detector 73 &# 39 ; outputs field synchronizing signals in addition to line synchronizing signals . the write buffer 71 &# 39 ; now accesses three picture stores 72 . 1 - 72 . 3 corresponding to the three color components , red , blue and green . the outputs of these picture stores 72 . 1 - 72 . 3 are applied to three digital - to - analog converters 75 . 1 - 75 . 3 , and then to three low - pass filters 76 . 1 - 76 . 3 for application to the other inputs of the three analog switches 61 . 1 - 61 . 3 . numerous alterations of the structure herein disclosed will suggest themselves to those skilled in the art . however it is to be understood that the embodiments herein disclosed are for purposes of illustration only and not to be construed as a limitation of the invention . all such modifications which do not depart from the spirit of the invention are intended to be included within the scope of the appended claims ."}
|
Is the category the most suitable category for the given patent?
| 0.25 |
8790d674f8c57a9e5539268a37bf38022b241b63e3f4f82bc8f6033d6eb96c97
| 0.000149 | 0.176758 | 0.014954 | 0.038574 | 0.066406 | 0.109863 |
null |
{"category": "Performing Operations; Transporting", "patent": "referring to fig2 of the drawings , one preferred embodiment of the invention is illustrated . a sinter or porous layer 10 is placed on top of the pot 6 to provide a further series resistance r pot2 to the pot i . e . an appropriate sinter 10 may have openings of microns in dimension and only be a few millimeters thick . this method may reduce the q b / q i by a factor of 10 . such an arrangement provides added benefits when used for membrane filter systems in a bio - reactor . the high solids feed in bio reactors leads to the sludge actually plugging the filter and self sealing the broken fibre totally . the may be extended to the general case by replacing the sinter with a membrane with the same pore size as the hollow fibre membrane and enabling achievement of this self plugging capability even with low solids feeds . it will be apparent the extra resistance of the sinter or membrane 10 will require an extra pressure to maintain the module filtrate flow , however , this is only an operating cost not a membrane process operating efficiency as it is operating over the pot assembly , not across the compressible dirt layer on the membrane . fouling of this membrane sinter can be reduced by a regular chemical cleaning backwash with chlorine or other suitable cleaners . the membrane / sinter 10 is desirably in intimate contact with the pot 6 to prevent sideways flow of filtrate / feed bypass . this may also be achieved with a replaceable sinter / membrane element . a highly asymmetric membrane 10 with the large pore side contacting the pot 6 ( so in normal filtrate flow the filtrate flows in the direction of reducing pore size ) is desirable . as shown in fig3 b - 3k a variety of methods may be used to increase the pot flow resistance . referring to fig3 a a normal pot 6 without modification is shown . fig3 b shows an increased length pot 6 which , while increasing pot flow resistance , has other disadvantages . fig3 c illustrates providing the fibre 5 with a non porous coating 7 adjacent the interface 8 between the fibre 5 and the pot 6 . this serves to increase pot flow resistance while also moving the fibre failure point away from the fibre - pot interface . fig3 d and 3e show a further method of reducing flow by reducing the inner diameter of the fibre lumen 8 using a layer of material 9 applied to part or whole of the inner surface 11 of the fibre lumen 8 in the region encompassed by the pot 6 . one method of providing such a layer 9 is to coat the inside of the lumen 8 near the end of the pot 6 with a thin layer of material that effectively reduces the diameter of the fibre lumen 8 at this point . this can be achieved by drawing up a material such as epoxy into the end of the fibre lumen 8 and then allowing it to run out again before it has time to set , leaving behind a thin coating 9 on the inner fibre lumen wall 12 that can then set over time . the embodiment shown in fig3 f illustrates smearing the surface of the pot with a suitable grout material 13 to reduce the diameter of the fibre lumen 8 adjacent its opening 14 from the pot 6 . fig3 g shows the insertion of hollow annulus 15 , for example , a hollow pin , into the end of the fibre lumen 8 in the region of the pot 6 to reduce the cross - sectional area of the lumen 8 in the region of the pot 6 . fig3 h shows the use of a porous layer of material 10 across the lumen opening 14 as also shown in the embodiment of fig2 . fig3 i shows an embodiment where a porous material is forced into the lumen opening 14 to form a plug 16 . this can be achieved by smearing a porous grout across and into the fibre lumen opening 14 . again this serves to reduce the flow resistance of the fibre lumen in the region of the pot 6 . fig3 j illustrates an embodiment of the invention where the fibre lumen 8 is narrowed within the region of the pot 6 by causing the potting material to swell or constricting the end of the fibre . fig3 k shows an embodiment where the fibre lumen end is narrowed prior to potting . fig4 shows the results of a test performed on two modules to illustrate the operation of the invention . two modules a and b were used in the test . for each module one hollow fibre membrane was potted . the end of the fibre which was not in the pot , was sealed . a stainless steel mesh was glued on the top of one of the pots in a way that prevented sideways flow of feed bypass during filtration in a similar manner to the embodiments shown in fig2 and 3h . the mesh had openings of 51 microns and was 56 microns thick . the characteristics of both of the modules are shown in table 1 . firstly , feed water was filtered through module a for 35 minutes . during this filtration , the transmembrane pressure ( tmp ) was measured . then the fibre of module a was cut as close to the pot as possible and module a filtered the same feed water for a further 35 minutes . during this filtration , the transmembrane pressure ( tmp ) was measured . the same test was repeated with the module b using the same feed water . the graph shown in fig4 compares the tmp of the modules a and b during the two filtrations before and after the fibre was cut . the first part of the graph shows that the two curves are very similar . in particular , it shows that tmp of both modules increased at the same rate . fibres of the modules were fouled at a similar rate . the small difference in tmp between the two modules is due to the mesh on module b which adds a small extra resistance to flow . the second part of the graph after the fibre of modules was cut shows that tmp of module a and b developed in a highly different way . the tmp of module a remained low and level whereas the tmp of module b increased sharply showing that the mesh was blocked by the feed contaminants . this test clearly shows the efficiency of a mesh as far as reduction of integrity loss is concerned . due to the addition of the mesh to the module , the cut fibre quickly sealed itself , preventing the feed from contaminating the filtrate . it will be apparent to those skilled in the art that a wide variety and number of techniques can be used to reduce the flow within the fibre lumen in the region of the pot and that such techniques fall within the scope of the invention described . it will also be appreciated that further embodiments and exemplifications of the invention are possible without departing from the spirit or scope of the invention described ."}
|
{"patent": "referring to fig2 of the drawings , one preferred embodiment of the invention is illustrated . a sinter or porous layer 10 is placed on top of the pot 6 to provide a further series resistance r pot2 to the pot i . e . an appropriate sinter 10 may have openings of microns in dimension and only be a few millimeters thick . this method may reduce the q b / q i by a factor of 10 . such an arrangement provides added benefits when used for membrane filter systems in a bio - reactor . the high solids feed in bio reactors leads to the sludge actually plugging the filter and self sealing the broken fibre totally . the may be extended to the general case by replacing the sinter with a membrane with the same pore size as the hollow fibre membrane and enabling achievement of this self plugging capability even with low solids feeds . it will be apparent the extra resistance of the sinter or membrane 10 will require an extra pressure to maintain the module filtrate flow , however , this is only an operating cost not a membrane process operating efficiency as it is operating over the pot assembly , not across the compressible dirt layer on the membrane . fouling of this membrane sinter can be reduced by a regular chemical cleaning backwash with chlorine or other suitable cleaners . the membrane / sinter 10 is desirably in intimate contact with the pot 6 to prevent sideways flow of filtrate / feed bypass . this may also be achieved with a replaceable sinter / membrane element . a highly asymmetric membrane 10 with the large pore side contacting the pot 6 ( so in normal filtrate flow the filtrate flows in the direction of reducing pore size ) is desirable . as shown in fig3 b - 3k a variety of methods may be used to increase the pot flow resistance . referring to fig3 a a normal pot 6 without modification is shown . fig3 b shows an increased length pot 6 which , while increasing pot flow resistance , has other disadvantages . fig3 c illustrates providing the fibre 5 with a non porous coating 7 adjacent the interface 8 between the fibre 5 and the pot 6 . this serves to increase pot flow resistance while also moving the fibre failure point away from the fibre - pot interface . fig3 d and 3e show a further method of reducing flow by reducing the inner diameter of the fibre lumen 8 using a layer of material 9 applied to part or whole of the inner surface 11 of the fibre lumen 8 in the region encompassed by the pot 6 . one method of providing such a layer 9 is to coat the inside of the lumen 8 near the end of the pot 6 with a thin layer of material that effectively reduces the diameter of the fibre lumen 8 at this point . this can be achieved by drawing up a material such as epoxy into the end of the fibre lumen 8 and then allowing it to run out again before it has time to set , leaving behind a thin coating 9 on the inner fibre lumen wall 12 that can then set over time . the embodiment shown in fig3 f illustrates smearing the surface of the pot with a suitable grout material 13 to reduce the diameter of the fibre lumen 8 adjacent its opening 14 from the pot 6 . fig3 g shows the insertion of hollow annulus 15 , for example , a hollow pin , into the end of the fibre lumen 8 in the region of the pot 6 to reduce the cross - sectional area of the lumen 8 in the region of the pot 6 . fig3 h shows the use of a porous layer of material 10 across the lumen opening 14 as also shown in the embodiment of fig2 . fig3 i shows an embodiment where a porous material is forced into the lumen opening 14 to form a plug 16 . this can be achieved by smearing a porous grout across and into the fibre lumen opening 14 . again this serves to reduce the flow resistance of the fibre lumen in the region of the pot 6 . fig3 j illustrates an embodiment of the invention where the fibre lumen 8 is narrowed within the region of the pot 6 by causing the potting material to swell or constricting the end of the fibre . fig3 k shows an embodiment where the fibre lumen end is narrowed prior to potting . fig4 shows the results of a test performed on two modules to illustrate the operation of the invention . two modules a and b were used in the test . for each module one hollow fibre membrane was potted . the end of the fibre which was not in the pot , was sealed . a stainless steel mesh was glued on the top of one of the pots in a way that prevented sideways flow of feed bypass during filtration in a similar manner to the embodiments shown in fig2 and 3h . the mesh had openings of 51 microns and was 56 microns thick . the characteristics of both of the modules are shown in table 1 . firstly , feed water was filtered through module a for 35 minutes . during this filtration , the transmembrane pressure ( tmp ) was measured . then the fibre of module a was cut as close to the pot as possible and module a filtered the same feed water for a further 35 minutes . during this filtration , the transmembrane pressure ( tmp ) was measured . the same test was repeated with the module b using the same feed water . the graph shown in fig4 compares the tmp of the modules a and b during the two filtrations before and after the fibre was cut . the first part of the graph shows that the two curves are very similar . in particular , it shows that tmp of both modules increased at the same rate . fibres of the modules were fouled at a similar rate . the small difference in tmp between the two modules is due to the mesh on module b which adds a small extra resistance to flow . the second part of the graph after the fibre of modules was cut shows that tmp of module a and b developed in a highly different way . the tmp of module a remained low and level whereas the tmp of module b increased sharply showing that the mesh was blocked by the feed contaminants . this test clearly shows the efficiency of a mesh as far as reduction of integrity loss is concerned . due to the addition of the mesh to the module , the cut fibre quickly sealed itself , preventing the feed from contaminating the filtrate . it will be apparent to those skilled in the art that a wide variety and number of techniques can be used to reduce the flow within the fibre lumen in the region of the pot and that such techniques fall within the scope of the invention described . it will also be appreciated that further embodiments and exemplifications of the invention are possible without departing from the spirit or scope of the invention described .", "category": "Human Necessities"}
|
Is the category the most suitable category for the given patent?
| 0.25 |
e19771a1309c1c24434e8f0ba17c4335b81021d51478af9c418ca7f4636745d0
| 0.068359 | 0.000231 | 0.017456 | 0.003937 | 0.143555 | 0.015442 |
null |
{"patent": "referring to fig2 of the drawings , one preferred embodiment of the invention is illustrated . a sinter or porous layer 10 is placed on top of the pot 6 to provide a further series resistance r pot2 to the pot i . e . an appropriate sinter 10 may have openings of microns in dimension and only be a few millimeters thick . this method may reduce the q b / q i by a factor of 10 . such an arrangement provides added benefits when used for membrane filter systems in a bio - reactor . the high solids feed in bio reactors leads to the sludge actually plugging the filter and self sealing the broken fibre totally . the may be extended to the general case by replacing the sinter with a membrane with the same pore size as the hollow fibre membrane and enabling achievement of this self plugging capability even with low solids feeds . it will be apparent the extra resistance of the sinter or membrane 10 will require an extra pressure to maintain the module filtrate flow , however , this is only an operating cost not a membrane process operating efficiency as it is operating over the pot assembly , not across the compressible dirt layer on the membrane . fouling of this membrane sinter can be reduced by a regular chemical cleaning backwash with chlorine or other suitable cleaners . the membrane / sinter 10 is desirably in intimate contact with the pot 6 to prevent sideways flow of filtrate / feed bypass . this may also be achieved with a replaceable sinter / membrane element . a highly asymmetric membrane 10 with the large pore side contacting the pot 6 ( so in normal filtrate flow the filtrate flows in the direction of reducing pore size ) is desirable . as shown in fig3 b - 3k a variety of methods may be used to increase the pot flow resistance . referring to fig3 a a normal pot 6 without modification is shown . fig3 b shows an increased length pot 6 which , while increasing pot flow resistance , has other disadvantages . fig3 c illustrates providing the fibre 5 with a non porous coating 7 adjacent the interface 8 between the fibre 5 and the pot 6 . this serves to increase pot flow resistance while also moving the fibre failure point away from the fibre - pot interface . fig3 d and 3e show a further method of reducing flow by reducing the inner diameter of the fibre lumen 8 using a layer of material 9 applied to part or whole of the inner surface 11 of the fibre lumen 8 in the region encompassed by the pot 6 . one method of providing such a layer 9 is to coat the inside of the lumen 8 near the end of the pot 6 with a thin layer of material that effectively reduces the diameter of the fibre lumen 8 at this point . this can be achieved by drawing up a material such as epoxy into the end of the fibre lumen 8 and then allowing it to run out again before it has time to set , leaving behind a thin coating 9 on the inner fibre lumen wall 12 that can then set over time . the embodiment shown in fig3 f illustrates smearing the surface of the pot with a suitable grout material 13 to reduce the diameter of the fibre lumen 8 adjacent its opening 14 from the pot 6 . fig3 g shows the insertion of hollow annulus 15 , for example , a hollow pin , into the end of the fibre lumen 8 in the region of the pot 6 to reduce the cross - sectional area of the lumen 8 in the region of the pot 6 . fig3 h shows the use of a porous layer of material 10 across the lumen opening 14 as also shown in the embodiment of fig2 . fig3 i shows an embodiment where a porous material is forced into the lumen opening 14 to form a plug 16 . this can be achieved by smearing a porous grout across and into the fibre lumen opening 14 . again this serves to reduce the flow resistance of the fibre lumen in the region of the pot 6 . fig3 j illustrates an embodiment of the invention where the fibre lumen 8 is narrowed within the region of the pot 6 by causing the potting material to swell or constricting the end of the fibre . fig3 k shows an embodiment where the fibre lumen end is narrowed prior to potting . fig4 shows the results of a test performed on two modules to illustrate the operation of the invention . two modules a and b were used in the test . for each module one hollow fibre membrane was potted . the end of the fibre which was not in the pot , was sealed . a stainless steel mesh was glued on the top of one of the pots in a way that prevented sideways flow of feed bypass during filtration in a similar manner to the embodiments shown in fig2 and 3h . the mesh had openings of 51 microns and was 56 microns thick . the characteristics of both of the modules are shown in table 1 . firstly , feed water was filtered through module a for 35 minutes . during this filtration , the transmembrane pressure ( tmp ) was measured . then the fibre of module a was cut as close to the pot as possible and module a filtered the same feed water for a further 35 minutes . during this filtration , the transmembrane pressure ( tmp ) was measured . the same test was repeated with the module b using the same feed water . the graph shown in fig4 compares the tmp of the modules a and b during the two filtrations before and after the fibre was cut . the first part of the graph shows that the two curves are very similar . in particular , it shows that tmp of both modules increased at the same rate . fibres of the modules were fouled at a similar rate . the small difference in tmp between the two modules is due to the mesh on module b which adds a small extra resistance to flow . the second part of the graph after the fibre of modules was cut shows that tmp of module a and b developed in a highly different way . the tmp of module a remained low and level whereas the tmp of module b increased sharply showing that the mesh was blocked by the feed contaminants . this test clearly shows the efficiency of a mesh as far as reduction of integrity loss is concerned . due to the addition of the mesh to the module , the cut fibre quickly sealed itself , preventing the feed from contaminating the filtrate . it will be apparent to those skilled in the art that a wide variety and number of techniques can be used to reduce the flow within the fibre lumen in the region of the pot and that such techniques fall within the scope of the invention described . it will also be appreciated that further embodiments and exemplifications of the invention are possible without departing from the spirit or scope of the invention described .", "category": "Performing Operations; Transporting"}
|
{"patent": "referring to fig2 of the drawings , one preferred embodiment of the invention is illustrated . a sinter or porous layer 10 is placed on top of the pot 6 to provide a further series resistance r pot2 to the pot i . e . an appropriate sinter 10 may have openings of microns in dimension and only be a few millimeters thick . this method may reduce the q b / q i by a factor of 10 . such an arrangement provides added benefits when used for membrane filter systems in a bio - reactor . the high solids feed in bio reactors leads to the sludge actually plugging the filter and self sealing the broken fibre totally . the may be extended to the general case by replacing the sinter with a membrane with the same pore size as the hollow fibre membrane and enabling achievement of this self plugging capability even with low solids feeds . it will be apparent the extra resistance of the sinter or membrane 10 will require an extra pressure to maintain the module filtrate flow , however , this is only an operating cost not a membrane process operating efficiency as it is operating over the pot assembly , not across the compressible dirt layer on the membrane . fouling of this membrane sinter can be reduced by a regular chemical cleaning backwash with chlorine or other suitable cleaners . the membrane / sinter 10 is desirably in intimate contact with the pot 6 to prevent sideways flow of filtrate / feed bypass . this may also be achieved with a replaceable sinter / membrane element . a highly asymmetric membrane 10 with the large pore side contacting the pot 6 ( so in normal filtrate flow the filtrate flows in the direction of reducing pore size ) is desirable . as shown in fig3 b - 3k a variety of methods may be used to increase the pot flow resistance . referring to fig3 a a normal pot 6 without modification is shown . fig3 b shows an increased length pot 6 which , while increasing pot flow resistance , has other disadvantages . fig3 c illustrates providing the fibre 5 with a non porous coating 7 adjacent the interface 8 between the fibre 5 and the pot 6 . this serves to increase pot flow resistance while also moving the fibre failure point away from the fibre - pot interface . fig3 d and 3e show a further method of reducing flow by reducing the inner diameter of the fibre lumen 8 using a layer of material 9 applied to part or whole of the inner surface 11 of the fibre lumen 8 in the region encompassed by the pot 6 . one method of providing such a layer 9 is to coat the inside of the lumen 8 near the end of the pot 6 with a thin layer of material that effectively reduces the diameter of the fibre lumen 8 at this point . this can be achieved by drawing up a material such as epoxy into the end of the fibre lumen 8 and then allowing it to run out again before it has time to set , leaving behind a thin coating 9 on the inner fibre lumen wall 12 that can then set over time . the embodiment shown in fig3 f illustrates smearing the surface of the pot with a suitable grout material 13 to reduce the diameter of the fibre lumen 8 adjacent its opening 14 from the pot 6 . fig3 g shows the insertion of hollow annulus 15 , for example , a hollow pin , into the end of the fibre lumen 8 in the region of the pot 6 to reduce the cross - sectional area of the lumen 8 in the region of the pot 6 . fig3 h shows the use of a porous layer of material 10 across the lumen opening 14 as also shown in the embodiment of fig2 . fig3 i shows an embodiment where a porous material is forced into the lumen opening 14 to form a plug 16 . this can be achieved by smearing a porous grout across and into the fibre lumen opening 14 . again this serves to reduce the flow resistance of the fibre lumen in the region of the pot 6 . fig3 j illustrates an embodiment of the invention where the fibre lumen 8 is narrowed within the region of the pot 6 by causing the potting material to swell or constricting the end of the fibre . fig3 k shows an embodiment where the fibre lumen end is narrowed prior to potting . fig4 shows the results of a test performed on two modules to illustrate the operation of the invention . two modules a and b were used in the test . for each module one hollow fibre membrane was potted . the end of the fibre which was not in the pot , was sealed . a stainless steel mesh was glued on the top of one of the pots in a way that prevented sideways flow of feed bypass during filtration in a similar manner to the embodiments shown in fig2 and 3h . the mesh had openings of 51 microns and was 56 microns thick . the characteristics of both of the modules are shown in table 1 . firstly , feed water was filtered through module a for 35 minutes . during this filtration , the transmembrane pressure ( tmp ) was measured . then the fibre of module a was cut as close to the pot as possible and module a filtered the same feed water for a further 35 minutes . during this filtration , the transmembrane pressure ( tmp ) was measured . the same test was repeated with the module b using the same feed water . the graph shown in fig4 compares the tmp of the modules a and b during the two filtrations before and after the fibre was cut . the first part of the graph shows that the two curves are very similar . in particular , it shows that tmp of both modules increased at the same rate . fibres of the modules were fouled at a similar rate . the small difference in tmp between the two modules is due to the mesh on module b which adds a small extra resistance to flow . the second part of the graph after the fibre of modules was cut shows that tmp of module a and b developed in a highly different way . the tmp of module a remained low and level whereas the tmp of module b increased sharply showing that the mesh was blocked by the feed contaminants . this test clearly shows the efficiency of a mesh as far as reduction of integrity loss is concerned . due to the addition of the mesh to the module , the cut fibre quickly sealed itself , preventing the feed from contaminating the filtrate . it will be apparent to those skilled in the art that a wide variety and number of techniques can be used to reduce the flow within the fibre lumen in the region of the pot and that such techniques fall within the scope of the invention described . it will also be appreciated that further embodiments and exemplifications of the invention are possible without departing from the spirit or scope of the invention described .", "category": "Chemistry; Metallurgy"}
|
Does the category match the content of the patent?
| 0.25 |
e19771a1309c1c24434e8f0ba17c4335b81021d51478af9c418ca7f4636745d0
| 0.031738 | 0.003708 | 0.083984 | 0.043457 | 0.177734 | 0.111328 |
null |
{"patent": "referring to fig2 of the drawings , one preferred embodiment of the invention is illustrated . a sinter or porous layer 10 is placed on top of the pot 6 to provide a further series resistance r pot2 to the pot i . e . an appropriate sinter 10 may have openings of microns in dimension and only be a few millimeters thick . this method may reduce the q b / q i by a factor of 10 . such an arrangement provides added benefits when used for membrane filter systems in a bio - reactor . the high solids feed in bio reactors leads to the sludge actually plugging the filter and self sealing the broken fibre totally . the may be extended to the general case by replacing the sinter with a membrane with the same pore size as the hollow fibre membrane and enabling achievement of this self plugging capability even with low solids feeds . it will be apparent the extra resistance of the sinter or membrane 10 will require an extra pressure to maintain the module filtrate flow , however , this is only an operating cost not a membrane process operating efficiency as it is operating over the pot assembly , not across the compressible dirt layer on the membrane . fouling of this membrane sinter can be reduced by a regular chemical cleaning backwash with chlorine or other suitable cleaners . the membrane / sinter 10 is desirably in intimate contact with the pot 6 to prevent sideways flow of filtrate / feed bypass . this may also be achieved with a replaceable sinter / membrane element . a highly asymmetric membrane 10 with the large pore side contacting the pot 6 ( so in normal filtrate flow the filtrate flows in the direction of reducing pore size ) is desirable . as shown in fig3 b - 3k a variety of methods may be used to increase the pot flow resistance . referring to fig3 a a normal pot 6 without modification is shown . fig3 b shows an increased length pot 6 which , while increasing pot flow resistance , has other disadvantages . fig3 c illustrates providing the fibre 5 with a non porous coating 7 adjacent the interface 8 between the fibre 5 and the pot 6 . this serves to increase pot flow resistance while also moving the fibre failure point away from the fibre - pot interface . fig3 d and 3e show a further method of reducing flow by reducing the inner diameter of the fibre lumen 8 using a layer of material 9 applied to part or whole of the inner surface 11 of the fibre lumen 8 in the region encompassed by the pot 6 . one method of providing such a layer 9 is to coat the inside of the lumen 8 near the end of the pot 6 with a thin layer of material that effectively reduces the diameter of the fibre lumen 8 at this point . this can be achieved by drawing up a material such as epoxy into the end of the fibre lumen 8 and then allowing it to run out again before it has time to set , leaving behind a thin coating 9 on the inner fibre lumen wall 12 that can then set over time . the embodiment shown in fig3 f illustrates smearing the surface of the pot with a suitable grout material 13 to reduce the diameter of the fibre lumen 8 adjacent its opening 14 from the pot 6 . fig3 g shows the insertion of hollow annulus 15 , for example , a hollow pin , into the end of the fibre lumen 8 in the region of the pot 6 to reduce the cross - sectional area of the lumen 8 in the region of the pot 6 . fig3 h shows the use of a porous layer of material 10 across the lumen opening 14 as also shown in the embodiment of fig2 . fig3 i shows an embodiment where a porous material is forced into the lumen opening 14 to form a plug 16 . this can be achieved by smearing a porous grout across and into the fibre lumen opening 14 . again this serves to reduce the flow resistance of the fibre lumen in the region of the pot 6 . fig3 j illustrates an embodiment of the invention where the fibre lumen 8 is narrowed within the region of the pot 6 by causing the potting material to swell or constricting the end of the fibre . fig3 k shows an embodiment where the fibre lumen end is narrowed prior to potting . fig4 shows the results of a test performed on two modules to illustrate the operation of the invention . two modules a and b were used in the test . for each module one hollow fibre membrane was potted . the end of the fibre which was not in the pot , was sealed . a stainless steel mesh was glued on the top of one of the pots in a way that prevented sideways flow of feed bypass during filtration in a similar manner to the embodiments shown in fig2 and 3h . the mesh had openings of 51 microns and was 56 microns thick . the characteristics of both of the modules are shown in table 1 . firstly , feed water was filtered through module a for 35 minutes . during this filtration , the transmembrane pressure ( tmp ) was measured . then the fibre of module a was cut as close to the pot as possible and module a filtered the same feed water for a further 35 minutes . during this filtration , the transmembrane pressure ( tmp ) was measured . the same test was repeated with the module b using the same feed water . the graph shown in fig4 compares the tmp of the modules a and b during the two filtrations before and after the fibre was cut . the first part of the graph shows that the two curves are very similar . in particular , it shows that tmp of both modules increased at the same rate . fibres of the modules were fouled at a similar rate . the small difference in tmp between the two modules is due to the mesh on module b which adds a small extra resistance to flow . the second part of the graph after the fibre of modules was cut shows that tmp of module a and b developed in a highly different way . the tmp of module a remained low and level whereas the tmp of module b increased sharply showing that the mesh was blocked by the feed contaminants . this test clearly shows the efficiency of a mesh as far as reduction of integrity loss is concerned . due to the addition of the mesh to the module , the cut fibre quickly sealed itself , preventing the feed from contaminating the filtrate . it will be apparent to those skilled in the art that a wide variety and number of techniques can be used to reduce the flow within the fibre lumen in the region of the pot and that such techniques fall within the scope of the invention described . it will also be appreciated that further embodiments and exemplifications of the invention are possible without departing from the spirit or scope of the invention described .", "category": "Performing Operations; Transporting"}
|
{"patent": "referring to fig2 of the drawings , one preferred embodiment of the invention is illustrated . a sinter or porous layer 10 is placed on top of the pot 6 to provide a further series resistance r pot2 to the pot i . e . an appropriate sinter 10 may have openings of microns in dimension and only be a few millimeters thick . this method may reduce the q b / q i by a factor of 10 . such an arrangement provides added benefits when used for membrane filter systems in a bio - reactor . the high solids feed in bio reactors leads to the sludge actually plugging the filter and self sealing the broken fibre totally . the may be extended to the general case by replacing the sinter with a membrane with the same pore size as the hollow fibre membrane and enabling achievement of this self plugging capability even with low solids feeds . it will be apparent the extra resistance of the sinter or membrane 10 will require an extra pressure to maintain the module filtrate flow , however , this is only an operating cost not a membrane process operating efficiency as it is operating over the pot assembly , not across the compressible dirt layer on the membrane . fouling of this membrane sinter can be reduced by a regular chemical cleaning backwash with chlorine or other suitable cleaners . the membrane / sinter 10 is desirably in intimate contact with the pot 6 to prevent sideways flow of filtrate / feed bypass . this may also be achieved with a replaceable sinter / membrane element . a highly asymmetric membrane 10 with the large pore side contacting the pot 6 ( so in normal filtrate flow the filtrate flows in the direction of reducing pore size ) is desirable . as shown in fig3 b - 3k a variety of methods may be used to increase the pot flow resistance . referring to fig3 a a normal pot 6 without modification is shown . fig3 b shows an increased length pot 6 which , while increasing pot flow resistance , has other disadvantages . fig3 c illustrates providing the fibre 5 with a non porous coating 7 adjacent the interface 8 between the fibre 5 and the pot 6 . this serves to increase pot flow resistance while also moving the fibre failure point away from the fibre - pot interface . fig3 d and 3e show a further method of reducing flow by reducing the inner diameter of the fibre lumen 8 using a layer of material 9 applied to part or whole of the inner surface 11 of the fibre lumen 8 in the region encompassed by the pot 6 . one method of providing such a layer 9 is to coat the inside of the lumen 8 near the end of the pot 6 with a thin layer of material that effectively reduces the diameter of the fibre lumen 8 at this point . this can be achieved by drawing up a material such as epoxy into the end of the fibre lumen 8 and then allowing it to run out again before it has time to set , leaving behind a thin coating 9 on the inner fibre lumen wall 12 that can then set over time . the embodiment shown in fig3 f illustrates smearing the surface of the pot with a suitable grout material 13 to reduce the diameter of the fibre lumen 8 adjacent its opening 14 from the pot 6 . fig3 g shows the insertion of hollow annulus 15 , for example , a hollow pin , into the end of the fibre lumen 8 in the region of the pot 6 to reduce the cross - sectional area of the lumen 8 in the region of the pot 6 . fig3 h shows the use of a porous layer of material 10 across the lumen opening 14 as also shown in the embodiment of fig2 . fig3 i shows an embodiment where a porous material is forced into the lumen opening 14 to form a plug 16 . this can be achieved by smearing a porous grout across and into the fibre lumen opening 14 . again this serves to reduce the flow resistance of the fibre lumen in the region of the pot 6 . fig3 j illustrates an embodiment of the invention where the fibre lumen 8 is narrowed within the region of the pot 6 by causing the potting material to swell or constricting the end of the fibre . fig3 k shows an embodiment where the fibre lumen end is narrowed prior to potting . fig4 shows the results of a test performed on two modules to illustrate the operation of the invention . two modules a and b were used in the test . for each module one hollow fibre membrane was potted . the end of the fibre which was not in the pot , was sealed . a stainless steel mesh was glued on the top of one of the pots in a way that prevented sideways flow of feed bypass during filtration in a similar manner to the embodiments shown in fig2 and 3h . the mesh had openings of 51 microns and was 56 microns thick . the characteristics of both of the modules are shown in table 1 . firstly , feed water was filtered through module a for 35 minutes . during this filtration , the transmembrane pressure ( tmp ) was measured . then the fibre of module a was cut as close to the pot as possible and module a filtered the same feed water for a further 35 minutes . during this filtration , the transmembrane pressure ( tmp ) was measured . the same test was repeated with the module b using the same feed water . the graph shown in fig4 compares the tmp of the modules a and b during the two filtrations before and after the fibre was cut . the first part of the graph shows that the two curves are very similar . in particular , it shows that tmp of both modules increased at the same rate . fibres of the modules were fouled at a similar rate . the small difference in tmp between the two modules is due to the mesh on module b which adds a small extra resistance to flow . the second part of the graph after the fibre of modules was cut shows that tmp of module a and b developed in a highly different way . the tmp of module a remained low and level whereas the tmp of module b increased sharply showing that the mesh was blocked by the feed contaminants . this test clearly shows the efficiency of a mesh as far as reduction of integrity loss is concerned . due to the addition of the mesh to the module , the cut fibre quickly sealed itself , preventing the feed from contaminating the filtrate . it will be apparent to those skilled in the art that a wide variety and number of techniques can be used to reduce the flow within the fibre lumen in the region of the pot and that such techniques fall within the scope of the invention described . it will also be appreciated that further embodiments and exemplifications of the invention are possible without departing from the spirit or scope of the invention described .", "category": "Textiles; Paper"}
|
Does the category match the content of the patent?
| 0.25 |
e19771a1309c1c24434e8f0ba17c4335b81021d51478af9c418ca7f4636745d0
| 0.031738 | 0.007813 | 0.083984 | 0.003601 | 0.177734 | 0.010315 |
null |
{"patent": "referring to fig2 of the drawings , one preferred embodiment of the invention is illustrated . a sinter or porous layer 10 is placed on top of the pot 6 to provide a further series resistance r pot2 to the pot i . e . an appropriate sinter 10 may have openings of microns in dimension and only be a few millimeters thick . this method may reduce the q b / q i by a factor of 10 . such an arrangement provides added benefits when used for membrane filter systems in a bio - reactor . the high solids feed in bio reactors leads to the sludge actually plugging the filter and self sealing the broken fibre totally . the may be extended to the general case by replacing the sinter with a membrane with the same pore size as the hollow fibre membrane and enabling achievement of this self plugging capability even with low solids feeds . it will be apparent the extra resistance of the sinter or membrane 10 will require an extra pressure to maintain the module filtrate flow , however , this is only an operating cost not a membrane process operating efficiency as it is operating over the pot assembly , not across the compressible dirt layer on the membrane . fouling of this membrane sinter can be reduced by a regular chemical cleaning backwash with chlorine or other suitable cleaners . the membrane / sinter 10 is desirably in intimate contact with the pot 6 to prevent sideways flow of filtrate / feed bypass . this may also be achieved with a replaceable sinter / membrane element . a highly asymmetric membrane 10 with the large pore side contacting the pot 6 ( so in normal filtrate flow the filtrate flows in the direction of reducing pore size ) is desirable . as shown in fig3 b - 3k a variety of methods may be used to increase the pot flow resistance . referring to fig3 a a normal pot 6 without modification is shown . fig3 b shows an increased length pot 6 which , while increasing pot flow resistance , has other disadvantages . fig3 c illustrates providing the fibre 5 with a non porous coating 7 adjacent the interface 8 between the fibre 5 and the pot 6 . this serves to increase pot flow resistance while also moving the fibre failure point away from the fibre - pot interface . fig3 d and 3e show a further method of reducing flow by reducing the inner diameter of the fibre lumen 8 using a layer of material 9 applied to part or whole of the inner surface 11 of the fibre lumen 8 in the region encompassed by the pot 6 . one method of providing such a layer 9 is to coat the inside of the lumen 8 near the end of the pot 6 with a thin layer of material that effectively reduces the diameter of the fibre lumen 8 at this point . this can be achieved by drawing up a material such as epoxy into the end of the fibre lumen 8 and then allowing it to run out again before it has time to set , leaving behind a thin coating 9 on the inner fibre lumen wall 12 that can then set over time . the embodiment shown in fig3 f illustrates smearing the surface of the pot with a suitable grout material 13 to reduce the diameter of the fibre lumen 8 adjacent its opening 14 from the pot 6 . fig3 g shows the insertion of hollow annulus 15 , for example , a hollow pin , into the end of the fibre lumen 8 in the region of the pot 6 to reduce the cross - sectional area of the lumen 8 in the region of the pot 6 . fig3 h shows the use of a porous layer of material 10 across the lumen opening 14 as also shown in the embodiment of fig2 . fig3 i shows an embodiment where a porous material is forced into the lumen opening 14 to form a plug 16 . this can be achieved by smearing a porous grout across and into the fibre lumen opening 14 . again this serves to reduce the flow resistance of the fibre lumen in the region of the pot 6 . fig3 j illustrates an embodiment of the invention where the fibre lumen 8 is narrowed within the region of the pot 6 by causing the potting material to swell or constricting the end of the fibre . fig3 k shows an embodiment where the fibre lumen end is narrowed prior to potting . fig4 shows the results of a test performed on two modules to illustrate the operation of the invention . two modules a and b were used in the test . for each module one hollow fibre membrane was potted . the end of the fibre which was not in the pot , was sealed . a stainless steel mesh was glued on the top of one of the pots in a way that prevented sideways flow of feed bypass during filtration in a similar manner to the embodiments shown in fig2 and 3h . the mesh had openings of 51 microns and was 56 microns thick . the characteristics of both of the modules are shown in table 1 . firstly , feed water was filtered through module a for 35 minutes . during this filtration , the transmembrane pressure ( tmp ) was measured . then the fibre of module a was cut as close to the pot as possible and module a filtered the same feed water for a further 35 minutes . during this filtration , the transmembrane pressure ( tmp ) was measured . the same test was repeated with the module b using the same feed water . the graph shown in fig4 compares the tmp of the modules a and b during the two filtrations before and after the fibre was cut . the first part of the graph shows that the two curves are very similar . in particular , it shows that tmp of both modules increased at the same rate . fibres of the modules were fouled at a similar rate . the small difference in tmp between the two modules is due to the mesh on module b which adds a small extra resistance to flow . the second part of the graph after the fibre of modules was cut shows that tmp of module a and b developed in a highly different way . the tmp of module a remained low and level whereas the tmp of module b increased sharply showing that the mesh was blocked by the feed contaminants . this test clearly shows the efficiency of a mesh as far as reduction of integrity loss is concerned . due to the addition of the mesh to the module , the cut fibre quickly sealed itself , preventing the feed from contaminating the filtrate . it will be apparent to those skilled in the art that a wide variety and number of techniques can be used to reduce the flow within the fibre lumen in the region of the pot and that such techniques fall within the scope of the invention described . it will also be appreciated that further embodiments and exemplifications of the invention are possible without departing from the spirit or scope of the invention described .", "category": "Performing Operations; Transporting"}
|
{"patent": "referring to fig2 of the drawings , one preferred embodiment of the invention is illustrated . a sinter or porous layer 10 is placed on top of the pot 6 to provide a further series resistance r pot2 to the pot i . e . an appropriate sinter 10 may have openings of microns in dimension and only be a few millimeters thick . this method may reduce the q b / q i by a factor of 10 . such an arrangement provides added benefits when used for membrane filter systems in a bio - reactor . the high solids feed in bio reactors leads to the sludge actually plugging the filter and self sealing the broken fibre totally . the may be extended to the general case by replacing the sinter with a membrane with the same pore size as the hollow fibre membrane and enabling achievement of this self plugging capability even with low solids feeds . it will be apparent the extra resistance of the sinter or membrane 10 will require an extra pressure to maintain the module filtrate flow , however , this is only an operating cost not a membrane process operating efficiency as it is operating over the pot assembly , not across the compressible dirt layer on the membrane . fouling of this membrane sinter can be reduced by a regular chemical cleaning backwash with chlorine or other suitable cleaners . the membrane / sinter 10 is desirably in intimate contact with the pot 6 to prevent sideways flow of filtrate / feed bypass . this may also be achieved with a replaceable sinter / membrane element . a highly asymmetric membrane 10 with the large pore side contacting the pot 6 ( so in normal filtrate flow the filtrate flows in the direction of reducing pore size ) is desirable . as shown in fig3 b - 3k a variety of methods may be used to increase the pot flow resistance . referring to fig3 a a normal pot 6 without modification is shown . fig3 b shows an increased length pot 6 which , while increasing pot flow resistance , has other disadvantages . fig3 c illustrates providing the fibre 5 with a non porous coating 7 adjacent the interface 8 between the fibre 5 and the pot 6 . this serves to increase pot flow resistance while also moving the fibre failure point away from the fibre - pot interface . fig3 d and 3e show a further method of reducing flow by reducing the inner diameter of the fibre lumen 8 using a layer of material 9 applied to part or whole of the inner surface 11 of the fibre lumen 8 in the region encompassed by the pot 6 . one method of providing such a layer 9 is to coat the inside of the lumen 8 near the end of the pot 6 with a thin layer of material that effectively reduces the diameter of the fibre lumen 8 at this point . this can be achieved by drawing up a material such as epoxy into the end of the fibre lumen 8 and then allowing it to run out again before it has time to set , leaving behind a thin coating 9 on the inner fibre lumen wall 12 that can then set over time . the embodiment shown in fig3 f illustrates smearing the surface of the pot with a suitable grout material 13 to reduce the diameter of the fibre lumen 8 adjacent its opening 14 from the pot 6 . fig3 g shows the insertion of hollow annulus 15 , for example , a hollow pin , into the end of the fibre lumen 8 in the region of the pot 6 to reduce the cross - sectional area of the lumen 8 in the region of the pot 6 . fig3 h shows the use of a porous layer of material 10 across the lumen opening 14 as also shown in the embodiment of fig2 . fig3 i shows an embodiment where a porous material is forced into the lumen opening 14 to form a plug 16 . this can be achieved by smearing a porous grout across and into the fibre lumen opening 14 . again this serves to reduce the flow resistance of the fibre lumen in the region of the pot 6 . fig3 j illustrates an embodiment of the invention where the fibre lumen 8 is narrowed within the region of the pot 6 by causing the potting material to swell or constricting the end of the fibre . fig3 k shows an embodiment where the fibre lumen end is narrowed prior to potting . fig4 shows the results of a test performed on two modules to illustrate the operation of the invention . two modules a and b were used in the test . for each module one hollow fibre membrane was potted . the end of the fibre which was not in the pot , was sealed . a stainless steel mesh was glued on the top of one of the pots in a way that prevented sideways flow of feed bypass during filtration in a similar manner to the embodiments shown in fig2 and 3h . the mesh had openings of 51 microns and was 56 microns thick . the characteristics of both of the modules are shown in table 1 . firstly , feed water was filtered through module a for 35 minutes . during this filtration , the transmembrane pressure ( tmp ) was measured . then the fibre of module a was cut as close to the pot as possible and module a filtered the same feed water for a further 35 minutes . during this filtration , the transmembrane pressure ( tmp ) was measured . the same test was repeated with the module b using the same feed water . the graph shown in fig4 compares the tmp of the modules a and b during the two filtrations before and after the fibre was cut . the first part of the graph shows that the two curves are very similar . in particular , it shows that tmp of both modules increased at the same rate . fibres of the modules were fouled at a similar rate . the small difference in tmp between the two modules is due to the mesh on module b which adds a small extra resistance to flow . the second part of the graph after the fibre of modules was cut shows that tmp of module a and b developed in a highly different way . the tmp of module a remained low and level whereas the tmp of module b increased sharply showing that the mesh was blocked by the feed contaminants . this test clearly shows the efficiency of a mesh as far as reduction of integrity loss is concerned . due to the addition of the mesh to the module , the cut fibre quickly sealed itself , preventing the feed from contaminating the filtrate . it will be apparent to those skilled in the art that a wide variety and number of techniques can be used to reduce the flow within the fibre lumen in the region of the pot and that such techniques fall within the scope of the invention described . it will also be appreciated that further embodiments and exemplifications of the invention are possible without departing from the spirit or scope of the invention described .", "category": "Fixed Constructions"}
|
Is the categorization of this patent accurate?
| 0.25 |
e19771a1309c1c24434e8f0ba17c4335b81021d51478af9c418ca7f4636745d0
| 0.044678 | 0.675781 | 0.182617 | 0.439453 | 0.179688 | 0.53125 |
null |
{"patent": "referring to fig2 of the drawings , one preferred embodiment of the invention is illustrated . a sinter or porous layer 10 is placed on top of the pot 6 to provide a further series resistance r pot2 to the pot i . e . an appropriate sinter 10 may have openings of microns in dimension and only be a few millimeters thick . this method may reduce the q b / q i by a factor of 10 . such an arrangement provides added benefits when used for membrane filter systems in a bio - reactor . the high solids feed in bio reactors leads to the sludge actually plugging the filter and self sealing the broken fibre totally . the may be extended to the general case by replacing the sinter with a membrane with the same pore size as the hollow fibre membrane and enabling achievement of this self plugging capability even with low solids feeds . it will be apparent the extra resistance of the sinter or membrane 10 will require an extra pressure to maintain the module filtrate flow , however , this is only an operating cost not a membrane process operating efficiency as it is operating over the pot assembly , not across the compressible dirt layer on the membrane . fouling of this membrane sinter can be reduced by a regular chemical cleaning backwash with chlorine or other suitable cleaners . the membrane / sinter 10 is desirably in intimate contact with the pot 6 to prevent sideways flow of filtrate / feed bypass . this may also be achieved with a replaceable sinter / membrane element . a highly asymmetric membrane 10 with the large pore side contacting the pot 6 ( so in normal filtrate flow the filtrate flows in the direction of reducing pore size ) is desirable . as shown in fig3 b - 3k a variety of methods may be used to increase the pot flow resistance . referring to fig3 a a normal pot 6 without modification is shown . fig3 b shows an increased length pot 6 which , while increasing pot flow resistance , has other disadvantages . fig3 c illustrates providing the fibre 5 with a non porous coating 7 adjacent the interface 8 between the fibre 5 and the pot 6 . this serves to increase pot flow resistance while also moving the fibre failure point away from the fibre - pot interface . fig3 d and 3e show a further method of reducing flow by reducing the inner diameter of the fibre lumen 8 using a layer of material 9 applied to part or whole of the inner surface 11 of the fibre lumen 8 in the region encompassed by the pot 6 . one method of providing such a layer 9 is to coat the inside of the lumen 8 near the end of the pot 6 with a thin layer of material that effectively reduces the diameter of the fibre lumen 8 at this point . this can be achieved by drawing up a material such as epoxy into the end of the fibre lumen 8 and then allowing it to run out again before it has time to set , leaving behind a thin coating 9 on the inner fibre lumen wall 12 that can then set over time . the embodiment shown in fig3 f illustrates smearing the surface of the pot with a suitable grout material 13 to reduce the diameter of the fibre lumen 8 adjacent its opening 14 from the pot 6 . fig3 g shows the insertion of hollow annulus 15 , for example , a hollow pin , into the end of the fibre lumen 8 in the region of the pot 6 to reduce the cross - sectional area of the lumen 8 in the region of the pot 6 . fig3 h shows the use of a porous layer of material 10 across the lumen opening 14 as also shown in the embodiment of fig2 . fig3 i shows an embodiment where a porous material is forced into the lumen opening 14 to form a plug 16 . this can be achieved by smearing a porous grout across and into the fibre lumen opening 14 . again this serves to reduce the flow resistance of the fibre lumen in the region of the pot 6 . fig3 j illustrates an embodiment of the invention where the fibre lumen 8 is narrowed within the region of the pot 6 by causing the potting material to swell or constricting the end of the fibre . fig3 k shows an embodiment where the fibre lumen end is narrowed prior to potting . fig4 shows the results of a test performed on two modules to illustrate the operation of the invention . two modules a and b were used in the test . for each module one hollow fibre membrane was potted . the end of the fibre which was not in the pot , was sealed . a stainless steel mesh was glued on the top of one of the pots in a way that prevented sideways flow of feed bypass during filtration in a similar manner to the embodiments shown in fig2 and 3h . the mesh had openings of 51 microns and was 56 microns thick . the characteristics of both of the modules are shown in table 1 . firstly , feed water was filtered through module a for 35 minutes . during this filtration , the transmembrane pressure ( tmp ) was measured . then the fibre of module a was cut as close to the pot as possible and module a filtered the same feed water for a further 35 minutes . during this filtration , the transmembrane pressure ( tmp ) was measured . the same test was repeated with the module b using the same feed water . the graph shown in fig4 compares the tmp of the modules a and b during the two filtrations before and after the fibre was cut . the first part of the graph shows that the two curves are very similar . in particular , it shows that tmp of both modules increased at the same rate . fibres of the modules were fouled at a similar rate . the small difference in tmp between the two modules is due to the mesh on module b which adds a small extra resistance to flow . the second part of the graph after the fibre of modules was cut shows that tmp of module a and b developed in a highly different way . the tmp of module a remained low and level whereas the tmp of module b increased sharply showing that the mesh was blocked by the feed contaminants . this test clearly shows the efficiency of a mesh as far as reduction of integrity loss is concerned . due to the addition of the mesh to the module , the cut fibre quickly sealed itself , preventing the feed from contaminating the filtrate . it will be apparent to those skilled in the art that a wide variety and number of techniques can be used to reduce the flow within the fibre lumen in the region of the pot and that such techniques fall within the scope of the invention described . it will also be appreciated that further embodiments and exemplifications of the invention are possible without departing from the spirit or scope of the invention described .", "category": "Performing Operations; Transporting"}
|
{"patent": "referring to fig2 of the drawings , one preferred embodiment of the invention is illustrated . a sinter or porous layer 10 is placed on top of the pot 6 to provide a further series resistance r pot2 to the pot i . e . an appropriate sinter 10 may have openings of microns in dimension and only be a few millimeters thick . this method may reduce the q b / q i by a factor of 10 . such an arrangement provides added benefits when used for membrane filter systems in a bio - reactor . the high solids feed in bio reactors leads to the sludge actually plugging the filter and self sealing the broken fibre totally . the may be extended to the general case by replacing the sinter with a membrane with the same pore size as the hollow fibre membrane and enabling achievement of this self plugging capability even with low solids feeds . it will be apparent the extra resistance of the sinter or membrane 10 will require an extra pressure to maintain the module filtrate flow , however , this is only an operating cost not a membrane process operating efficiency as it is operating over the pot assembly , not across the compressible dirt layer on the membrane . fouling of this membrane sinter can be reduced by a regular chemical cleaning backwash with chlorine or other suitable cleaners . the membrane / sinter 10 is desirably in intimate contact with the pot 6 to prevent sideways flow of filtrate / feed bypass . this may also be achieved with a replaceable sinter / membrane element . a highly asymmetric membrane 10 with the large pore side contacting the pot 6 ( so in normal filtrate flow the filtrate flows in the direction of reducing pore size ) is desirable . as shown in fig3 b - 3k a variety of methods may be used to increase the pot flow resistance . referring to fig3 a a normal pot 6 without modification is shown . fig3 b shows an increased length pot 6 which , while increasing pot flow resistance , has other disadvantages . fig3 c illustrates providing the fibre 5 with a non porous coating 7 adjacent the interface 8 between the fibre 5 and the pot 6 . this serves to increase pot flow resistance while also moving the fibre failure point away from the fibre - pot interface . fig3 d and 3e show a further method of reducing flow by reducing the inner diameter of the fibre lumen 8 using a layer of material 9 applied to part or whole of the inner surface 11 of the fibre lumen 8 in the region encompassed by the pot 6 . one method of providing such a layer 9 is to coat the inside of the lumen 8 near the end of the pot 6 with a thin layer of material that effectively reduces the diameter of the fibre lumen 8 at this point . this can be achieved by drawing up a material such as epoxy into the end of the fibre lumen 8 and then allowing it to run out again before it has time to set , leaving behind a thin coating 9 on the inner fibre lumen wall 12 that can then set over time . the embodiment shown in fig3 f illustrates smearing the surface of the pot with a suitable grout material 13 to reduce the diameter of the fibre lumen 8 adjacent its opening 14 from the pot 6 . fig3 g shows the insertion of hollow annulus 15 , for example , a hollow pin , into the end of the fibre lumen 8 in the region of the pot 6 to reduce the cross - sectional area of the lumen 8 in the region of the pot 6 . fig3 h shows the use of a porous layer of material 10 across the lumen opening 14 as also shown in the embodiment of fig2 . fig3 i shows an embodiment where a porous material is forced into the lumen opening 14 to form a plug 16 . this can be achieved by smearing a porous grout across and into the fibre lumen opening 14 . again this serves to reduce the flow resistance of the fibre lumen in the region of the pot 6 . fig3 j illustrates an embodiment of the invention where the fibre lumen 8 is narrowed within the region of the pot 6 by causing the potting material to swell or constricting the end of the fibre . fig3 k shows an embodiment where the fibre lumen end is narrowed prior to potting . fig4 shows the results of a test performed on two modules to illustrate the operation of the invention . two modules a and b were used in the test . for each module one hollow fibre membrane was potted . the end of the fibre which was not in the pot , was sealed . a stainless steel mesh was glued on the top of one of the pots in a way that prevented sideways flow of feed bypass during filtration in a similar manner to the embodiments shown in fig2 and 3h . the mesh had openings of 51 microns and was 56 microns thick . the characteristics of both of the modules are shown in table 1 . firstly , feed water was filtered through module a for 35 minutes . during this filtration , the transmembrane pressure ( tmp ) was measured . then the fibre of module a was cut as close to the pot as possible and module a filtered the same feed water for a further 35 minutes . during this filtration , the transmembrane pressure ( tmp ) was measured . the same test was repeated with the module b using the same feed water . the graph shown in fig4 compares the tmp of the modules a and b during the two filtrations before and after the fibre was cut . the first part of the graph shows that the two curves are very similar . in particular , it shows that tmp of both modules increased at the same rate . fibres of the modules were fouled at a similar rate . the small difference in tmp between the two modules is due to the mesh on module b which adds a small extra resistance to flow . the second part of the graph after the fibre of modules was cut shows that tmp of module a and b developed in a highly different way . the tmp of module a remained low and level whereas the tmp of module b increased sharply showing that the mesh was blocked by the feed contaminants . this test clearly shows the efficiency of a mesh as far as reduction of integrity loss is concerned . due to the addition of the mesh to the module , the cut fibre quickly sealed itself , preventing the feed from contaminating the filtrate . it will be apparent to those skilled in the art that a wide variety and number of techniques can be used to reduce the flow within the fibre lumen in the region of the pot and that such techniques fall within the scope of the invention described . it will also be appreciated that further embodiments and exemplifications of the invention are possible without departing from the spirit or scope of the invention described .", "category": "Mechanical Engineering; Lightning; Heating; Weapons; Blasting"}
|
Does the patent belong in this category?
| 0.25 |
e19771a1309c1c24434e8f0ba17c4335b81021d51478af9c418ca7f4636745d0
| 0.025146 | 0.004333 | 0.039063 | 0.00885 | 0.192383 | 0.019165 |
null |
{"patent": "referring to fig2 of the drawings , one preferred embodiment of the invention is illustrated . a sinter or porous layer 10 is placed on top of the pot 6 to provide a further series resistance r pot2 to the pot i . e . an appropriate sinter 10 may have openings of microns in dimension and only be a few millimeters thick . this method may reduce the q b / q i by a factor of 10 . such an arrangement provides added benefits when used for membrane filter systems in a bio - reactor . the high solids feed in bio reactors leads to the sludge actually plugging the filter and self sealing the broken fibre totally . the may be extended to the general case by replacing the sinter with a membrane with the same pore size as the hollow fibre membrane and enabling achievement of this self plugging capability even with low solids feeds . it will be apparent the extra resistance of the sinter or membrane 10 will require an extra pressure to maintain the module filtrate flow , however , this is only an operating cost not a membrane process operating efficiency as it is operating over the pot assembly , not across the compressible dirt layer on the membrane . fouling of this membrane sinter can be reduced by a regular chemical cleaning backwash with chlorine or other suitable cleaners . the membrane / sinter 10 is desirably in intimate contact with the pot 6 to prevent sideways flow of filtrate / feed bypass . this may also be achieved with a replaceable sinter / membrane element . a highly asymmetric membrane 10 with the large pore side contacting the pot 6 ( so in normal filtrate flow the filtrate flows in the direction of reducing pore size ) is desirable . as shown in fig3 b - 3k a variety of methods may be used to increase the pot flow resistance . referring to fig3 a a normal pot 6 without modification is shown . fig3 b shows an increased length pot 6 which , while increasing pot flow resistance , has other disadvantages . fig3 c illustrates providing the fibre 5 with a non porous coating 7 adjacent the interface 8 between the fibre 5 and the pot 6 . this serves to increase pot flow resistance while also moving the fibre failure point away from the fibre - pot interface . fig3 d and 3e show a further method of reducing flow by reducing the inner diameter of the fibre lumen 8 using a layer of material 9 applied to part or whole of the inner surface 11 of the fibre lumen 8 in the region encompassed by the pot 6 . one method of providing such a layer 9 is to coat the inside of the lumen 8 near the end of the pot 6 with a thin layer of material that effectively reduces the diameter of the fibre lumen 8 at this point . this can be achieved by drawing up a material such as epoxy into the end of the fibre lumen 8 and then allowing it to run out again before it has time to set , leaving behind a thin coating 9 on the inner fibre lumen wall 12 that can then set over time . the embodiment shown in fig3 f illustrates smearing the surface of the pot with a suitable grout material 13 to reduce the diameter of the fibre lumen 8 adjacent its opening 14 from the pot 6 . fig3 g shows the insertion of hollow annulus 15 , for example , a hollow pin , into the end of the fibre lumen 8 in the region of the pot 6 to reduce the cross - sectional area of the lumen 8 in the region of the pot 6 . fig3 h shows the use of a porous layer of material 10 across the lumen opening 14 as also shown in the embodiment of fig2 . fig3 i shows an embodiment where a porous material is forced into the lumen opening 14 to form a plug 16 . this can be achieved by smearing a porous grout across and into the fibre lumen opening 14 . again this serves to reduce the flow resistance of the fibre lumen in the region of the pot 6 . fig3 j illustrates an embodiment of the invention where the fibre lumen 8 is narrowed within the region of the pot 6 by causing the potting material to swell or constricting the end of the fibre . fig3 k shows an embodiment where the fibre lumen end is narrowed prior to potting . fig4 shows the results of a test performed on two modules to illustrate the operation of the invention . two modules a and b were used in the test . for each module one hollow fibre membrane was potted . the end of the fibre which was not in the pot , was sealed . a stainless steel mesh was glued on the top of one of the pots in a way that prevented sideways flow of feed bypass during filtration in a similar manner to the embodiments shown in fig2 and 3h . the mesh had openings of 51 microns and was 56 microns thick . the characteristics of both of the modules are shown in table 1 . firstly , feed water was filtered through module a for 35 minutes . during this filtration , the transmembrane pressure ( tmp ) was measured . then the fibre of module a was cut as close to the pot as possible and module a filtered the same feed water for a further 35 minutes . during this filtration , the transmembrane pressure ( tmp ) was measured . the same test was repeated with the module b using the same feed water . the graph shown in fig4 compares the tmp of the modules a and b during the two filtrations before and after the fibre was cut . the first part of the graph shows that the two curves are very similar . in particular , it shows that tmp of both modules increased at the same rate . fibres of the modules were fouled at a similar rate . the small difference in tmp between the two modules is due to the mesh on module b which adds a small extra resistance to flow . the second part of the graph after the fibre of modules was cut shows that tmp of module a and b developed in a highly different way . the tmp of module a remained low and level whereas the tmp of module b increased sharply showing that the mesh was blocked by the feed contaminants . this test clearly shows the efficiency of a mesh as far as reduction of integrity loss is concerned . due to the addition of the mesh to the module , the cut fibre quickly sealed itself , preventing the feed from contaminating the filtrate . it will be apparent to those skilled in the art that a wide variety and number of techniques can be used to reduce the flow within the fibre lumen in the region of the pot and that such techniques fall within the scope of the invention described . it will also be appreciated that further embodiments and exemplifications of the invention are possible without departing from the spirit or scope of the invention described .", "category": "Performing Operations; Transporting"}
|
{"patent": "referring to fig2 of the drawings , one preferred embodiment of the invention is illustrated . a sinter or porous layer 10 is placed on top of the pot 6 to provide a further series resistance r pot2 to the pot i . e . an appropriate sinter 10 may have openings of microns in dimension and only be a few millimeters thick . this method may reduce the q b / q i by a factor of 10 . such an arrangement provides added benefits when used for membrane filter systems in a bio - reactor . the high solids feed in bio reactors leads to the sludge actually plugging the filter and self sealing the broken fibre totally . the may be extended to the general case by replacing the sinter with a membrane with the same pore size as the hollow fibre membrane and enabling achievement of this self plugging capability even with low solids feeds . it will be apparent the extra resistance of the sinter or membrane 10 will require an extra pressure to maintain the module filtrate flow , however , this is only an operating cost not a membrane process operating efficiency as it is operating over the pot assembly , not across the compressible dirt layer on the membrane . fouling of this membrane sinter can be reduced by a regular chemical cleaning backwash with chlorine or other suitable cleaners . the membrane / sinter 10 is desirably in intimate contact with the pot 6 to prevent sideways flow of filtrate / feed bypass . this may also be achieved with a replaceable sinter / membrane element . a highly asymmetric membrane 10 with the large pore side contacting the pot 6 ( so in normal filtrate flow the filtrate flows in the direction of reducing pore size ) is desirable . as shown in fig3 b - 3k a variety of methods may be used to increase the pot flow resistance . referring to fig3 a a normal pot 6 without modification is shown . fig3 b shows an increased length pot 6 which , while increasing pot flow resistance , has other disadvantages . fig3 c illustrates providing the fibre 5 with a non porous coating 7 adjacent the interface 8 between the fibre 5 and the pot 6 . this serves to increase pot flow resistance while also moving the fibre failure point away from the fibre - pot interface . fig3 d and 3e show a further method of reducing flow by reducing the inner diameter of the fibre lumen 8 using a layer of material 9 applied to part or whole of the inner surface 11 of the fibre lumen 8 in the region encompassed by the pot 6 . one method of providing such a layer 9 is to coat the inside of the lumen 8 near the end of the pot 6 with a thin layer of material that effectively reduces the diameter of the fibre lumen 8 at this point . this can be achieved by drawing up a material such as epoxy into the end of the fibre lumen 8 and then allowing it to run out again before it has time to set , leaving behind a thin coating 9 on the inner fibre lumen wall 12 that can then set over time . the embodiment shown in fig3 f illustrates smearing the surface of the pot with a suitable grout material 13 to reduce the diameter of the fibre lumen 8 adjacent its opening 14 from the pot 6 . fig3 g shows the insertion of hollow annulus 15 , for example , a hollow pin , into the end of the fibre lumen 8 in the region of the pot 6 to reduce the cross - sectional area of the lumen 8 in the region of the pot 6 . fig3 h shows the use of a porous layer of material 10 across the lumen opening 14 as also shown in the embodiment of fig2 . fig3 i shows an embodiment where a porous material is forced into the lumen opening 14 to form a plug 16 . this can be achieved by smearing a porous grout across and into the fibre lumen opening 14 . again this serves to reduce the flow resistance of the fibre lumen in the region of the pot 6 . fig3 j illustrates an embodiment of the invention where the fibre lumen 8 is narrowed within the region of the pot 6 by causing the potting material to swell or constricting the end of the fibre . fig3 k shows an embodiment where the fibre lumen end is narrowed prior to potting . fig4 shows the results of a test performed on two modules to illustrate the operation of the invention . two modules a and b were used in the test . for each module one hollow fibre membrane was potted . the end of the fibre which was not in the pot , was sealed . a stainless steel mesh was glued on the top of one of the pots in a way that prevented sideways flow of feed bypass during filtration in a similar manner to the embodiments shown in fig2 and 3h . the mesh had openings of 51 microns and was 56 microns thick . the characteristics of both of the modules are shown in table 1 . firstly , feed water was filtered through module a for 35 minutes . during this filtration , the transmembrane pressure ( tmp ) was measured . then the fibre of module a was cut as close to the pot as possible and module a filtered the same feed water for a further 35 minutes . during this filtration , the transmembrane pressure ( tmp ) was measured . the same test was repeated with the module b using the same feed water . the graph shown in fig4 compares the tmp of the modules a and b during the two filtrations before and after the fibre was cut . the first part of the graph shows that the two curves are very similar . in particular , it shows that tmp of both modules increased at the same rate . fibres of the modules were fouled at a similar rate . the small difference in tmp between the two modules is due to the mesh on module b which adds a small extra resistance to flow . the second part of the graph after the fibre of modules was cut shows that tmp of module a and b developed in a highly different way . the tmp of module a remained low and level whereas the tmp of module b increased sharply showing that the mesh was blocked by the feed contaminants . this test clearly shows the efficiency of a mesh as far as reduction of integrity loss is concerned . due to the addition of the mesh to the module , the cut fibre quickly sealed itself , preventing the feed from contaminating the filtrate . it will be apparent to those skilled in the art that a wide variety and number of techniques can be used to reduce the flow within the fibre lumen in the region of the pot and that such techniques fall within the scope of the invention described . it will also be appreciated that further embodiments and exemplifications of the invention are possible without departing from the spirit or scope of the invention described .", "category": "Physics"}
|
Is the categorization of this patent accurate?
| 0.25 |
e19771a1309c1c24434e8f0ba17c4335b81021d51478af9c418ca7f4636745d0
| 0.047363 | 0.07373 | 0.182617 | 0.098145 | 0.185547 | 0.115723 |
null |
{"category": "Performing Operations; Transporting", "patent": "referring to fig2 of the drawings , one preferred embodiment of the invention is illustrated . a sinter or porous layer 10 is placed on top of the pot 6 to provide a further series resistance r pot2 to the pot i . e . an appropriate sinter 10 may have openings of microns in dimension and only be a few millimeters thick . this method may reduce the q b / q i by a factor of 10 . such an arrangement provides added benefits when used for membrane filter systems in a bio - reactor . the high solids feed in bio reactors leads to the sludge actually plugging the filter and self sealing the broken fibre totally . the may be extended to the general case by replacing the sinter with a membrane with the same pore size as the hollow fibre membrane and enabling achievement of this self plugging capability even with low solids feeds . it will be apparent the extra resistance of the sinter or membrane 10 will require an extra pressure to maintain the module filtrate flow , however , this is only an operating cost not a membrane process operating efficiency as it is operating over the pot assembly , not across the compressible dirt layer on the membrane . fouling of this membrane sinter can be reduced by a regular chemical cleaning backwash with chlorine or other suitable cleaners . the membrane / sinter 10 is desirably in intimate contact with the pot 6 to prevent sideways flow of filtrate / feed bypass . this may also be achieved with a replaceable sinter / membrane element . a highly asymmetric membrane 10 with the large pore side contacting the pot 6 ( so in normal filtrate flow the filtrate flows in the direction of reducing pore size ) is desirable . as shown in fig3 b - 3k a variety of methods may be used to increase the pot flow resistance . referring to fig3 a a normal pot 6 without modification is shown . fig3 b shows an increased length pot 6 which , while increasing pot flow resistance , has other disadvantages . fig3 c illustrates providing the fibre 5 with a non porous coating 7 adjacent the interface 8 between the fibre 5 and the pot 6 . this serves to increase pot flow resistance while also moving the fibre failure point away from the fibre - pot interface . fig3 d and 3e show a further method of reducing flow by reducing the inner diameter of the fibre lumen 8 using a layer of material 9 applied to part or whole of the inner surface 11 of the fibre lumen 8 in the region encompassed by the pot 6 . one method of providing such a layer 9 is to coat the inside of the lumen 8 near the end of the pot 6 with a thin layer of material that effectively reduces the diameter of the fibre lumen 8 at this point . this can be achieved by drawing up a material such as epoxy into the end of the fibre lumen 8 and then allowing it to run out again before it has time to set , leaving behind a thin coating 9 on the inner fibre lumen wall 12 that can then set over time . the embodiment shown in fig3 f illustrates smearing the surface of the pot with a suitable grout material 13 to reduce the diameter of the fibre lumen 8 adjacent its opening 14 from the pot 6 . fig3 g shows the insertion of hollow annulus 15 , for example , a hollow pin , into the end of the fibre lumen 8 in the region of the pot 6 to reduce the cross - sectional area of the lumen 8 in the region of the pot 6 . fig3 h shows the use of a porous layer of material 10 across the lumen opening 14 as also shown in the embodiment of fig2 . fig3 i shows an embodiment where a porous material is forced into the lumen opening 14 to form a plug 16 . this can be achieved by smearing a porous grout across and into the fibre lumen opening 14 . again this serves to reduce the flow resistance of the fibre lumen in the region of the pot 6 . fig3 j illustrates an embodiment of the invention where the fibre lumen 8 is narrowed within the region of the pot 6 by causing the potting material to swell or constricting the end of the fibre . fig3 k shows an embodiment where the fibre lumen end is narrowed prior to potting . fig4 shows the results of a test performed on two modules to illustrate the operation of the invention . two modules a and b were used in the test . for each module one hollow fibre membrane was potted . the end of the fibre which was not in the pot , was sealed . a stainless steel mesh was glued on the top of one of the pots in a way that prevented sideways flow of feed bypass during filtration in a similar manner to the embodiments shown in fig2 and 3h . the mesh had openings of 51 microns and was 56 microns thick . the characteristics of both of the modules are shown in table 1 . firstly , feed water was filtered through module a for 35 minutes . during this filtration , the transmembrane pressure ( tmp ) was measured . then the fibre of module a was cut as close to the pot as possible and module a filtered the same feed water for a further 35 minutes . during this filtration , the transmembrane pressure ( tmp ) was measured . the same test was repeated with the module b using the same feed water . the graph shown in fig4 compares the tmp of the modules a and b during the two filtrations before and after the fibre was cut . the first part of the graph shows that the two curves are very similar . in particular , it shows that tmp of both modules increased at the same rate . fibres of the modules were fouled at a similar rate . the small difference in tmp between the two modules is due to the mesh on module b which adds a small extra resistance to flow . the second part of the graph after the fibre of modules was cut shows that tmp of module a and b developed in a highly different way . the tmp of module a remained low and level whereas the tmp of module b increased sharply showing that the mesh was blocked by the feed contaminants . this test clearly shows the efficiency of a mesh as far as reduction of integrity loss is concerned . due to the addition of the mesh to the module , the cut fibre quickly sealed itself , preventing the feed from contaminating the filtrate . it will be apparent to those skilled in the art that a wide variety and number of techniques can be used to reduce the flow within the fibre lumen in the region of the pot and that such techniques fall within the scope of the invention described . it will also be appreciated that further embodiments and exemplifications of the invention are possible without departing from the spirit or scope of the invention described ."}
|
{"patent": "referring to fig2 of the drawings , one preferred embodiment of the invention is illustrated . a sinter or porous layer 10 is placed on top of the pot 6 to provide a further series resistance r pot2 to the pot i . e . an appropriate sinter 10 may have openings of microns in dimension and only be a few millimeters thick . this method may reduce the q b / q i by a factor of 10 . such an arrangement provides added benefits when used for membrane filter systems in a bio - reactor . the high solids feed in bio reactors leads to the sludge actually plugging the filter and self sealing the broken fibre totally . the may be extended to the general case by replacing the sinter with a membrane with the same pore size as the hollow fibre membrane and enabling achievement of this self plugging capability even with low solids feeds . it will be apparent the extra resistance of the sinter or membrane 10 will require an extra pressure to maintain the module filtrate flow , however , this is only an operating cost not a membrane process operating efficiency as it is operating over the pot assembly , not across the compressible dirt layer on the membrane . fouling of this membrane sinter can be reduced by a regular chemical cleaning backwash with chlorine or other suitable cleaners . the membrane / sinter 10 is desirably in intimate contact with the pot 6 to prevent sideways flow of filtrate / feed bypass . this may also be achieved with a replaceable sinter / membrane element . a highly asymmetric membrane 10 with the large pore side contacting the pot 6 ( so in normal filtrate flow the filtrate flows in the direction of reducing pore size ) is desirable . as shown in fig3 b - 3k a variety of methods may be used to increase the pot flow resistance . referring to fig3 a a normal pot 6 without modification is shown . fig3 b shows an increased length pot 6 which , while increasing pot flow resistance , has other disadvantages . fig3 c illustrates providing the fibre 5 with a non porous coating 7 adjacent the interface 8 between the fibre 5 and the pot 6 . this serves to increase pot flow resistance while also moving the fibre failure point away from the fibre - pot interface . fig3 d and 3e show a further method of reducing flow by reducing the inner diameter of the fibre lumen 8 using a layer of material 9 applied to part or whole of the inner surface 11 of the fibre lumen 8 in the region encompassed by the pot 6 . one method of providing such a layer 9 is to coat the inside of the lumen 8 near the end of the pot 6 with a thin layer of material that effectively reduces the diameter of the fibre lumen 8 at this point . this can be achieved by drawing up a material such as epoxy into the end of the fibre lumen 8 and then allowing it to run out again before it has time to set , leaving behind a thin coating 9 on the inner fibre lumen wall 12 that can then set over time . the embodiment shown in fig3 f illustrates smearing the surface of the pot with a suitable grout material 13 to reduce the diameter of the fibre lumen 8 adjacent its opening 14 from the pot 6 . fig3 g shows the insertion of hollow annulus 15 , for example , a hollow pin , into the end of the fibre lumen 8 in the region of the pot 6 to reduce the cross - sectional area of the lumen 8 in the region of the pot 6 . fig3 h shows the use of a porous layer of material 10 across the lumen opening 14 as also shown in the embodiment of fig2 . fig3 i shows an embodiment where a porous material is forced into the lumen opening 14 to form a plug 16 . this can be achieved by smearing a porous grout across and into the fibre lumen opening 14 . again this serves to reduce the flow resistance of the fibre lumen in the region of the pot 6 . fig3 j illustrates an embodiment of the invention where the fibre lumen 8 is narrowed within the region of the pot 6 by causing the potting material to swell or constricting the end of the fibre . fig3 k shows an embodiment where the fibre lumen end is narrowed prior to potting . fig4 shows the results of a test performed on two modules to illustrate the operation of the invention . two modules a and b were used in the test . for each module one hollow fibre membrane was potted . the end of the fibre which was not in the pot , was sealed . a stainless steel mesh was glued on the top of one of the pots in a way that prevented sideways flow of feed bypass during filtration in a similar manner to the embodiments shown in fig2 and 3h . the mesh had openings of 51 microns and was 56 microns thick . the characteristics of both of the modules are shown in table 1 . firstly , feed water was filtered through module a for 35 minutes . during this filtration , the transmembrane pressure ( tmp ) was measured . then the fibre of module a was cut as close to the pot as possible and module a filtered the same feed water for a further 35 minutes . during this filtration , the transmembrane pressure ( tmp ) was measured . the same test was repeated with the module b using the same feed water . the graph shown in fig4 compares the tmp of the modules a and b during the two filtrations before and after the fibre was cut . the first part of the graph shows that the two curves are very similar . in particular , it shows that tmp of both modules increased at the same rate . fibres of the modules were fouled at a similar rate . the small difference in tmp between the two modules is due to the mesh on module b which adds a small extra resistance to flow . the second part of the graph after the fibre of modules was cut shows that tmp of module a and b developed in a highly different way . the tmp of module a remained low and level whereas the tmp of module b increased sharply showing that the mesh was blocked by the feed contaminants . this test clearly shows the efficiency of a mesh as far as reduction of integrity loss is concerned . due to the addition of the mesh to the module , the cut fibre quickly sealed itself , preventing the feed from contaminating the filtrate . it will be apparent to those skilled in the art that a wide variety and number of techniques can be used to reduce the flow within the fibre lumen in the region of the pot and that such techniques fall within the scope of the invention described . it will also be appreciated that further embodiments and exemplifications of the invention are possible without departing from the spirit or scope of the invention described .", "category": "Electricity"}
|
Is the category the most suitable category for the given patent?
| 0.25 |
e19771a1309c1c24434e8f0ba17c4335b81021d51478af9c418ca7f4636745d0
| 0.068359 | 0.003372 | 0.017456 | 0.005737 | 0.143555 | 0.068359 |
null |
{"patent": "referring to fig2 of the drawings , one preferred embodiment of the invention is illustrated . a sinter or porous layer 10 is placed on top of the pot 6 to provide a further series resistance r pot2 to the pot i . e . an appropriate sinter 10 may have openings of microns in dimension and only be a few millimeters thick . this method may reduce the q b / q i by a factor of 10 . such an arrangement provides added benefits when used for membrane filter systems in a bio - reactor . the high solids feed in bio reactors leads to the sludge actually plugging the filter and self sealing the broken fibre totally . the may be extended to the general case by replacing the sinter with a membrane with the same pore size as the hollow fibre membrane and enabling achievement of this self plugging capability even with low solids feeds . it will be apparent the extra resistance of the sinter or membrane 10 will require an extra pressure to maintain the module filtrate flow , however , this is only an operating cost not a membrane process operating efficiency as it is operating over the pot assembly , not across the compressible dirt layer on the membrane . fouling of this membrane sinter can be reduced by a regular chemical cleaning backwash with chlorine or other suitable cleaners . the membrane / sinter 10 is desirably in intimate contact with the pot 6 to prevent sideways flow of filtrate / feed bypass . this may also be achieved with a replaceable sinter / membrane element . a highly asymmetric membrane 10 with the large pore side contacting the pot 6 ( so in normal filtrate flow the filtrate flows in the direction of reducing pore size ) is desirable . as shown in fig3 b - 3k a variety of methods may be used to increase the pot flow resistance . referring to fig3 a a normal pot 6 without modification is shown . fig3 b shows an increased length pot 6 which , while increasing pot flow resistance , has other disadvantages . fig3 c illustrates providing the fibre 5 with a non porous coating 7 adjacent the interface 8 between the fibre 5 and the pot 6 . this serves to increase pot flow resistance while also moving the fibre failure point away from the fibre - pot interface . fig3 d and 3e show a further method of reducing flow by reducing the inner diameter of the fibre lumen 8 using a layer of material 9 applied to part or whole of the inner surface 11 of the fibre lumen 8 in the region encompassed by the pot 6 . one method of providing such a layer 9 is to coat the inside of the lumen 8 near the end of the pot 6 with a thin layer of material that effectively reduces the diameter of the fibre lumen 8 at this point . this can be achieved by drawing up a material such as epoxy into the end of the fibre lumen 8 and then allowing it to run out again before it has time to set , leaving behind a thin coating 9 on the inner fibre lumen wall 12 that can then set over time . the embodiment shown in fig3 f illustrates smearing the surface of the pot with a suitable grout material 13 to reduce the diameter of the fibre lumen 8 adjacent its opening 14 from the pot 6 . fig3 g shows the insertion of hollow annulus 15 , for example , a hollow pin , into the end of the fibre lumen 8 in the region of the pot 6 to reduce the cross - sectional area of the lumen 8 in the region of the pot 6 . fig3 h shows the use of a porous layer of material 10 across the lumen opening 14 as also shown in the embodiment of fig2 . fig3 i shows an embodiment where a porous material is forced into the lumen opening 14 to form a plug 16 . this can be achieved by smearing a porous grout across and into the fibre lumen opening 14 . again this serves to reduce the flow resistance of the fibre lumen in the region of the pot 6 . fig3 j illustrates an embodiment of the invention where the fibre lumen 8 is narrowed within the region of the pot 6 by causing the potting material to swell or constricting the end of the fibre . fig3 k shows an embodiment where the fibre lumen end is narrowed prior to potting . fig4 shows the results of a test performed on two modules to illustrate the operation of the invention . two modules a and b were used in the test . for each module one hollow fibre membrane was potted . the end of the fibre which was not in the pot , was sealed . a stainless steel mesh was glued on the top of one of the pots in a way that prevented sideways flow of feed bypass during filtration in a similar manner to the embodiments shown in fig2 and 3h . the mesh had openings of 51 microns and was 56 microns thick . the characteristics of both of the modules are shown in table 1 . firstly , feed water was filtered through module a for 35 minutes . during this filtration , the transmembrane pressure ( tmp ) was measured . then the fibre of module a was cut as close to the pot as possible and module a filtered the same feed water for a further 35 minutes . during this filtration , the transmembrane pressure ( tmp ) was measured . the same test was repeated with the module b using the same feed water . the graph shown in fig4 compares the tmp of the modules a and b during the two filtrations before and after the fibre was cut . the first part of the graph shows that the two curves are very similar . in particular , it shows that tmp of both modules increased at the same rate . fibres of the modules were fouled at a similar rate . the small difference in tmp between the two modules is due to the mesh on module b which adds a small extra resistance to flow . the second part of the graph after the fibre of modules was cut shows that tmp of module a and b developed in a highly different way . the tmp of module a remained low and level whereas the tmp of module b increased sharply showing that the mesh was blocked by the feed contaminants . this test clearly shows the efficiency of a mesh as far as reduction of integrity loss is concerned . due to the addition of the mesh to the module , the cut fibre quickly sealed itself , preventing the feed from contaminating the filtrate . it will be apparent to those skilled in the art that a wide variety and number of techniques can be used to reduce the flow within the fibre lumen in the region of the pot and that such techniques fall within the scope of the invention described . it will also be appreciated that further embodiments and exemplifications of the invention are possible without departing from the spirit or scope of the invention described .", "category": "Performing Operations; Transporting"}
|
{"patent": "referring to fig2 of the drawings , one preferred embodiment of the invention is illustrated . a sinter or porous layer 10 is placed on top of the pot 6 to provide a further series resistance r pot2 to the pot i . e . an appropriate sinter 10 may have openings of microns in dimension and only be a few millimeters thick . this method may reduce the q b / q i by a factor of 10 . such an arrangement provides added benefits when used for membrane filter systems in a bio - reactor . the high solids feed in bio reactors leads to the sludge actually plugging the filter and self sealing the broken fibre totally . the may be extended to the general case by replacing the sinter with a membrane with the same pore size as the hollow fibre membrane and enabling achievement of this self plugging capability even with low solids feeds . it will be apparent the extra resistance of the sinter or membrane 10 will require an extra pressure to maintain the module filtrate flow , however , this is only an operating cost not a membrane process operating efficiency as it is operating over the pot assembly , not across the compressible dirt layer on the membrane . fouling of this membrane sinter can be reduced by a regular chemical cleaning backwash with chlorine or other suitable cleaners . the membrane / sinter 10 is desirably in intimate contact with the pot 6 to prevent sideways flow of filtrate / feed bypass . this may also be achieved with a replaceable sinter / membrane element . a highly asymmetric membrane 10 with the large pore side contacting the pot 6 ( so in normal filtrate flow the filtrate flows in the direction of reducing pore size ) is desirable . as shown in fig3 b - 3k a variety of methods may be used to increase the pot flow resistance . referring to fig3 a a normal pot 6 without modification is shown . fig3 b shows an increased length pot 6 which , while increasing pot flow resistance , has other disadvantages . fig3 c illustrates providing the fibre 5 with a non porous coating 7 adjacent the interface 8 between the fibre 5 and the pot 6 . this serves to increase pot flow resistance while also moving the fibre failure point away from the fibre - pot interface . fig3 d and 3e show a further method of reducing flow by reducing the inner diameter of the fibre lumen 8 using a layer of material 9 applied to part or whole of the inner surface 11 of the fibre lumen 8 in the region encompassed by the pot 6 . one method of providing such a layer 9 is to coat the inside of the lumen 8 near the end of the pot 6 with a thin layer of material that effectively reduces the diameter of the fibre lumen 8 at this point . this can be achieved by drawing up a material such as epoxy into the end of the fibre lumen 8 and then allowing it to run out again before it has time to set , leaving behind a thin coating 9 on the inner fibre lumen wall 12 that can then set over time . the embodiment shown in fig3 f illustrates smearing the surface of the pot with a suitable grout material 13 to reduce the diameter of the fibre lumen 8 adjacent its opening 14 from the pot 6 . fig3 g shows the insertion of hollow annulus 15 , for example , a hollow pin , into the end of the fibre lumen 8 in the region of the pot 6 to reduce the cross - sectional area of the lumen 8 in the region of the pot 6 . fig3 h shows the use of a porous layer of material 10 across the lumen opening 14 as also shown in the embodiment of fig2 . fig3 i shows an embodiment where a porous material is forced into the lumen opening 14 to form a plug 16 . this can be achieved by smearing a porous grout across and into the fibre lumen opening 14 . again this serves to reduce the flow resistance of the fibre lumen in the region of the pot 6 . fig3 j illustrates an embodiment of the invention where the fibre lumen 8 is narrowed within the region of the pot 6 by causing the potting material to swell or constricting the end of the fibre . fig3 k shows an embodiment where the fibre lumen end is narrowed prior to potting . fig4 shows the results of a test performed on two modules to illustrate the operation of the invention . two modules a and b were used in the test . for each module one hollow fibre membrane was potted . the end of the fibre which was not in the pot , was sealed . a stainless steel mesh was glued on the top of one of the pots in a way that prevented sideways flow of feed bypass during filtration in a similar manner to the embodiments shown in fig2 and 3h . the mesh had openings of 51 microns and was 56 microns thick . the characteristics of both of the modules are shown in table 1 . firstly , feed water was filtered through module a for 35 minutes . during this filtration , the transmembrane pressure ( tmp ) was measured . then the fibre of module a was cut as close to the pot as possible and module a filtered the same feed water for a further 35 minutes . during this filtration , the transmembrane pressure ( tmp ) was measured . the same test was repeated with the module b using the same feed water . the graph shown in fig4 compares the tmp of the modules a and b during the two filtrations before and after the fibre was cut . the first part of the graph shows that the two curves are very similar . in particular , it shows that tmp of both modules increased at the same rate . fibres of the modules were fouled at a similar rate . the small difference in tmp between the two modules is due to the mesh on module b which adds a small extra resistance to flow . the second part of the graph after the fibre of modules was cut shows that tmp of module a and b developed in a highly different way . the tmp of module a remained low and level whereas the tmp of module b increased sharply showing that the mesh was blocked by the feed contaminants . this test clearly shows the efficiency of a mesh as far as reduction of integrity loss is concerned . due to the addition of the mesh to the module , the cut fibre quickly sealed itself , preventing the feed from contaminating the filtrate . it will be apparent to those skilled in the art that a wide variety and number of techniques can be used to reduce the flow within the fibre lumen in the region of the pot and that such techniques fall within the scope of the invention described . it will also be appreciated that further embodiments and exemplifications of the invention are possible without departing from the spirit or scope of the invention described .", "category": "General tagging of new or cross-sectional technology"}
|
Does the category match the content of the patent?
| 0.25 |
e19771a1309c1c24434e8f0ba17c4335b81021d51478af9c418ca7f4636745d0
| 0.031738 | 0.03418 | 0.083984 | 0.117676 | 0.177734 | 0.074707 |
null |
{"category": "Physics", "patent": "fig1 a depicts the inventive instruction cache ( icache ) 100 of a processor , and includes long cache lines 101 , 102 , 103 , and 104 . only four lines are depicted for simplicity , icache 100 size is implementation dependent . each cache line includes a tag 105 , which is used to tell a cache hit or miss , a plurality of instruction bundles 106 , and counter / branch information 107 . fig1 b depicts the contents of instruction bundles 106 . each bundle is comprised of a group of instructions 108 that can be issued in the same cycle , for example , bundle 0 includes a load instruction , an add instruction , and a compare - branch instruction . note that each bundle has a fixed number of instructions , however , some of the instructions may be nops . fig1 c depicts the counter / branch information associated with each instruction bundle . each instruction bundle has counter information 109 , which is used to determine whether the code within the bundle is hot code . when the bundle is brought into the icache , the counter is initialized to a threshold value . depending on the threshold value desired , the counter can be as small as 8 to 10 bits . the counter is updated when the instruction bundle is retired from the execution pipeline . each update decrements the counter by 1 . note that the counter could initially be set to zero and increment with each retirement . however , this would require a comparison with a non - zero threshold number , e . g . 100 , which requires more work than comparing with a zero threshold number . each instruction bundle 106 in the icache 100 also maintains a branch history 110 , 111 for each instruction within the bundle . this history describes whether the comparisons in the branch instructions have resulted in a fall through to the next instruction or a branch taken to another instruction . branch history 110 is associated with bundle 0 , including slots a , b , c , which correspond to the instructions within the bundle 0 . thus , it appears one slot in the history is allocated for each instruction in the bundle , whether the instruction is a branch instruction or not . when the instructions from the original binary are brought into the icache , the branch history is cleared . the branch history information is updated when the instruction bundle is retired from fie pipeline . note that the number of instructions ( and thus the number of slots ) is by way of example only , as each bundle could have more or fewer instructions . since the third instruction in bundle 0 is a branch instruction , then slot 110 c has branch information . binary zeros indicate a fall through , and binary ones indicate a branch taken . thus , the information in 110 c , i . e . 00100 , indicates that of the last five times that this instruction has been executed , that the instruction br 1 has fallen through , fallen through , been taken , fallen through , and fallen through . note that the number of bits in the history is by way of example only , and more bits could be used to provide a more detailed history ( while requiring more space ), while fewer bits could be used to save space ( while providing less history ). note that either the most significant bit or the last significant bit may represent the most urgent execution instruction . similarly , the information in 111 b and 111 c describe the histories of instruction br 2 and br 3 respectively . note that br 2 has not recently branched , whereas the previous four executions of br 3 have resulted in the branch taken . in operation , once the counter of a bundle reaches zero , a software component known as the trace selector 201 is invoked , via a special trap , to select a trace . diagnose instructions ( special instructions to diagnose hardware ) are used by the trace selector to examine the icache and the branch history information to form a trace . regular instructions cannot read i - cache contents since i - cache is not part of the architecture states . each processor has a set of diagnose instructions defined ( not visible to application programmer ) which can be used to examine i - cache contents . fig2 a and 2b depict trace formation . fig2 a depicts instruction bundles 106 and their associated branch information 11110 . assume that the counter of bundle 1 ( not shown ) has reached zero , and that bundles 5 - 9 and 14 - 99 are not shown for reason of simplicity . note that bundles 1 - 101 may be in one or more cache lines of icache 100 . the trace selector 201 begins building the trace 202 from the hot code , in this case bundle 1 . the trace selector 201 examines the branch information ( if any ) in bundle 1 to predict whether the branch will be taken or fall through . if there are no branch instructions in the bundle , then bundle will fall through to the next sequential bundle . if the trace selector determines that the branch is most likely to fall through , then the next sequential bundle is added to the trace 202 , in this case it would be bundle 2 . note that if a branch instruction that is in the middle of bundle is assumed to be taken , the remaining instructions of the bundle are not included in the trace . the trace 202 is stored in the trace memory 203 . if the trace selector determines that the branch is most likely to be taken , then the target bundle of the branch is added to the trace 202 , in this case it would be bundle 30 . after examining the branch history 112 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 30 will not be taken , and will add the next sequential bundle , bundle 2 , to the trace 202 , and then will examine bundle 2 . after examining the branch history 113 of bundle 2 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken four times and fallen through once . therefore , the trace selector will predict that the branch to bundle 10 will be taken , and will add the target bundle , bundle 10 , to the trace 202 , and then will examine bundle 10 . after examining the branch history 114 of bundle 10 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 20 will not be taken , and will add the next sequential bundle , bundle 11 , to the trace 202 , and then will examine bundle 11 . bundle 11 does not contain any branch instructions , and therefore will not have a branch history , thus the trace selector 201 will add the next sequential bundle , bundle 12 , to the trace 202 , and then will examine bundle 12 . after examining the branch history 115 of bundle 12 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 24 will not be taken , and will add the next sequential bundle , bundle 13 , to the trace 202 , and then will examine bundle 13 . after examining the branch history 116 of bundle 13 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken four times and fallen through once . therefore , the trace selector will predict that the branch to bundle 101 will be taken , and will add the target bundle , bundle 101 , to the trace 202 , and then will examine bundle 101 . after examining the branch history 117 of bundle 101 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken five times . therefore , the trace selector will predict that the branch to bundle 1 will be taken . the trace selector notes that bundle 1 is already part of the trace 202 in trace memory 203 , via the trace of a sequence of bundles , by examining the address of a backward branch , it can be detected whether the target bundle is already part of a trace . the trace selector then ends the trace or passes the formed trace to the optimizer . the branch to bundle 1 from bundle 101 is known as a backward branch , which forms a loop . at this point , the trace may be stopped , as the trace would merely repeat bundles that are already present in the trace . the trace selector may also end the trace based on other criteria from a set of heuristics including the length of the trace , the number of conditional branches encountered , the probability of accumulated branch predictions and other considerations . thus , a trace may end when its length is a multiple of a cache line size . this would make cache operations easier , as the entire line could be loaded or overwritten without having to be concerned about starting and stopping points in the middle of a cache line . the trace could also end after a certain , predetermined number of conditional branches has been encountered . note that branch histories 113 and 116 do indicate that branch falls through occasionally , and thus the trace would be inaccurate as the trace predicts that the branch will be taken . the predetermined number could be based on the probability of error of the trace . for example , the predetermined number would be low if many of the branches have histories of 00011 or 00111 . on the other hand , the predetermined number would be high if many of the branches have histories of 00000 or 11111 . note that a trace may terminate at an indirect branch since the target address is not known . an indirect branch is different from an ip - relative ( or pc - relative ) branch in that the branch target address cannot be computed directly from the branch instruction . its target is stored either in a register or in a memory location . so the target address is unknown unless the instruction is actually executed . for example , however , the trace selector may decide to grow the trace by predicting its most recent target from the target address cache ( tac ), which is a structure commonly used to predict branch target address . for a return branch which is an indirect branch , with its target being dependent on the call site , the trace selector would know the return address if the call instruction is in the trace , if the call instruction is not in the trace , the trace selector can predict the call site using the top address of the return stack buffer ( rsb ), which is a commonly used data structure to predict return branches . the tac and the rsb are discussed in the co - pending and commonly assigned u . s . patent application entitled efficient mapping to optimized code for processor embedded run - time optimizer ser . no . 09 / 252 , 367 , filed feb . 18 , 1999 , and issued feb . 6 , 2001 , as u . s . pat . no . 6 , 185 , 669 ; which is hereby incorporated by reference . the trace 202 will be stored in the trace memory 203 . there is a mapping from the trace starting instruction bundle in the original binary to the trace in the trace memory . when the trace starting bundle is executed , the mapping will automatically lead the execution to the trace stored in the trace memory 203 . typically , an executed branch instruction has its target in the trace memory . this is discussed in the co - pending and commonly assigned u . s . patent application entitled system and method using a hardware embedded run - time optimizer ser . no . 09 / 252 , 170 , filed feb . 18 , 1999 , and issued sep . 17 , 2002 , as u . s . pat . no . 6 , 453 , 411 , which is hereby incorporated by reference . note that the trace may require more than one cache line . as stated previously , long cache lines are inefficient for original binary . this is because the original binary is loaded sequentially , i . e . bundle 1 , 2 , 3 , 4 , 5 , 6 , etc ., and branches taken within the bundles may result in many of the loaded bundles not being used . for example , suppose bundle 6 has a branch taken to bundle 50 , then the loading of bundles 7 - 49 represent wasted time and cache space as they are not going to be used . however , when the trace is loaded into the cache , the entire trace is almost certain to be used . thus , the long cache lines are much more efficient , because of the sequential locality , as the bundles of the trace will ( almost always ) fall through to the next bundle of the trace . note that a trace usually spans several cache lines . it may not end at the end of a cache line . in this case , the remaining part of the cache line can be the start of another trace . note that since traces are also brought into the icache , the profiling and trace selection may end up generating a trace on top of an existing trace . traces can be identified since their addresses are preserved addresses in physical memory . if their participation in subsequent trace selection is not desired , then when the trace is moved into the icache , the counters associated with the trace will not be initialized to the threshold value , and instead are set to a null value . thus , the trace will not participate in profiling . however , subsequent profiling and trace selection could be used to determine whether the trace is considered \u201c good .\u201d for example , if a trace has frequent early exits , then the trace may need to be regenerated . note that more bits of branch history will allow for more accurate predictions to be made by the trace selector . however , this will require more cache space . alternatively , a multi - tiered system may be used such that the trace selector would not to select a trace when a bundle traps for the first time . instead , the trace selector may record the branch history information of the bundle in another location of memory , and then set the threshold back to a second value , which could be smaller , larger or the same as the original threshold value , and return to execution . when this bundle traps again , the trace selector can accumulate the current branch history with the branch history from the first trap to make more accurate branch predictions . although the present invention and its advantages have been described in detail , it should be understood that various changes , substitutions and alterations can be made herein without departing from the spirit and scope of the invention as defined by the appended claims ."}
|
{"category": "Human Necessities", "patent": "fig1 a depicts the inventive instruction cache ( icache ) 100 of a processor , and includes long cache lines 101 , 102 , 103 , and 104 . only four lines are depicted for simplicity , icache 100 size is implementation dependent . each cache line includes a tag 105 , which is used to tell a cache hit or miss , a plurality of instruction bundles 106 , and counter / branch information 107 . fig1 b depicts the contents of instruction bundles 106 . each bundle is comprised of a group of instructions 108 that can be issued in the same cycle , for example , bundle 0 includes a load instruction , an add instruction , and a compare - branch instruction . note that each bundle has a fixed number of instructions , however , some of the instructions may be nops . fig1 c depicts the counter / branch information associated with each instruction bundle . each instruction bundle has counter information 109 , which is used to determine whether the code within the bundle is hot code . when the bundle is brought into the icache , the counter is initialized to a threshold value . depending on the threshold value desired , the counter can be as small as 8 to 10 bits . the counter is updated when the instruction bundle is retired from the execution pipeline . each update decrements the counter by 1 . note that the counter could initially be set to zero and increment with each retirement . however , this would require a comparison with a non - zero threshold number , e . g . 100 , which requires more work than comparing with a zero threshold number . each instruction bundle 106 in the icache 100 also maintains a branch history 110 , 111 for each instruction within the bundle . this history describes whether the comparisons in the branch instructions have resulted in a fall through to the next instruction or a branch taken to another instruction . branch history 110 is associated with bundle 0 , including slots a , b , c , which correspond to the instructions within the bundle 0 . thus , it appears one slot in the history is allocated for each instruction in the bundle , whether the instruction is a branch instruction or not . when the instructions from the original binary are brought into the icache , the branch history is cleared . the branch history information is updated when the instruction bundle is retired from fie pipeline . note that the number of instructions ( and thus the number of slots ) is by way of example only , as each bundle could have more or fewer instructions . since the third instruction in bundle 0 is a branch instruction , then slot 110 c has branch information . binary zeros indicate a fall through , and binary ones indicate a branch taken . thus , the information in 110 c , i . e . 00100 , indicates that of the last five times that this instruction has been executed , that the instruction br 1 has fallen through , fallen through , been taken , fallen through , and fallen through . note that the number of bits in the history is by way of example only , and more bits could be used to provide a more detailed history ( while requiring more space ), while fewer bits could be used to save space ( while providing less history ). note that either the most significant bit or the last significant bit may represent the most urgent execution instruction . similarly , the information in 111 b and 111 c describe the histories of instruction br 2 and br 3 respectively . note that br 2 has not recently branched , whereas the previous four executions of br 3 have resulted in the branch taken . in operation , once the counter of a bundle reaches zero , a software component known as the trace selector 201 is invoked , via a special trap , to select a trace . diagnose instructions ( special instructions to diagnose hardware ) are used by the trace selector to examine the icache and the branch history information to form a trace . regular instructions cannot read i - cache contents since i - cache is not part of the architecture states . each processor has a set of diagnose instructions defined ( not visible to application programmer ) which can be used to examine i - cache contents . fig2 a and 2b depict trace formation . fig2 a depicts instruction bundles 106 and their associated branch information 11110 . assume that the counter of bundle 1 ( not shown ) has reached zero , and that bundles 5 - 9 and 14 - 99 are not shown for reason of simplicity . note that bundles 1 - 101 may be in one or more cache lines of icache 100 . the trace selector 201 begins building the trace 202 from the hot code , in this case bundle 1 . the trace selector 201 examines the branch information ( if any ) in bundle 1 to predict whether the branch will be taken or fall through . if there are no branch instructions in the bundle , then bundle will fall through to the next sequential bundle . if the trace selector determines that the branch is most likely to fall through , then the next sequential bundle is added to the trace 202 , in this case it would be bundle 2 . note that if a branch instruction that is in the middle of bundle is assumed to be taken , the remaining instructions of the bundle are not included in the trace . the trace 202 is stored in the trace memory 203 . if the trace selector determines that the branch is most likely to be taken , then the target bundle of the branch is added to the trace 202 , in this case it would be bundle 30 . after examining the branch history 112 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 30 will not be taken , and will add the next sequential bundle , bundle 2 , to the trace 202 , and then will examine bundle 2 . after examining the branch history 113 of bundle 2 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken four times and fallen through once . therefore , the trace selector will predict that the branch to bundle 10 will be taken , and will add the target bundle , bundle 10 , to the trace 202 , and then will examine bundle 10 . after examining the branch history 114 of bundle 10 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 20 will not be taken , and will add the next sequential bundle , bundle 11 , to the trace 202 , and then will examine bundle 11 . bundle 11 does not contain any branch instructions , and therefore will not have a branch history , thus the trace selector 201 will add the next sequential bundle , bundle 12 , to the trace 202 , and then will examine bundle 12 . after examining the branch history 115 of bundle 12 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 24 will not be taken , and will add the next sequential bundle , bundle 13 , to the trace 202 , and then will examine bundle 13 . after examining the branch history 116 of bundle 13 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken four times and fallen through once . therefore , the trace selector will predict that the branch to bundle 101 will be taken , and will add the target bundle , bundle 101 , to the trace 202 , and then will examine bundle 101 . after examining the branch history 117 of bundle 101 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken five times . therefore , the trace selector will predict that the branch to bundle 1 will be taken . the trace selector notes that bundle 1 is already part of the trace 202 in trace memory 203 , via the trace of a sequence of bundles , by examining the address of a backward branch , it can be detected whether the target bundle is already part of a trace . the trace selector then ends the trace or passes the formed trace to the optimizer . the branch to bundle 1 from bundle 101 is known as a backward branch , which forms a loop . at this point , the trace may be stopped , as the trace would merely repeat bundles that are already present in the trace . the trace selector may also end the trace based on other criteria from a set of heuristics including the length of the trace , the number of conditional branches encountered , the probability of accumulated branch predictions and other considerations . thus , a trace may end when its length is a multiple of a cache line size . this would make cache operations easier , as the entire line could be loaded or overwritten without having to be concerned about starting and stopping points in the middle of a cache line . the trace could also end after a certain , predetermined number of conditional branches has been encountered . note that branch histories 113 and 116 do indicate that branch falls through occasionally , and thus the trace would be inaccurate as the trace predicts that the branch will be taken . the predetermined number could be based on the probability of error of the trace . for example , the predetermined number would be low if many of the branches have histories of 00011 or 00111 . on the other hand , the predetermined number would be high if many of the branches have histories of 00000 or 11111 . note that a trace may terminate at an indirect branch since the target address is not known . an indirect branch is different from an ip - relative ( or pc - relative ) branch in that the branch target address cannot be computed directly from the branch instruction . its target is stored either in a register or in a memory location . so the target address is unknown unless the instruction is actually executed . for example , however , the trace selector may decide to grow the trace by predicting its most recent target from the target address cache ( tac ), which is a structure commonly used to predict branch target address . for a return branch which is an indirect branch , with its target being dependent on the call site , the trace selector would know the return address if the call instruction is in the trace , if the call instruction is not in the trace , the trace selector can predict the call site using the top address of the return stack buffer ( rsb ), which is a commonly used data structure to predict return branches . the tac and the rsb are discussed in the co - pending and commonly assigned u . s . patent application entitled efficient mapping to optimized code for processor embedded run - time optimizer ser . no . 09 / 252 , 367 , filed feb . 18 , 1999 , and issued feb . 6 , 2001 , as u . s . pat . no . 6 , 185 , 669 ; which is hereby incorporated by reference . the trace 202 will be stored in the trace memory 203 . there is a mapping from the trace starting instruction bundle in the original binary to the trace in the trace memory . when the trace starting bundle is executed , the mapping will automatically lead the execution to the trace stored in the trace memory 203 . typically , an executed branch instruction has its target in the trace memory . this is discussed in the co - pending and commonly assigned u . s . patent application entitled system and method using a hardware embedded run - time optimizer ser . no . 09 / 252 , 170 , filed feb . 18 , 1999 , and issued sep . 17 , 2002 , as u . s . pat . no . 6 , 453 , 411 , which is hereby incorporated by reference . note that the trace may require more than one cache line . as stated previously , long cache lines are inefficient for original binary . this is because the original binary is loaded sequentially , i . e . bundle 1 , 2 , 3 , 4 , 5 , 6 , etc ., and branches taken within the bundles may result in many of the loaded bundles not being used . for example , suppose bundle 6 has a branch taken to bundle 50 , then the loading of bundles 7 - 49 represent wasted time and cache space as they are not going to be used . however , when the trace is loaded into the cache , the entire trace is almost certain to be used . thus , the long cache lines are much more efficient , because of the sequential locality , as the bundles of the trace will ( almost always ) fall through to the next bundle of the trace . note that a trace usually spans several cache lines . it may not end at the end of a cache line . in this case , the remaining part of the cache line can be the start of another trace . note that since traces are also brought into the icache , the profiling and trace selection may end up generating a trace on top of an existing trace . traces can be identified since their addresses are preserved addresses in physical memory . if their participation in subsequent trace selection is not desired , then when the trace is moved into the icache , the counters associated with the trace will not be initialized to the threshold value , and instead are set to a null value . thus , the trace will not participate in profiling . however , subsequent profiling and trace selection could be used to determine whether the trace is considered \u201c good .\u201d for example , if a trace has frequent early exits , then the trace may need to be regenerated . note that more bits of branch history will allow for more accurate predictions to be made by the trace selector . however , this will require more cache space . alternatively , a multi - tiered system may be used such that the trace selector would not to select a trace when a bundle traps for the first time . instead , the trace selector may record the branch history information of the bundle in another location of memory , and then set the threshold back to a second value , which could be smaller , larger or the same as the original threshold value , and return to execution . when this bundle traps again , the trace selector can accumulate the current branch history with the branch history from the first trap to make more accurate branch predictions . although the present invention and its advantages have been described in detail , it should be understood that various changes , substitutions and alterations can be made herein without departing from the spirit and scope of the invention as defined by the appended claims ."}
|
Is the patent correctly categorized?
| 0.25 |
fdcc34483564ad9019b21f93490188a72a8ae33eb03472de41376137951476c1
| 0.078125 | 0.014526 | 0.094238 | 0.002975 | 0.279297 | 0.010986 |
null |
{"patent": "fig1 a depicts the inventive instruction cache ( icache ) 100 of a processor , and includes long cache lines 101 , 102 , 103 , and 104 . only four lines are depicted for simplicity , icache 100 size is implementation dependent . each cache line includes a tag 105 , which is used to tell a cache hit or miss , a plurality of instruction bundles 106 , and counter / branch information 107 . fig1 b depicts the contents of instruction bundles 106 . each bundle is comprised of a group of instructions 108 that can be issued in the same cycle , for example , bundle 0 includes a load instruction , an add instruction , and a compare - branch instruction . note that each bundle has a fixed number of instructions , however , some of the instructions may be nops . fig1 c depicts the counter / branch information associated with each instruction bundle . each instruction bundle has counter information 109 , which is used to determine whether the code within the bundle is hot code . when the bundle is brought into the icache , the counter is initialized to a threshold value . depending on the threshold value desired , the counter can be as small as 8 to 10 bits . the counter is updated when the instruction bundle is retired from the execution pipeline . each update decrements the counter by 1 . note that the counter could initially be set to zero and increment with each retirement . however , this would require a comparison with a non - zero threshold number , e . g . 100 , which requires more work than comparing with a zero threshold number . each instruction bundle 106 in the icache 100 also maintains a branch history 110 , 111 for each instruction within the bundle . this history describes whether the comparisons in the branch instructions have resulted in a fall through to the next instruction or a branch taken to another instruction . branch history 110 is associated with bundle 0 , including slots a , b , c , which correspond to the instructions within the bundle 0 . thus , it appears one slot in the history is allocated for each instruction in the bundle , whether the instruction is a branch instruction or not . when the instructions from the original binary are brought into the icache , the branch history is cleared . the branch history information is updated when the instruction bundle is retired from fie pipeline . note that the number of instructions ( and thus the number of slots ) is by way of example only , as each bundle could have more or fewer instructions . since the third instruction in bundle 0 is a branch instruction , then slot 110 c has branch information . binary zeros indicate a fall through , and binary ones indicate a branch taken . thus , the information in 110 c , i . e . 00100 , indicates that of the last five times that this instruction has been executed , that the instruction br 1 has fallen through , fallen through , been taken , fallen through , and fallen through . note that the number of bits in the history is by way of example only , and more bits could be used to provide a more detailed history ( while requiring more space ), while fewer bits could be used to save space ( while providing less history ). note that either the most significant bit or the last significant bit may represent the most urgent execution instruction . similarly , the information in 111 b and 111 c describe the histories of instruction br 2 and br 3 respectively . note that br 2 has not recently branched , whereas the previous four executions of br 3 have resulted in the branch taken . in operation , once the counter of a bundle reaches zero , a software component known as the trace selector 201 is invoked , via a special trap , to select a trace . diagnose instructions ( special instructions to diagnose hardware ) are used by the trace selector to examine the icache and the branch history information to form a trace . regular instructions cannot read i - cache contents since i - cache is not part of the architecture states . each processor has a set of diagnose instructions defined ( not visible to application programmer ) which can be used to examine i - cache contents . fig2 a and 2b depict trace formation . fig2 a depicts instruction bundles 106 and their associated branch information 11110 . assume that the counter of bundle 1 ( not shown ) has reached zero , and that bundles 5 - 9 and 14 - 99 are not shown for reason of simplicity . note that bundles 1 - 101 may be in one or more cache lines of icache 100 . the trace selector 201 begins building the trace 202 from the hot code , in this case bundle 1 . the trace selector 201 examines the branch information ( if any ) in bundle 1 to predict whether the branch will be taken or fall through . if there are no branch instructions in the bundle , then bundle will fall through to the next sequential bundle . if the trace selector determines that the branch is most likely to fall through , then the next sequential bundle is added to the trace 202 , in this case it would be bundle 2 . note that if a branch instruction that is in the middle of bundle is assumed to be taken , the remaining instructions of the bundle are not included in the trace . the trace 202 is stored in the trace memory 203 . if the trace selector determines that the branch is most likely to be taken , then the target bundle of the branch is added to the trace 202 , in this case it would be bundle 30 . after examining the branch history 112 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 30 will not be taken , and will add the next sequential bundle , bundle 2 , to the trace 202 , and then will examine bundle 2 . after examining the branch history 113 of bundle 2 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken four times and fallen through once . therefore , the trace selector will predict that the branch to bundle 10 will be taken , and will add the target bundle , bundle 10 , to the trace 202 , and then will examine bundle 10 . after examining the branch history 114 of bundle 10 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 20 will not be taken , and will add the next sequential bundle , bundle 11 , to the trace 202 , and then will examine bundle 11 . bundle 11 does not contain any branch instructions , and therefore will not have a branch history , thus the trace selector 201 will add the next sequential bundle , bundle 12 , to the trace 202 , and then will examine bundle 12 . after examining the branch history 115 of bundle 12 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 24 will not be taken , and will add the next sequential bundle , bundle 13 , to the trace 202 , and then will examine bundle 13 . after examining the branch history 116 of bundle 13 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken four times and fallen through once . therefore , the trace selector will predict that the branch to bundle 101 will be taken , and will add the target bundle , bundle 101 , to the trace 202 , and then will examine bundle 101 . after examining the branch history 117 of bundle 101 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken five times . therefore , the trace selector will predict that the branch to bundle 1 will be taken . the trace selector notes that bundle 1 is already part of the trace 202 in trace memory 203 , via the trace of a sequence of bundles , by examining the address of a backward branch , it can be detected whether the target bundle is already part of a trace . the trace selector then ends the trace or passes the formed trace to the optimizer . the branch to bundle 1 from bundle 101 is known as a backward branch , which forms a loop . at this point , the trace may be stopped , as the trace would merely repeat bundles that are already present in the trace . the trace selector may also end the trace based on other criteria from a set of heuristics including the length of the trace , the number of conditional branches encountered , the probability of accumulated branch predictions and other considerations . thus , a trace may end when its length is a multiple of a cache line size . this would make cache operations easier , as the entire line could be loaded or overwritten without having to be concerned about starting and stopping points in the middle of a cache line . the trace could also end after a certain , predetermined number of conditional branches has been encountered . note that branch histories 113 and 116 do indicate that branch falls through occasionally , and thus the trace would be inaccurate as the trace predicts that the branch will be taken . the predetermined number could be based on the probability of error of the trace . for example , the predetermined number would be low if many of the branches have histories of 00011 or 00111 . on the other hand , the predetermined number would be high if many of the branches have histories of 00000 or 11111 . note that a trace may terminate at an indirect branch since the target address is not known . an indirect branch is different from an ip - relative ( or pc - relative ) branch in that the branch target address cannot be computed directly from the branch instruction . its target is stored either in a register or in a memory location . so the target address is unknown unless the instruction is actually executed . for example , however , the trace selector may decide to grow the trace by predicting its most recent target from the target address cache ( tac ), which is a structure commonly used to predict branch target address . for a return branch which is an indirect branch , with its target being dependent on the call site , the trace selector would know the return address if the call instruction is in the trace , if the call instruction is not in the trace , the trace selector can predict the call site using the top address of the return stack buffer ( rsb ), which is a commonly used data structure to predict return branches . the tac and the rsb are discussed in the co - pending and commonly assigned u . s . patent application entitled efficient mapping to optimized code for processor embedded run - time optimizer ser . no . 09 / 252 , 367 , filed feb . 18 , 1999 , and issued feb . 6 , 2001 , as u . s . pat . no . 6 , 185 , 669 ; which is hereby incorporated by reference . the trace 202 will be stored in the trace memory 203 . there is a mapping from the trace starting instruction bundle in the original binary to the trace in the trace memory . when the trace starting bundle is executed , the mapping will automatically lead the execution to the trace stored in the trace memory 203 . typically , an executed branch instruction has its target in the trace memory . this is discussed in the co - pending and commonly assigned u . s . patent application entitled system and method using a hardware embedded run - time optimizer ser . no . 09 / 252 , 170 , filed feb . 18 , 1999 , and issued sep . 17 , 2002 , as u . s . pat . no . 6 , 453 , 411 , which is hereby incorporated by reference . note that the trace may require more than one cache line . as stated previously , long cache lines are inefficient for original binary . this is because the original binary is loaded sequentially , i . e . bundle 1 , 2 , 3 , 4 , 5 , 6 , etc ., and branches taken within the bundles may result in many of the loaded bundles not being used . for example , suppose bundle 6 has a branch taken to bundle 50 , then the loading of bundles 7 - 49 represent wasted time and cache space as they are not going to be used . however , when the trace is loaded into the cache , the entire trace is almost certain to be used . thus , the long cache lines are much more efficient , because of the sequential locality , as the bundles of the trace will ( almost always ) fall through to the next bundle of the trace . note that a trace usually spans several cache lines . it may not end at the end of a cache line . in this case , the remaining part of the cache line can be the start of another trace . note that since traces are also brought into the icache , the profiling and trace selection may end up generating a trace on top of an existing trace . traces can be identified since their addresses are preserved addresses in physical memory . if their participation in subsequent trace selection is not desired , then when the trace is moved into the icache , the counters associated with the trace will not be initialized to the threshold value , and instead are set to a null value . thus , the trace will not participate in profiling . however , subsequent profiling and trace selection could be used to determine whether the trace is considered \u201c good .\u201d for example , if a trace has frequent early exits , then the trace may need to be regenerated . note that more bits of branch history will allow for more accurate predictions to be made by the trace selector . however , this will require more cache space . alternatively , a multi - tiered system may be used such that the trace selector would not to select a trace when a bundle traps for the first time . instead , the trace selector may record the branch history information of the bundle in another location of memory , and then set the threshold back to a second value , which could be smaller , larger or the same as the original threshold value , and return to execution . when this bundle traps again , the trace selector can accumulate the current branch history with the branch history from the first trap to make more accurate branch predictions . although the present invention and its advantages have been described in detail , it should be understood that various changes , substitutions and alterations can be made herein without departing from the spirit and scope of the invention as defined by the appended claims .", "category": "Physics"}
|
{"patent": "fig1 a depicts the inventive instruction cache ( icache ) 100 of a processor , and includes long cache lines 101 , 102 , 103 , and 104 . only four lines are depicted for simplicity , icache 100 size is implementation dependent . each cache line includes a tag 105 , which is used to tell a cache hit or miss , a plurality of instruction bundles 106 , and counter / branch information 107 . fig1 b depicts the contents of instruction bundles 106 . each bundle is comprised of a group of instructions 108 that can be issued in the same cycle , for example , bundle 0 includes a load instruction , an add instruction , and a compare - branch instruction . note that each bundle has a fixed number of instructions , however , some of the instructions may be nops . fig1 c depicts the counter / branch information associated with each instruction bundle . each instruction bundle has counter information 109 , which is used to determine whether the code within the bundle is hot code . when the bundle is brought into the icache , the counter is initialized to a threshold value . depending on the threshold value desired , the counter can be as small as 8 to 10 bits . the counter is updated when the instruction bundle is retired from the execution pipeline . each update decrements the counter by 1 . note that the counter could initially be set to zero and increment with each retirement . however , this would require a comparison with a non - zero threshold number , e . g . 100 , which requires more work than comparing with a zero threshold number . each instruction bundle 106 in the icache 100 also maintains a branch history 110 , 111 for each instruction within the bundle . this history describes whether the comparisons in the branch instructions have resulted in a fall through to the next instruction or a branch taken to another instruction . branch history 110 is associated with bundle 0 , including slots a , b , c , which correspond to the instructions within the bundle 0 . thus , it appears one slot in the history is allocated for each instruction in the bundle , whether the instruction is a branch instruction or not . when the instructions from the original binary are brought into the icache , the branch history is cleared . the branch history information is updated when the instruction bundle is retired from fie pipeline . note that the number of instructions ( and thus the number of slots ) is by way of example only , as each bundle could have more or fewer instructions . since the third instruction in bundle 0 is a branch instruction , then slot 110 c has branch information . binary zeros indicate a fall through , and binary ones indicate a branch taken . thus , the information in 110 c , i . e . 00100 , indicates that of the last five times that this instruction has been executed , that the instruction br 1 has fallen through , fallen through , been taken , fallen through , and fallen through . note that the number of bits in the history is by way of example only , and more bits could be used to provide a more detailed history ( while requiring more space ), while fewer bits could be used to save space ( while providing less history ). note that either the most significant bit or the last significant bit may represent the most urgent execution instruction . similarly , the information in 111 b and 111 c describe the histories of instruction br 2 and br 3 respectively . note that br 2 has not recently branched , whereas the previous four executions of br 3 have resulted in the branch taken . in operation , once the counter of a bundle reaches zero , a software component known as the trace selector 201 is invoked , via a special trap , to select a trace . diagnose instructions ( special instructions to diagnose hardware ) are used by the trace selector to examine the icache and the branch history information to form a trace . regular instructions cannot read i - cache contents since i - cache is not part of the architecture states . each processor has a set of diagnose instructions defined ( not visible to application programmer ) which can be used to examine i - cache contents . fig2 a and 2b depict trace formation . fig2 a depicts instruction bundles 106 and their associated branch information 11110 . assume that the counter of bundle 1 ( not shown ) has reached zero , and that bundles 5 - 9 and 14 - 99 are not shown for reason of simplicity . note that bundles 1 - 101 may be in one or more cache lines of icache 100 . the trace selector 201 begins building the trace 202 from the hot code , in this case bundle 1 . the trace selector 201 examines the branch information ( if any ) in bundle 1 to predict whether the branch will be taken or fall through . if there are no branch instructions in the bundle , then bundle will fall through to the next sequential bundle . if the trace selector determines that the branch is most likely to fall through , then the next sequential bundle is added to the trace 202 , in this case it would be bundle 2 . note that if a branch instruction that is in the middle of bundle is assumed to be taken , the remaining instructions of the bundle are not included in the trace . the trace 202 is stored in the trace memory 203 . if the trace selector determines that the branch is most likely to be taken , then the target bundle of the branch is added to the trace 202 , in this case it would be bundle 30 . after examining the branch history 112 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 30 will not be taken , and will add the next sequential bundle , bundle 2 , to the trace 202 , and then will examine bundle 2 . after examining the branch history 113 of bundle 2 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken four times and fallen through once . therefore , the trace selector will predict that the branch to bundle 10 will be taken , and will add the target bundle , bundle 10 , to the trace 202 , and then will examine bundle 10 . after examining the branch history 114 of bundle 10 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 20 will not be taken , and will add the next sequential bundle , bundle 11 , to the trace 202 , and then will examine bundle 11 . bundle 11 does not contain any branch instructions , and therefore will not have a branch history , thus the trace selector 201 will add the next sequential bundle , bundle 12 , to the trace 202 , and then will examine bundle 12 . after examining the branch history 115 of bundle 12 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 24 will not be taken , and will add the next sequential bundle , bundle 13 , to the trace 202 , and then will examine bundle 13 . after examining the branch history 116 of bundle 13 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken four times and fallen through once . therefore , the trace selector will predict that the branch to bundle 101 will be taken , and will add the target bundle , bundle 101 , to the trace 202 , and then will examine bundle 101 . after examining the branch history 117 of bundle 101 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken five times . therefore , the trace selector will predict that the branch to bundle 1 will be taken . the trace selector notes that bundle 1 is already part of the trace 202 in trace memory 203 , via the trace of a sequence of bundles , by examining the address of a backward branch , it can be detected whether the target bundle is already part of a trace . the trace selector then ends the trace or passes the formed trace to the optimizer . the branch to bundle 1 from bundle 101 is known as a backward branch , which forms a loop . at this point , the trace may be stopped , as the trace would merely repeat bundles that are already present in the trace . the trace selector may also end the trace based on other criteria from a set of heuristics including the length of the trace , the number of conditional branches encountered , the probability of accumulated branch predictions and other considerations . thus , a trace may end when its length is a multiple of a cache line size . this would make cache operations easier , as the entire line could be loaded or overwritten without having to be concerned about starting and stopping points in the middle of a cache line . the trace could also end after a certain , predetermined number of conditional branches has been encountered . note that branch histories 113 and 116 do indicate that branch falls through occasionally , and thus the trace would be inaccurate as the trace predicts that the branch will be taken . the predetermined number could be based on the probability of error of the trace . for example , the predetermined number would be low if many of the branches have histories of 00011 or 00111 . on the other hand , the predetermined number would be high if many of the branches have histories of 00000 or 11111 . note that a trace may terminate at an indirect branch since the target address is not known . an indirect branch is different from an ip - relative ( or pc - relative ) branch in that the branch target address cannot be computed directly from the branch instruction . its target is stored either in a register or in a memory location . so the target address is unknown unless the instruction is actually executed . for example , however , the trace selector may decide to grow the trace by predicting its most recent target from the target address cache ( tac ), which is a structure commonly used to predict branch target address . for a return branch which is an indirect branch , with its target being dependent on the call site , the trace selector would know the return address if the call instruction is in the trace , if the call instruction is not in the trace , the trace selector can predict the call site using the top address of the return stack buffer ( rsb ), which is a commonly used data structure to predict return branches . the tac and the rsb are discussed in the co - pending and commonly assigned u . s . patent application entitled efficient mapping to optimized code for processor embedded run - time optimizer ser . no . 09 / 252 , 367 , filed feb . 18 , 1999 , and issued feb . 6 , 2001 , as u . s . pat . no . 6 , 185 , 669 ; which is hereby incorporated by reference . the trace 202 will be stored in the trace memory 203 . there is a mapping from the trace starting instruction bundle in the original binary to the trace in the trace memory . when the trace starting bundle is executed , the mapping will automatically lead the execution to the trace stored in the trace memory 203 . typically , an executed branch instruction has its target in the trace memory . this is discussed in the co - pending and commonly assigned u . s . patent application entitled system and method using a hardware embedded run - time optimizer ser . no . 09 / 252 , 170 , filed feb . 18 , 1999 , and issued sep . 17 , 2002 , as u . s . pat . no . 6 , 453 , 411 , which is hereby incorporated by reference . note that the trace may require more than one cache line . as stated previously , long cache lines are inefficient for original binary . this is because the original binary is loaded sequentially , i . e . bundle 1 , 2 , 3 , 4 , 5 , 6 , etc ., and branches taken within the bundles may result in many of the loaded bundles not being used . for example , suppose bundle 6 has a branch taken to bundle 50 , then the loading of bundles 7 - 49 represent wasted time and cache space as they are not going to be used . however , when the trace is loaded into the cache , the entire trace is almost certain to be used . thus , the long cache lines are much more efficient , because of the sequential locality , as the bundles of the trace will ( almost always ) fall through to the next bundle of the trace . note that a trace usually spans several cache lines . it may not end at the end of a cache line . in this case , the remaining part of the cache line can be the start of another trace . note that since traces are also brought into the icache , the profiling and trace selection may end up generating a trace on top of an existing trace . traces can be identified since their addresses are preserved addresses in physical memory . if their participation in subsequent trace selection is not desired , then when the trace is moved into the icache , the counters associated with the trace will not be initialized to the threshold value , and instead are set to a null value . thus , the trace will not participate in profiling . however , subsequent profiling and trace selection could be used to determine whether the trace is considered \u201c good .\u201d for example , if a trace has frequent early exits , then the trace may need to be regenerated . note that more bits of branch history will allow for more accurate predictions to be made by the trace selector . however , this will require more cache space . alternatively , a multi - tiered system may be used such that the trace selector would not to select a trace when a bundle traps for the first time . instead , the trace selector may record the branch history information of the bundle in another location of memory , and then set the threshold back to a second value , which could be smaller , larger or the same as the original threshold value , and return to execution . when this bundle traps again , the trace selector can accumulate the current branch history with the branch history from the first trap to make more accurate branch predictions . although the present invention and its advantages have been described in detail , it should be understood that various changes , substitutions and alterations can be made herein without departing from the spirit and scope of the invention as defined by the appended claims .", "category": "Performing Operations; Transporting"}
|
Is the patent correctly categorized?
| 0.25 |
fdcc34483564ad9019b21f93490188a72a8ae33eb03472de41376137951476c1
| 0.003479 | 0.004608 | 0.027588 | 0.038574 | 0.083984 | 0.177734 |
null |
{"category": "Physics", "patent": "fig1 a depicts the inventive instruction cache ( icache ) 100 of a processor , and includes long cache lines 101 , 102 , 103 , and 104 . only four lines are depicted for simplicity , icache 100 size is implementation dependent . each cache line includes a tag 105 , which is used to tell a cache hit or miss , a plurality of instruction bundles 106 , and counter / branch information 107 . fig1 b depicts the contents of instruction bundles 106 . each bundle is comprised of a group of instructions 108 that can be issued in the same cycle , for example , bundle 0 includes a load instruction , an add instruction , and a compare - branch instruction . note that each bundle has a fixed number of instructions , however , some of the instructions may be nops . fig1 c depicts the counter / branch information associated with each instruction bundle . each instruction bundle has counter information 109 , which is used to determine whether the code within the bundle is hot code . when the bundle is brought into the icache , the counter is initialized to a threshold value . depending on the threshold value desired , the counter can be as small as 8 to 10 bits . the counter is updated when the instruction bundle is retired from the execution pipeline . each update decrements the counter by 1 . note that the counter could initially be set to zero and increment with each retirement . however , this would require a comparison with a non - zero threshold number , e . g . 100 , which requires more work than comparing with a zero threshold number . each instruction bundle 106 in the icache 100 also maintains a branch history 110 , 111 for each instruction within the bundle . this history describes whether the comparisons in the branch instructions have resulted in a fall through to the next instruction or a branch taken to another instruction . branch history 110 is associated with bundle 0 , including slots a , b , c , which correspond to the instructions within the bundle 0 . thus , it appears one slot in the history is allocated for each instruction in the bundle , whether the instruction is a branch instruction or not . when the instructions from the original binary are brought into the icache , the branch history is cleared . the branch history information is updated when the instruction bundle is retired from fie pipeline . note that the number of instructions ( and thus the number of slots ) is by way of example only , as each bundle could have more or fewer instructions . since the third instruction in bundle 0 is a branch instruction , then slot 110 c has branch information . binary zeros indicate a fall through , and binary ones indicate a branch taken . thus , the information in 110 c , i . e . 00100 , indicates that of the last five times that this instruction has been executed , that the instruction br 1 has fallen through , fallen through , been taken , fallen through , and fallen through . note that the number of bits in the history is by way of example only , and more bits could be used to provide a more detailed history ( while requiring more space ), while fewer bits could be used to save space ( while providing less history ). note that either the most significant bit or the last significant bit may represent the most urgent execution instruction . similarly , the information in 111 b and 111 c describe the histories of instruction br 2 and br 3 respectively . note that br 2 has not recently branched , whereas the previous four executions of br 3 have resulted in the branch taken . in operation , once the counter of a bundle reaches zero , a software component known as the trace selector 201 is invoked , via a special trap , to select a trace . diagnose instructions ( special instructions to diagnose hardware ) are used by the trace selector to examine the icache and the branch history information to form a trace . regular instructions cannot read i - cache contents since i - cache is not part of the architecture states . each processor has a set of diagnose instructions defined ( not visible to application programmer ) which can be used to examine i - cache contents . fig2 a and 2b depict trace formation . fig2 a depicts instruction bundles 106 and their associated branch information 11110 . assume that the counter of bundle 1 ( not shown ) has reached zero , and that bundles 5 - 9 and 14 - 99 are not shown for reason of simplicity . note that bundles 1 - 101 may be in one or more cache lines of icache 100 . the trace selector 201 begins building the trace 202 from the hot code , in this case bundle 1 . the trace selector 201 examines the branch information ( if any ) in bundle 1 to predict whether the branch will be taken or fall through . if there are no branch instructions in the bundle , then bundle will fall through to the next sequential bundle . if the trace selector determines that the branch is most likely to fall through , then the next sequential bundle is added to the trace 202 , in this case it would be bundle 2 . note that if a branch instruction that is in the middle of bundle is assumed to be taken , the remaining instructions of the bundle are not included in the trace . the trace 202 is stored in the trace memory 203 . if the trace selector determines that the branch is most likely to be taken , then the target bundle of the branch is added to the trace 202 , in this case it would be bundle 30 . after examining the branch history 112 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 30 will not be taken , and will add the next sequential bundle , bundle 2 , to the trace 202 , and then will examine bundle 2 . after examining the branch history 113 of bundle 2 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken four times and fallen through once . therefore , the trace selector will predict that the branch to bundle 10 will be taken , and will add the target bundle , bundle 10 , to the trace 202 , and then will examine bundle 10 . after examining the branch history 114 of bundle 10 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 20 will not be taken , and will add the next sequential bundle , bundle 11 , to the trace 202 , and then will examine bundle 11 . bundle 11 does not contain any branch instructions , and therefore will not have a branch history , thus the trace selector 201 will add the next sequential bundle , bundle 12 , to the trace 202 , and then will examine bundle 12 . after examining the branch history 115 of bundle 12 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 24 will not be taken , and will add the next sequential bundle , bundle 13 , to the trace 202 , and then will examine bundle 13 . after examining the branch history 116 of bundle 13 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken four times and fallen through once . therefore , the trace selector will predict that the branch to bundle 101 will be taken , and will add the target bundle , bundle 101 , to the trace 202 , and then will examine bundle 101 . after examining the branch history 117 of bundle 101 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken five times . therefore , the trace selector will predict that the branch to bundle 1 will be taken . the trace selector notes that bundle 1 is already part of the trace 202 in trace memory 203 , via the trace of a sequence of bundles , by examining the address of a backward branch , it can be detected whether the target bundle is already part of a trace . the trace selector then ends the trace or passes the formed trace to the optimizer . the branch to bundle 1 from bundle 101 is known as a backward branch , which forms a loop . at this point , the trace may be stopped , as the trace would merely repeat bundles that are already present in the trace . the trace selector may also end the trace based on other criteria from a set of heuristics including the length of the trace , the number of conditional branches encountered , the probability of accumulated branch predictions and other considerations . thus , a trace may end when its length is a multiple of a cache line size . this would make cache operations easier , as the entire line could be loaded or overwritten without having to be concerned about starting and stopping points in the middle of a cache line . the trace could also end after a certain , predetermined number of conditional branches has been encountered . note that branch histories 113 and 116 do indicate that branch falls through occasionally , and thus the trace would be inaccurate as the trace predicts that the branch will be taken . the predetermined number could be based on the probability of error of the trace . for example , the predetermined number would be low if many of the branches have histories of 00011 or 00111 . on the other hand , the predetermined number would be high if many of the branches have histories of 00000 or 11111 . note that a trace may terminate at an indirect branch since the target address is not known . an indirect branch is different from an ip - relative ( or pc - relative ) branch in that the branch target address cannot be computed directly from the branch instruction . its target is stored either in a register or in a memory location . so the target address is unknown unless the instruction is actually executed . for example , however , the trace selector may decide to grow the trace by predicting its most recent target from the target address cache ( tac ), which is a structure commonly used to predict branch target address . for a return branch which is an indirect branch , with its target being dependent on the call site , the trace selector would know the return address if the call instruction is in the trace , if the call instruction is not in the trace , the trace selector can predict the call site using the top address of the return stack buffer ( rsb ), which is a commonly used data structure to predict return branches . the tac and the rsb are discussed in the co - pending and commonly assigned u . s . patent application entitled efficient mapping to optimized code for processor embedded run - time optimizer ser . no . 09 / 252 , 367 , filed feb . 18 , 1999 , and issued feb . 6 , 2001 , as u . s . pat . no . 6 , 185 , 669 ; which is hereby incorporated by reference . the trace 202 will be stored in the trace memory 203 . there is a mapping from the trace starting instruction bundle in the original binary to the trace in the trace memory . when the trace starting bundle is executed , the mapping will automatically lead the execution to the trace stored in the trace memory 203 . typically , an executed branch instruction has its target in the trace memory . this is discussed in the co - pending and commonly assigned u . s . patent application entitled system and method using a hardware embedded run - time optimizer ser . no . 09 / 252 , 170 , filed feb . 18 , 1999 , and issued sep . 17 , 2002 , as u . s . pat . no . 6 , 453 , 411 , which is hereby incorporated by reference . note that the trace may require more than one cache line . as stated previously , long cache lines are inefficient for original binary . this is because the original binary is loaded sequentially , i . e . bundle 1 , 2 , 3 , 4 , 5 , 6 , etc ., and branches taken within the bundles may result in many of the loaded bundles not being used . for example , suppose bundle 6 has a branch taken to bundle 50 , then the loading of bundles 7 - 49 represent wasted time and cache space as they are not going to be used . however , when the trace is loaded into the cache , the entire trace is almost certain to be used . thus , the long cache lines are much more efficient , because of the sequential locality , as the bundles of the trace will ( almost always ) fall through to the next bundle of the trace . note that a trace usually spans several cache lines . it may not end at the end of a cache line . in this case , the remaining part of the cache line can be the start of another trace . note that since traces are also brought into the icache , the profiling and trace selection may end up generating a trace on top of an existing trace . traces can be identified since their addresses are preserved addresses in physical memory . if their participation in subsequent trace selection is not desired , then when the trace is moved into the icache , the counters associated with the trace will not be initialized to the threshold value , and instead are set to a null value . thus , the trace will not participate in profiling . however , subsequent profiling and trace selection could be used to determine whether the trace is considered \u201c good .\u201d for example , if a trace has frequent early exits , then the trace may need to be regenerated . note that more bits of branch history will allow for more accurate predictions to be made by the trace selector . however , this will require more cache space . alternatively , a multi - tiered system may be used such that the trace selector would not to select a trace when a bundle traps for the first time . instead , the trace selector may record the branch history information of the bundle in another location of memory , and then set the threshold back to a second value , which could be smaller , larger or the same as the original threshold value , and return to execution . when this bundle traps again , the trace selector can accumulate the current branch history with the branch history from the first trap to make more accurate branch predictions . although the present invention and its advantages have been described in detail , it should be understood that various changes , substitutions and alterations can be made herein without departing from the spirit and scope of the invention as defined by the appended claims ."}
|
{"category": "Chemistry; Metallurgy", "patent": "fig1 a depicts the inventive instruction cache ( icache ) 100 of a processor , and includes long cache lines 101 , 102 , 103 , and 104 . only four lines are depicted for simplicity , icache 100 size is implementation dependent . each cache line includes a tag 105 , which is used to tell a cache hit or miss , a plurality of instruction bundles 106 , and counter / branch information 107 . fig1 b depicts the contents of instruction bundles 106 . each bundle is comprised of a group of instructions 108 that can be issued in the same cycle , for example , bundle 0 includes a load instruction , an add instruction , and a compare - branch instruction . note that each bundle has a fixed number of instructions , however , some of the instructions may be nops . fig1 c depicts the counter / branch information associated with each instruction bundle . each instruction bundle has counter information 109 , which is used to determine whether the code within the bundle is hot code . when the bundle is brought into the icache , the counter is initialized to a threshold value . depending on the threshold value desired , the counter can be as small as 8 to 10 bits . the counter is updated when the instruction bundle is retired from the execution pipeline . each update decrements the counter by 1 . note that the counter could initially be set to zero and increment with each retirement . however , this would require a comparison with a non - zero threshold number , e . g . 100 , which requires more work than comparing with a zero threshold number . each instruction bundle 106 in the icache 100 also maintains a branch history 110 , 111 for each instruction within the bundle . this history describes whether the comparisons in the branch instructions have resulted in a fall through to the next instruction or a branch taken to another instruction . branch history 110 is associated with bundle 0 , including slots a , b , c , which correspond to the instructions within the bundle 0 . thus , it appears one slot in the history is allocated for each instruction in the bundle , whether the instruction is a branch instruction or not . when the instructions from the original binary are brought into the icache , the branch history is cleared . the branch history information is updated when the instruction bundle is retired from fie pipeline . note that the number of instructions ( and thus the number of slots ) is by way of example only , as each bundle could have more or fewer instructions . since the third instruction in bundle 0 is a branch instruction , then slot 110 c has branch information . binary zeros indicate a fall through , and binary ones indicate a branch taken . thus , the information in 110 c , i . e . 00100 , indicates that of the last five times that this instruction has been executed , that the instruction br 1 has fallen through , fallen through , been taken , fallen through , and fallen through . note that the number of bits in the history is by way of example only , and more bits could be used to provide a more detailed history ( while requiring more space ), while fewer bits could be used to save space ( while providing less history ). note that either the most significant bit or the last significant bit may represent the most urgent execution instruction . similarly , the information in 111 b and 111 c describe the histories of instruction br 2 and br 3 respectively . note that br 2 has not recently branched , whereas the previous four executions of br 3 have resulted in the branch taken . in operation , once the counter of a bundle reaches zero , a software component known as the trace selector 201 is invoked , via a special trap , to select a trace . diagnose instructions ( special instructions to diagnose hardware ) are used by the trace selector to examine the icache and the branch history information to form a trace . regular instructions cannot read i - cache contents since i - cache is not part of the architecture states . each processor has a set of diagnose instructions defined ( not visible to application programmer ) which can be used to examine i - cache contents . fig2 a and 2b depict trace formation . fig2 a depicts instruction bundles 106 and their associated branch information 11110 . assume that the counter of bundle 1 ( not shown ) has reached zero , and that bundles 5 - 9 and 14 - 99 are not shown for reason of simplicity . note that bundles 1 - 101 may be in one or more cache lines of icache 100 . the trace selector 201 begins building the trace 202 from the hot code , in this case bundle 1 . the trace selector 201 examines the branch information ( if any ) in bundle 1 to predict whether the branch will be taken or fall through . if there are no branch instructions in the bundle , then bundle will fall through to the next sequential bundle . if the trace selector determines that the branch is most likely to fall through , then the next sequential bundle is added to the trace 202 , in this case it would be bundle 2 . note that if a branch instruction that is in the middle of bundle is assumed to be taken , the remaining instructions of the bundle are not included in the trace . the trace 202 is stored in the trace memory 203 . if the trace selector determines that the branch is most likely to be taken , then the target bundle of the branch is added to the trace 202 , in this case it would be bundle 30 . after examining the branch history 112 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 30 will not be taken , and will add the next sequential bundle , bundle 2 , to the trace 202 , and then will examine bundle 2 . after examining the branch history 113 of bundle 2 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken four times and fallen through once . therefore , the trace selector will predict that the branch to bundle 10 will be taken , and will add the target bundle , bundle 10 , to the trace 202 , and then will examine bundle 10 . after examining the branch history 114 of bundle 10 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 20 will not be taken , and will add the next sequential bundle , bundle 11 , to the trace 202 , and then will examine bundle 11 . bundle 11 does not contain any branch instructions , and therefore will not have a branch history , thus the trace selector 201 will add the next sequential bundle , bundle 12 , to the trace 202 , and then will examine bundle 12 . after examining the branch history 115 of bundle 12 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 24 will not be taken , and will add the next sequential bundle , bundle 13 , to the trace 202 , and then will examine bundle 13 . after examining the branch history 116 of bundle 13 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken four times and fallen through once . therefore , the trace selector will predict that the branch to bundle 101 will be taken , and will add the target bundle , bundle 101 , to the trace 202 , and then will examine bundle 101 . after examining the branch history 117 of bundle 101 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken five times . therefore , the trace selector will predict that the branch to bundle 1 will be taken . the trace selector notes that bundle 1 is already part of the trace 202 in trace memory 203 , via the trace of a sequence of bundles , by examining the address of a backward branch , it can be detected whether the target bundle is already part of a trace . the trace selector then ends the trace or passes the formed trace to the optimizer . the branch to bundle 1 from bundle 101 is known as a backward branch , which forms a loop . at this point , the trace may be stopped , as the trace would merely repeat bundles that are already present in the trace . the trace selector may also end the trace based on other criteria from a set of heuristics including the length of the trace , the number of conditional branches encountered , the probability of accumulated branch predictions and other considerations . thus , a trace may end when its length is a multiple of a cache line size . this would make cache operations easier , as the entire line could be loaded or overwritten without having to be concerned about starting and stopping points in the middle of a cache line . the trace could also end after a certain , predetermined number of conditional branches has been encountered . note that branch histories 113 and 116 do indicate that branch falls through occasionally , and thus the trace would be inaccurate as the trace predicts that the branch will be taken . the predetermined number could be based on the probability of error of the trace . for example , the predetermined number would be low if many of the branches have histories of 00011 or 00111 . on the other hand , the predetermined number would be high if many of the branches have histories of 00000 or 11111 . note that a trace may terminate at an indirect branch since the target address is not known . an indirect branch is different from an ip - relative ( or pc - relative ) branch in that the branch target address cannot be computed directly from the branch instruction . its target is stored either in a register or in a memory location . so the target address is unknown unless the instruction is actually executed . for example , however , the trace selector may decide to grow the trace by predicting its most recent target from the target address cache ( tac ), which is a structure commonly used to predict branch target address . for a return branch which is an indirect branch , with its target being dependent on the call site , the trace selector would know the return address if the call instruction is in the trace , if the call instruction is not in the trace , the trace selector can predict the call site using the top address of the return stack buffer ( rsb ), which is a commonly used data structure to predict return branches . the tac and the rsb are discussed in the co - pending and commonly assigned u . s . patent application entitled efficient mapping to optimized code for processor embedded run - time optimizer ser . no . 09 / 252 , 367 , filed feb . 18 , 1999 , and issued feb . 6 , 2001 , as u . s . pat . no . 6 , 185 , 669 ; which is hereby incorporated by reference . the trace 202 will be stored in the trace memory 203 . there is a mapping from the trace starting instruction bundle in the original binary to the trace in the trace memory . when the trace starting bundle is executed , the mapping will automatically lead the execution to the trace stored in the trace memory 203 . typically , an executed branch instruction has its target in the trace memory . this is discussed in the co - pending and commonly assigned u . s . patent application entitled system and method using a hardware embedded run - time optimizer ser . no . 09 / 252 , 170 , filed feb . 18 , 1999 , and issued sep . 17 , 2002 , as u . s . pat . no . 6 , 453 , 411 , which is hereby incorporated by reference . note that the trace may require more than one cache line . as stated previously , long cache lines are inefficient for original binary . this is because the original binary is loaded sequentially , i . e . bundle 1 , 2 , 3 , 4 , 5 , 6 , etc ., and branches taken within the bundles may result in many of the loaded bundles not being used . for example , suppose bundle 6 has a branch taken to bundle 50 , then the loading of bundles 7 - 49 represent wasted time and cache space as they are not going to be used . however , when the trace is loaded into the cache , the entire trace is almost certain to be used . thus , the long cache lines are much more efficient , because of the sequential locality , as the bundles of the trace will ( almost always ) fall through to the next bundle of the trace . note that a trace usually spans several cache lines . it may not end at the end of a cache line . in this case , the remaining part of the cache line can be the start of another trace . note that since traces are also brought into the icache , the profiling and trace selection may end up generating a trace on top of an existing trace . traces can be identified since their addresses are preserved addresses in physical memory . if their participation in subsequent trace selection is not desired , then when the trace is moved into the icache , the counters associated with the trace will not be initialized to the threshold value , and instead are set to a null value . thus , the trace will not participate in profiling . however , subsequent profiling and trace selection could be used to determine whether the trace is considered \u201c good .\u201d for example , if a trace has frequent early exits , then the trace may need to be regenerated . note that more bits of branch history will allow for more accurate predictions to be made by the trace selector . however , this will require more cache space . alternatively , a multi - tiered system may be used such that the trace selector would not to select a trace when a bundle traps for the first time . instead , the trace selector may record the branch history information of the bundle in another location of memory , and then set the threshold back to a second value , which could be smaller , larger or the same as the original threshold value , and return to execution . when this bundle traps again , the trace selector can accumulate the current branch history with the branch history from the first trap to make more accurate branch predictions . although the present invention and its advantages have been described in detail , it should be understood that various changes , substitutions and alterations can be made herein without departing from the spirit and scope of the invention as defined by the appended claims ."}
|
Does the category match the content of the patent?
| 0.25 |
fdcc34483564ad9019b21f93490188a72a8ae33eb03472de41376137951476c1
| 0.021606 | 0.00008 | 0.05835 | 0.000534 | 0.03064 | 0.001328 |
null |
{"category": "Physics", "patent": "fig1 a depicts the inventive instruction cache ( icache ) 100 of a processor , and includes long cache lines 101 , 102 , 103 , and 104 . only four lines are depicted for simplicity , icache 100 size is implementation dependent . each cache line includes a tag 105 , which is used to tell a cache hit or miss , a plurality of instruction bundles 106 , and counter / branch information 107 . fig1 b depicts the contents of instruction bundles 106 . each bundle is comprised of a group of instructions 108 that can be issued in the same cycle , for example , bundle 0 includes a load instruction , an add instruction , and a compare - branch instruction . note that each bundle has a fixed number of instructions , however , some of the instructions may be nops . fig1 c depicts the counter / branch information associated with each instruction bundle . each instruction bundle has counter information 109 , which is used to determine whether the code within the bundle is hot code . when the bundle is brought into the icache , the counter is initialized to a threshold value . depending on the threshold value desired , the counter can be as small as 8 to 10 bits . the counter is updated when the instruction bundle is retired from the execution pipeline . each update decrements the counter by 1 . note that the counter could initially be set to zero and increment with each retirement . however , this would require a comparison with a non - zero threshold number , e . g . 100 , which requires more work than comparing with a zero threshold number . each instruction bundle 106 in the icache 100 also maintains a branch history 110 , 111 for each instruction within the bundle . this history describes whether the comparisons in the branch instructions have resulted in a fall through to the next instruction or a branch taken to another instruction . branch history 110 is associated with bundle 0 , including slots a , b , c , which correspond to the instructions within the bundle 0 . thus , it appears one slot in the history is allocated for each instruction in the bundle , whether the instruction is a branch instruction or not . when the instructions from the original binary are brought into the icache , the branch history is cleared . the branch history information is updated when the instruction bundle is retired from fie pipeline . note that the number of instructions ( and thus the number of slots ) is by way of example only , as each bundle could have more or fewer instructions . since the third instruction in bundle 0 is a branch instruction , then slot 110 c has branch information . binary zeros indicate a fall through , and binary ones indicate a branch taken . thus , the information in 110 c , i . e . 00100 , indicates that of the last five times that this instruction has been executed , that the instruction br 1 has fallen through , fallen through , been taken , fallen through , and fallen through . note that the number of bits in the history is by way of example only , and more bits could be used to provide a more detailed history ( while requiring more space ), while fewer bits could be used to save space ( while providing less history ). note that either the most significant bit or the last significant bit may represent the most urgent execution instruction . similarly , the information in 111 b and 111 c describe the histories of instruction br 2 and br 3 respectively . note that br 2 has not recently branched , whereas the previous four executions of br 3 have resulted in the branch taken . in operation , once the counter of a bundle reaches zero , a software component known as the trace selector 201 is invoked , via a special trap , to select a trace . diagnose instructions ( special instructions to diagnose hardware ) are used by the trace selector to examine the icache and the branch history information to form a trace . regular instructions cannot read i - cache contents since i - cache is not part of the architecture states . each processor has a set of diagnose instructions defined ( not visible to application programmer ) which can be used to examine i - cache contents . fig2 a and 2b depict trace formation . fig2 a depicts instruction bundles 106 and their associated branch information 11110 . assume that the counter of bundle 1 ( not shown ) has reached zero , and that bundles 5 - 9 and 14 - 99 are not shown for reason of simplicity . note that bundles 1 - 101 may be in one or more cache lines of icache 100 . the trace selector 201 begins building the trace 202 from the hot code , in this case bundle 1 . the trace selector 201 examines the branch information ( if any ) in bundle 1 to predict whether the branch will be taken or fall through . if there are no branch instructions in the bundle , then bundle will fall through to the next sequential bundle . if the trace selector determines that the branch is most likely to fall through , then the next sequential bundle is added to the trace 202 , in this case it would be bundle 2 . note that if a branch instruction that is in the middle of bundle is assumed to be taken , the remaining instructions of the bundle are not included in the trace . the trace 202 is stored in the trace memory 203 . if the trace selector determines that the branch is most likely to be taken , then the target bundle of the branch is added to the trace 202 , in this case it would be bundle 30 . after examining the branch history 112 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 30 will not be taken , and will add the next sequential bundle , bundle 2 , to the trace 202 , and then will examine bundle 2 . after examining the branch history 113 of bundle 2 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken four times and fallen through once . therefore , the trace selector will predict that the branch to bundle 10 will be taken , and will add the target bundle , bundle 10 , to the trace 202 , and then will examine bundle 10 . after examining the branch history 114 of bundle 10 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 20 will not be taken , and will add the next sequential bundle , bundle 11 , to the trace 202 , and then will examine bundle 11 . bundle 11 does not contain any branch instructions , and therefore will not have a branch history , thus the trace selector 201 will add the next sequential bundle , bundle 12 , to the trace 202 , and then will examine bundle 12 . after examining the branch history 115 of bundle 12 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 24 will not be taken , and will add the next sequential bundle , bundle 13 , to the trace 202 , and then will examine bundle 13 . after examining the branch history 116 of bundle 13 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken four times and fallen through once . therefore , the trace selector will predict that the branch to bundle 101 will be taken , and will add the target bundle , bundle 101 , to the trace 202 , and then will examine bundle 101 . after examining the branch history 117 of bundle 101 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken five times . therefore , the trace selector will predict that the branch to bundle 1 will be taken . the trace selector notes that bundle 1 is already part of the trace 202 in trace memory 203 , via the trace of a sequence of bundles , by examining the address of a backward branch , it can be detected whether the target bundle is already part of a trace . the trace selector then ends the trace or passes the formed trace to the optimizer . the branch to bundle 1 from bundle 101 is known as a backward branch , which forms a loop . at this point , the trace may be stopped , as the trace would merely repeat bundles that are already present in the trace . the trace selector may also end the trace based on other criteria from a set of heuristics including the length of the trace , the number of conditional branches encountered , the probability of accumulated branch predictions and other considerations . thus , a trace may end when its length is a multiple of a cache line size . this would make cache operations easier , as the entire line could be loaded or overwritten without having to be concerned about starting and stopping points in the middle of a cache line . the trace could also end after a certain , predetermined number of conditional branches has been encountered . note that branch histories 113 and 116 do indicate that branch falls through occasionally , and thus the trace would be inaccurate as the trace predicts that the branch will be taken . the predetermined number could be based on the probability of error of the trace . for example , the predetermined number would be low if many of the branches have histories of 00011 or 00111 . on the other hand , the predetermined number would be high if many of the branches have histories of 00000 or 11111 . note that a trace may terminate at an indirect branch since the target address is not known . an indirect branch is different from an ip - relative ( or pc - relative ) branch in that the branch target address cannot be computed directly from the branch instruction . its target is stored either in a register or in a memory location . so the target address is unknown unless the instruction is actually executed . for example , however , the trace selector may decide to grow the trace by predicting its most recent target from the target address cache ( tac ), which is a structure commonly used to predict branch target address . for a return branch which is an indirect branch , with its target being dependent on the call site , the trace selector would know the return address if the call instruction is in the trace , if the call instruction is not in the trace , the trace selector can predict the call site using the top address of the return stack buffer ( rsb ), which is a commonly used data structure to predict return branches . the tac and the rsb are discussed in the co - pending and commonly assigned u . s . patent application entitled efficient mapping to optimized code for processor embedded run - time optimizer ser . no . 09 / 252 , 367 , filed feb . 18 , 1999 , and issued feb . 6 , 2001 , as u . s . pat . no . 6 , 185 , 669 ; which is hereby incorporated by reference . the trace 202 will be stored in the trace memory 203 . there is a mapping from the trace starting instruction bundle in the original binary to the trace in the trace memory . when the trace starting bundle is executed , the mapping will automatically lead the execution to the trace stored in the trace memory 203 . typically , an executed branch instruction has its target in the trace memory . this is discussed in the co - pending and commonly assigned u . s . patent application entitled system and method using a hardware embedded run - time optimizer ser . no . 09 / 252 , 170 , filed feb . 18 , 1999 , and issued sep . 17 , 2002 , as u . s . pat . no . 6 , 453 , 411 , which is hereby incorporated by reference . note that the trace may require more than one cache line . as stated previously , long cache lines are inefficient for original binary . this is because the original binary is loaded sequentially , i . e . bundle 1 , 2 , 3 , 4 , 5 , 6 , etc ., and branches taken within the bundles may result in many of the loaded bundles not being used . for example , suppose bundle 6 has a branch taken to bundle 50 , then the loading of bundles 7 - 49 represent wasted time and cache space as they are not going to be used . however , when the trace is loaded into the cache , the entire trace is almost certain to be used . thus , the long cache lines are much more efficient , because of the sequential locality , as the bundles of the trace will ( almost always ) fall through to the next bundle of the trace . note that a trace usually spans several cache lines . it may not end at the end of a cache line . in this case , the remaining part of the cache line can be the start of another trace . note that since traces are also brought into the icache , the profiling and trace selection may end up generating a trace on top of an existing trace . traces can be identified since their addresses are preserved addresses in physical memory . if their participation in subsequent trace selection is not desired , then when the trace is moved into the icache , the counters associated with the trace will not be initialized to the threshold value , and instead are set to a null value . thus , the trace will not participate in profiling . however , subsequent profiling and trace selection could be used to determine whether the trace is considered \u201c good .\u201d for example , if a trace has frequent early exits , then the trace may need to be regenerated . note that more bits of branch history will allow for more accurate predictions to be made by the trace selector . however , this will require more cache space . alternatively , a multi - tiered system may be used such that the trace selector would not to select a trace when a bundle traps for the first time . instead , the trace selector may record the branch history information of the bundle in another location of memory , and then set the threshold back to a second value , which could be smaller , larger or the same as the original threshold value , and return to execution . when this bundle traps again , the trace selector can accumulate the current branch history with the branch history from the first trap to make more accurate branch predictions . although the present invention and its advantages have been described in detail , it should be understood that various changes , substitutions and alterations can be made herein without departing from the spirit and scope of the invention as defined by the appended claims ."}
|
{"patent": "fig1 a depicts the inventive instruction cache ( icache ) 100 of a processor , and includes long cache lines 101 , 102 , 103 , and 104 . only four lines are depicted for simplicity , icache 100 size is implementation dependent . each cache line includes a tag 105 , which is used to tell a cache hit or miss , a plurality of instruction bundles 106 , and counter / branch information 107 . fig1 b depicts the contents of instruction bundles 106 . each bundle is comprised of a group of instructions 108 that can be issued in the same cycle , for example , bundle 0 includes a load instruction , an add instruction , and a compare - branch instruction . note that each bundle has a fixed number of instructions , however , some of the instructions may be nops . fig1 c depicts the counter / branch information associated with each instruction bundle . each instruction bundle has counter information 109 , which is used to determine whether the code within the bundle is hot code . when the bundle is brought into the icache , the counter is initialized to a threshold value . depending on the threshold value desired , the counter can be as small as 8 to 10 bits . the counter is updated when the instruction bundle is retired from the execution pipeline . each update decrements the counter by 1 . note that the counter could initially be set to zero and increment with each retirement . however , this would require a comparison with a non - zero threshold number , e . g . 100 , which requires more work than comparing with a zero threshold number . each instruction bundle 106 in the icache 100 also maintains a branch history 110 , 111 for each instruction within the bundle . this history describes whether the comparisons in the branch instructions have resulted in a fall through to the next instruction or a branch taken to another instruction . branch history 110 is associated with bundle 0 , including slots a , b , c , which correspond to the instructions within the bundle 0 . thus , it appears one slot in the history is allocated for each instruction in the bundle , whether the instruction is a branch instruction or not . when the instructions from the original binary are brought into the icache , the branch history is cleared . the branch history information is updated when the instruction bundle is retired from fie pipeline . note that the number of instructions ( and thus the number of slots ) is by way of example only , as each bundle could have more or fewer instructions . since the third instruction in bundle 0 is a branch instruction , then slot 110 c has branch information . binary zeros indicate a fall through , and binary ones indicate a branch taken . thus , the information in 110 c , i . e . 00100 , indicates that of the last five times that this instruction has been executed , that the instruction br 1 has fallen through , fallen through , been taken , fallen through , and fallen through . note that the number of bits in the history is by way of example only , and more bits could be used to provide a more detailed history ( while requiring more space ), while fewer bits could be used to save space ( while providing less history ). note that either the most significant bit or the last significant bit may represent the most urgent execution instruction . similarly , the information in 111 b and 111 c describe the histories of instruction br 2 and br 3 respectively . note that br 2 has not recently branched , whereas the previous four executions of br 3 have resulted in the branch taken . in operation , once the counter of a bundle reaches zero , a software component known as the trace selector 201 is invoked , via a special trap , to select a trace . diagnose instructions ( special instructions to diagnose hardware ) are used by the trace selector to examine the icache and the branch history information to form a trace . regular instructions cannot read i - cache contents since i - cache is not part of the architecture states . each processor has a set of diagnose instructions defined ( not visible to application programmer ) which can be used to examine i - cache contents . fig2 a and 2b depict trace formation . fig2 a depicts instruction bundles 106 and their associated branch information 11110 . assume that the counter of bundle 1 ( not shown ) has reached zero , and that bundles 5 - 9 and 14 - 99 are not shown for reason of simplicity . note that bundles 1 - 101 may be in one or more cache lines of icache 100 . the trace selector 201 begins building the trace 202 from the hot code , in this case bundle 1 . the trace selector 201 examines the branch information ( if any ) in bundle 1 to predict whether the branch will be taken or fall through . if there are no branch instructions in the bundle , then bundle will fall through to the next sequential bundle . if the trace selector determines that the branch is most likely to fall through , then the next sequential bundle is added to the trace 202 , in this case it would be bundle 2 . note that if a branch instruction that is in the middle of bundle is assumed to be taken , the remaining instructions of the bundle are not included in the trace . the trace 202 is stored in the trace memory 203 . if the trace selector determines that the branch is most likely to be taken , then the target bundle of the branch is added to the trace 202 , in this case it would be bundle 30 . after examining the branch history 112 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 30 will not be taken , and will add the next sequential bundle , bundle 2 , to the trace 202 , and then will examine bundle 2 . after examining the branch history 113 of bundle 2 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken four times and fallen through once . therefore , the trace selector will predict that the branch to bundle 10 will be taken , and will add the target bundle , bundle 10 , to the trace 202 , and then will examine bundle 10 . after examining the branch history 114 of bundle 10 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 20 will not be taken , and will add the next sequential bundle , bundle 11 , to the trace 202 , and then will examine bundle 11 . bundle 11 does not contain any branch instructions , and therefore will not have a branch history , thus the trace selector 201 will add the next sequential bundle , bundle 12 , to the trace 202 , and then will examine bundle 12 . after examining the branch history 115 of bundle 12 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 24 will not be taken , and will add the next sequential bundle , bundle 13 , to the trace 202 , and then will examine bundle 13 . after examining the branch history 116 of bundle 13 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken four times and fallen through once . therefore , the trace selector will predict that the branch to bundle 101 will be taken , and will add the target bundle , bundle 101 , to the trace 202 , and then will examine bundle 101 . after examining the branch history 117 of bundle 101 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken five times . therefore , the trace selector will predict that the branch to bundle 1 will be taken . the trace selector notes that bundle 1 is already part of the trace 202 in trace memory 203 , via the trace of a sequence of bundles , by examining the address of a backward branch , it can be detected whether the target bundle is already part of a trace . the trace selector then ends the trace or passes the formed trace to the optimizer . the branch to bundle 1 from bundle 101 is known as a backward branch , which forms a loop . at this point , the trace may be stopped , as the trace would merely repeat bundles that are already present in the trace . the trace selector may also end the trace based on other criteria from a set of heuristics including the length of the trace , the number of conditional branches encountered , the probability of accumulated branch predictions and other considerations . thus , a trace may end when its length is a multiple of a cache line size . this would make cache operations easier , as the entire line could be loaded or overwritten without having to be concerned about starting and stopping points in the middle of a cache line . the trace could also end after a certain , predetermined number of conditional branches has been encountered . note that branch histories 113 and 116 do indicate that branch falls through occasionally , and thus the trace would be inaccurate as the trace predicts that the branch will be taken . the predetermined number could be based on the probability of error of the trace . for example , the predetermined number would be low if many of the branches have histories of 00011 or 00111 . on the other hand , the predetermined number would be high if many of the branches have histories of 00000 or 11111 . note that a trace may terminate at an indirect branch since the target address is not known . an indirect branch is different from an ip - relative ( or pc - relative ) branch in that the branch target address cannot be computed directly from the branch instruction . its target is stored either in a register or in a memory location . so the target address is unknown unless the instruction is actually executed . for example , however , the trace selector may decide to grow the trace by predicting its most recent target from the target address cache ( tac ), which is a structure commonly used to predict branch target address . for a return branch which is an indirect branch , with its target being dependent on the call site , the trace selector would know the return address if the call instruction is in the trace , if the call instruction is not in the trace , the trace selector can predict the call site using the top address of the return stack buffer ( rsb ), which is a commonly used data structure to predict return branches . the tac and the rsb are discussed in the co - pending and commonly assigned u . s . patent application entitled efficient mapping to optimized code for processor embedded run - time optimizer ser . no . 09 / 252 , 367 , filed feb . 18 , 1999 , and issued feb . 6 , 2001 , as u . s . pat . no . 6 , 185 , 669 ; which is hereby incorporated by reference . the trace 202 will be stored in the trace memory 203 . there is a mapping from the trace starting instruction bundle in the original binary to the trace in the trace memory . when the trace starting bundle is executed , the mapping will automatically lead the execution to the trace stored in the trace memory 203 . typically , an executed branch instruction has its target in the trace memory . this is discussed in the co - pending and commonly assigned u . s . patent application entitled system and method using a hardware embedded run - time optimizer ser . no . 09 / 252 , 170 , filed feb . 18 , 1999 , and issued sep . 17 , 2002 , as u . s . pat . no . 6 , 453 , 411 , which is hereby incorporated by reference . note that the trace may require more than one cache line . as stated previously , long cache lines are inefficient for original binary . this is because the original binary is loaded sequentially , i . e . bundle 1 , 2 , 3 , 4 , 5 , 6 , etc ., and branches taken within the bundles may result in many of the loaded bundles not being used . for example , suppose bundle 6 has a branch taken to bundle 50 , then the loading of bundles 7 - 49 represent wasted time and cache space as they are not going to be used . however , when the trace is loaded into the cache , the entire trace is almost certain to be used . thus , the long cache lines are much more efficient , because of the sequential locality , as the bundles of the trace will ( almost always ) fall through to the next bundle of the trace . note that a trace usually spans several cache lines . it may not end at the end of a cache line . in this case , the remaining part of the cache line can be the start of another trace . note that since traces are also brought into the icache , the profiling and trace selection may end up generating a trace on top of an existing trace . traces can be identified since their addresses are preserved addresses in physical memory . if their participation in subsequent trace selection is not desired , then when the trace is moved into the icache , the counters associated with the trace will not be initialized to the threshold value , and instead are set to a null value . thus , the trace will not participate in profiling . however , subsequent profiling and trace selection could be used to determine whether the trace is considered \u201c good .\u201d for example , if a trace has frequent early exits , then the trace may need to be regenerated . note that more bits of branch history will allow for more accurate predictions to be made by the trace selector . however , this will require more cache space . alternatively , a multi - tiered system may be used such that the trace selector would not to select a trace when a bundle traps for the first time . instead , the trace selector may record the branch history information of the bundle in another location of memory , and then set the threshold back to a second value , which could be smaller , larger or the same as the original threshold value , and return to execution . when this bundle traps again , the trace selector can accumulate the current branch history with the branch history from the first trap to make more accurate branch predictions . although the present invention and its advantages have been described in detail , it should be understood that various changes , substitutions and alterations can be made herein without departing from the spirit and scope of the invention as defined by the appended claims .", "category": "Textiles; Paper"}
|
Does the category match the content of the patent?
| 0.25 |
fdcc34483564ad9019b21f93490188a72a8ae33eb03472de41376137951476c1
| 0.02063 | 0.000075 | 0.05835 | 0.001167 | 0.03064 | 0.000504 |
null |
{"category": "Physics", "patent": "fig1 a depicts the inventive instruction cache ( icache ) 100 of a processor , and includes long cache lines 101 , 102 , 103 , and 104 . only four lines are depicted for simplicity , icache 100 size is implementation dependent . each cache line includes a tag 105 , which is used to tell a cache hit or miss , a plurality of instruction bundles 106 , and counter / branch information 107 . fig1 b depicts the contents of instruction bundles 106 . each bundle is comprised of a group of instructions 108 that can be issued in the same cycle , for example , bundle 0 includes a load instruction , an add instruction , and a compare - branch instruction . note that each bundle has a fixed number of instructions , however , some of the instructions may be nops . fig1 c depicts the counter / branch information associated with each instruction bundle . each instruction bundle has counter information 109 , which is used to determine whether the code within the bundle is hot code . when the bundle is brought into the icache , the counter is initialized to a threshold value . depending on the threshold value desired , the counter can be as small as 8 to 10 bits . the counter is updated when the instruction bundle is retired from the execution pipeline . each update decrements the counter by 1 . note that the counter could initially be set to zero and increment with each retirement . however , this would require a comparison with a non - zero threshold number , e . g . 100 , which requires more work than comparing with a zero threshold number . each instruction bundle 106 in the icache 100 also maintains a branch history 110 , 111 for each instruction within the bundle . this history describes whether the comparisons in the branch instructions have resulted in a fall through to the next instruction or a branch taken to another instruction . branch history 110 is associated with bundle 0 , including slots a , b , c , which correspond to the instructions within the bundle 0 . thus , it appears one slot in the history is allocated for each instruction in the bundle , whether the instruction is a branch instruction or not . when the instructions from the original binary are brought into the icache , the branch history is cleared . the branch history information is updated when the instruction bundle is retired from fie pipeline . note that the number of instructions ( and thus the number of slots ) is by way of example only , as each bundle could have more or fewer instructions . since the third instruction in bundle 0 is a branch instruction , then slot 110 c has branch information . binary zeros indicate a fall through , and binary ones indicate a branch taken . thus , the information in 110 c , i . e . 00100 , indicates that of the last five times that this instruction has been executed , that the instruction br 1 has fallen through , fallen through , been taken , fallen through , and fallen through . note that the number of bits in the history is by way of example only , and more bits could be used to provide a more detailed history ( while requiring more space ), while fewer bits could be used to save space ( while providing less history ). note that either the most significant bit or the last significant bit may represent the most urgent execution instruction . similarly , the information in 111 b and 111 c describe the histories of instruction br 2 and br 3 respectively . note that br 2 has not recently branched , whereas the previous four executions of br 3 have resulted in the branch taken . in operation , once the counter of a bundle reaches zero , a software component known as the trace selector 201 is invoked , via a special trap , to select a trace . diagnose instructions ( special instructions to diagnose hardware ) are used by the trace selector to examine the icache and the branch history information to form a trace . regular instructions cannot read i - cache contents since i - cache is not part of the architecture states . each processor has a set of diagnose instructions defined ( not visible to application programmer ) which can be used to examine i - cache contents . fig2 a and 2b depict trace formation . fig2 a depicts instruction bundles 106 and their associated branch information 11110 . assume that the counter of bundle 1 ( not shown ) has reached zero , and that bundles 5 - 9 and 14 - 99 are not shown for reason of simplicity . note that bundles 1 - 101 may be in one or more cache lines of icache 100 . the trace selector 201 begins building the trace 202 from the hot code , in this case bundle 1 . the trace selector 201 examines the branch information ( if any ) in bundle 1 to predict whether the branch will be taken or fall through . if there are no branch instructions in the bundle , then bundle will fall through to the next sequential bundle . if the trace selector determines that the branch is most likely to fall through , then the next sequential bundle is added to the trace 202 , in this case it would be bundle 2 . note that if a branch instruction that is in the middle of bundle is assumed to be taken , the remaining instructions of the bundle are not included in the trace . the trace 202 is stored in the trace memory 203 . if the trace selector determines that the branch is most likely to be taken , then the target bundle of the branch is added to the trace 202 , in this case it would be bundle 30 . after examining the branch history 112 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 30 will not be taken , and will add the next sequential bundle , bundle 2 , to the trace 202 , and then will examine bundle 2 . after examining the branch history 113 of bundle 2 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken four times and fallen through once . therefore , the trace selector will predict that the branch to bundle 10 will be taken , and will add the target bundle , bundle 10 , to the trace 202 , and then will examine bundle 10 . after examining the branch history 114 of bundle 10 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 20 will not be taken , and will add the next sequential bundle , bundle 11 , to the trace 202 , and then will examine bundle 11 . bundle 11 does not contain any branch instructions , and therefore will not have a branch history , thus the trace selector 201 will add the next sequential bundle , bundle 12 , to the trace 202 , and then will examine bundle 12 . after examining the branch history 115 of bundle 12 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 24 will not be taken , and will add the next sequential bundle , bundle 13 , to the trace 202 , and then will examine bundle 13 . after examining the branch history 116 of bundle 13 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken four times and fallen through once . therefore , the trace selector will predict that the branch to bundle 101 will be taken , and will add the target bundle , bundle 101 , to the trace 202 , and then will examine bundle 101 . after examining the branch history 117 of bundle 101 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken five times . therefore , the trace selector will predict that the branch to bundle 1 will be taken . the trace selector notes that bundle 1 is already part of the trace 202 in trace memory 203 , via the trace of a sequence of bundles , by examining the address of a backward branch , it can be detected whether the target bundle is already part of a trace . the trace selector then ends the trace or passes the formed trace to the optimizer . the branch to bundle 1 from bundle 101 is known as a backward branch , which forms a loop . at this point , the trace may be stopped , as the trace would merely repeat bundles that are already present in the trace . the trace selector may also end the trace based on other criteria from a set of heuristics including the length of the trace , the number of conditional branches encountered , the probability of accumulated branch predictions and other considerations . thus , a trace may end when its length is a multiple of a cache line size . this would make cache operations easier , as the entire line could be loaded or overwritten without having to be concerned about starting and stopping points in the middle of a cache line . the trace could also end after a certain , predetermined number of conditional branches has been encountered . note that branch histories 113 and 116 do indicate that branch falls through occasionally , and thus the trace would be inaccurate as the trace predicts that the branch will be taken . the predetermined number could be based on the probability of error of the trace . for example , the predetermined number would be low if many of the branches have histories of 00011 or 00111 . on the other hand , the predetermined number would be high if many of the branches have histories of 00000 or 11111 . note that a trace may terminate at an indirect branch since the target address is not known . an indirect branch is different from an ip - relative ( or pc - relative ) branch in that the branch target address cannot be computed directly from the branch instruction . its target is stored either in a register or in a memory location . so the target address is unknown unless the instruction is actually executed . for example , however , the trace selector may decide to grow the trace by predicting its most recent target from the target address cache ( tac ), which is a structure commonly used to predict branch target address . for a return branch which is an indirect branch , with its target being dependent on the call site , the trace selector would know the return address if the call instruction is in the trace , if the call instruction is not in the trace , the trace selector can predict the call site using the top address of the return stack buffer ( rsb ), which is a commonly used data structure to predict return branches . the tac and the rsb are discussed in the co - pending and commonly assigned u . s . patent application entitled efficient mapping to optimized code for processor embedded run - time optimizer ser . no . 09 / 252 , 367 , filed feb . 18 , 1999 , and issued feb . 6 , 2001 , as u . s . pat . no . 6 , 185 , 669 ; which is hereby incorporated by reference . the trace 202 will be stored in the trace memory 203 . there is a mapping from the trace starting instruction bundle in the original binary to the trace in the trace memory . when the trace starting bundle is executed , the mapping will automatically lead the execution to the trace stored in the trace memory 203 . typically , an executed branch instruction has its target in the trace memory . this is discussed in the co - pending and commonly assigned u . s . patent application entitled system and method using a hardware embedded run - time optimizer ser . no . 09 / 252 , 170 , filed feb . 18 , 1999 , and issued sep . 17 , 2002 , as u . s . pat . no . 6 , 453 , 411 , which is hereby incorporated by reference . note that the trace may require more than one cache line . as stated previously , long cache lines are inefficient for original binary . this is because the original binary is loaded sequentially , i . e . bundle 1 , 2 , 3 , 4 , 5 , 6 , etc ., and branches taken within the bundles may result in many of the loaded bundles not being used . for example , suppose bundle 6 has a branch taken to bundle 50 , then the loading of bundles 7 - 49 represent wasted time and cache space as they are not going to be used . however , when the trace is loaded into the cache , the entire trace is almost certain to be used . thus , the long cache lines are much more efficient , because of the sequential locality , as the bundles of the trace will ( almost always ) fall through to the next bundle of the trace . note that a trace usually spans several cache lines . it may not end at the end of a cache line . in this case , the remaining part of the cache line can be the start of another trace . note that since traces are also brought into the icache , the profiling and trace selection may end up generating a trace on top of an existing trace . traces can be identified since their addresses are preserved addresses in physical memory . if their participation in subsequent trace selection is not desired , then when the trace is moved into the icache , the counters associated with the trace will not be initialized to the threshold value , and instead are set to a null value . thus , the trace will not participate in profiling . however , subsequent profiling and trace selection could be used to determine whether the trace is considered \u201c good .\u201d for example , if a trace has frequent early exits , then the trace may need to be regenerated . note that more bits of branch history will allow for more accurate predictions to be made by the trace selector . however , this will require more cache space . alternatively , a multi - tiered system may be used such that the trace selector would not to select a trace when a bundle traps for the first time . instead , the trace selector may record the branch history information of the bundle in another location of memory , and then set the threshold back to a second value , which could be smaller , larger or the same as the original threshold value , and return to execution . when this bundle traps again , the trace selector can accumulate the current branch history with the branch history from the first trap to make more accurate branch predictions . although the present invention and its advantages have been described in detail , it should be understood that various changes , substitutions and alterations can be made herein without departing from the spirit and scope of the invention as defined by the appended claims ."}
|
{"patent": "fig1 a depicts the inventive instruction cache ( icache ) 100 of a processor , and includes long cache lines 101 , 102 , 103 , and 104 . only four lines are depicted for simplicity , icache 100 size is implementation dependent . each cache line includes a tag 105 , which is used to tell a cache hit or miss , a plurality of instruction bundles 106 , and counter / branch information 107 . fig1 b depicts the contents of instruction bundles 106 . each bundle is comprised of a group of instructions 108 that can be issued in the same cycle , for example , bundle 0 includes a load instruction , an add instruction , and a compare - branch instruction . note that each bundle has a fixed number of instructions , however , some of the instructions may be nops . fig1 c depicts the counter / branch information associated with each instruction bundle . each instruction bundle has counter information 109 , which is used to determine whether the code within the bundle is hot code . when the bundle is brought into the icache , the counter is initialized to a threshold value . depending on the threshold value desired , the counter can be as small as 8 to 10 bits . the counter is updated when the instruction bundle is retired from the execution pipeline . each update decrements the counter by 1 . note that the counter could initially be set to zero and increment with each retirement . however , this would require a comparison with a non - zero threshold number , e . g . 100 , which requires more work than comparing with a zero threshold number . each instruction bundle 106 in the icache 100 also maintains a branch history 110 , 111 for each instruction within the bundle . this history describes whether the comparisons in the branch instructions have resulted in a fall through to the next instruction or a branch taken to another instruction . branch history 110 is associated with bundle 0 , including slots a , b , c , which correspond to the instructions within the bundle 0 . thus , it appears one slot in the history is allocated for each instruction in the bundle , whether the instruction is a branch instruction or not . when the instructions from the original binary are brought into the icache , the branch history is cleared . the branch history information is updated when the instruction bundle is retired from fie pipeline . note that the number of instructions ( and thus the number of slots ) is by way of example only , as each bundle could have more or fewer instructions . since the third instruction in bundle 0 is a branch instruction , then slot 110 c has branch information . binary zeros indicate a fall through , and binary ones indicate a branch taken . thus , the information in 110 c , i . e . 00100 , indicates that of the last five times that this instruction has been executed , that the instruction br 1 has fallen through , fallen through , been taken , fallen through , and fallen through . note that the number of bits in the history is by way of example only , and more bits could be used to provide a more detailed history ( while requiring more space ), while fewer bits could be used to save space ( while providing less history ). note that either the most significant bit or the last significant bit may represent the most urgent execution instruction . similarly , the information in 111 b and 111 c describe the histories of instruction br 2 and br 3 respectively . note that br 2 has not recently branched , whereas the previous four executions of br 3 have resulted in the branch taken . in operation , once the counter of a bundle reaches zero , a software component known as the trace selector 201 is invoked , via a special trap , to select a trace . diagnose instructions ( special instructions to diagnose hardware ) are used by the trace selector to examine the icache and the branch history information to form a trace . regular instructions cannot read i - cache contents since i - cache is not part of the architecture states . each processor has a set of diagnose instructions defined ( not visible to application programmer ) which can be used to examine i - cache contents . fig2 a and 2b depict trace formation . fig2 a depicts instruction bundles 106 and their associated branch information 11110 . assume that the counter of bundle 1 ( not shown ) has reached zero , and that bundles 5 - 9 and 14 - 99 are not shown for reason of simplicity . note that bundles 1 - 101 may be in one or more cache lines of icache 100 . the trace selector 201 begins building the trace 202 from the hot code , in this case bundle 1 . the trace selector 201 examines the branch information ( if any ) in bundle 1 to predict whether the branch will be taken or fall through . if there are no branch instructions in the bundle , then bundle will fall through to the next sequential bundle . if the trace selector determines that the branch is most likely to fall through , then the next sequential bundle is added to the trace 202 , in this case it would be bundle 2 . note that if a branch instruction that is in the middle of bundle is assumed to be taken , the remaining instructions of the bundle are not included in the trace . the trace 202 is stored in the trace memory 203 . if the trace selector determines that the branch is most likely to be taken , then the target bundle of the branch is added to the trace 202 , in this case it would be bundle 30 . after examining the branch history 112 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 30 will not be taken , and will add the next sequential bundle , bundle 2 , to the trace 202 , and then will examine bundle 2 . after examining the branch history 113 of bundle 2 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken four times and fallen through once . therefore , the trace selector will predict that the branch to bundle 10 will be taken , and will add the target bundle , bundle 10 , to the trace 202 , and then will examine bundle 10 . after examining the branch history 114 of bundle 10 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 20 will not be taken , and will add the next sequential bundle , bundle 11 , to the trace 202 , and then will examine bundle 11 . bundle 11 does not contain any branch instructions , and therefore will not have a branch history , thus the trace selector 201 will add the next sequential bundle , bundle 12 , to the trace 202 , and then will examine bundle 12 . after examining the branch history 115 of bundle 12 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 24 will not be taken , and will add the next sequential bundle , bundle 13 , to the trace 202 , and then will examine bundle 13 . after examining the branch history 116 of bundle 13 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken four times and fallen through once . therefore , the trace selector will predict that the branch to bundle 101 will be taken , and will add the target bundle , bundle 101 , to the trace 202 , and then will examine bundle 101 . after examining the branch history 117 of bundle 101 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken five times . therefore , the trace selector will predict that the branch to bundle 1 will be taken . the trace selector notes that bundle 1 is already part of the trace 202 in trace memory 203 , via the trace of a sequence of bundles , by examining the address of a backward branch , it can be detected whether the target bundle is already part of a trace . the trace selector then ends the trace or passes the formed trace to the optimizer . the branch to bundle 1 from bundle 101 is known as a backward branch , which forms a loop . at this point , the trace may be stopped , as the trace would merely repeat bundles that are already present in the trace . the trace selector may also end the trace based on other criteria from a set of heuristics including the length of the trace , the number of conditional branches encountered , the probability of accumulated branch predictions and other considerations . thus , a trace may end when its length is a multiple of a cache line size . this would make cache operations easier , as the entire line could be loaded or overwritten without having to be concerned about starting and stopping points in the middle of a cache line . the trace could also end after a certain , predetermined number of conditional branches has been encountered . note that branch histories 113 and 116 do indicate that branch falls through occasionally , and thus the trace would be inaccurate as the trace predicts that the branch will be taken . the predetermined number could be based on the probability of error of the trace . for example , the predetermined number would be low if many of the branches have histories of 00011 or 00111 . on the other hand , the predetermined number would be high if many of the branches have histories of 00000 or 11111 . note that a trace may terminate at an indirect branch since the target address is not known . an indirect branch is different from an ip - relative ( or pc - relative ) branch in that the branch target address cannot be computed directly from the branch instruction . its target is stored either in a register or in a memory location . so the target address is unknown unless the instruction is actually executed . for example , however , the trace selector may decide to grow the trace by predicting its most recent target from the target address cache ( tac ), which is a structure commonly used to predict branch target address . for a return branch which is an indirect branch , with its target being dependent on the call site , the trace selector would know the return address if the call instruction is in the trace , if the call instruction is not in the trace , the trace selector can predict the call site using the top address of the return stack buffer ( rsb ), which is a commonly used data structure to predict return branches . the tac and the rsb are discussed in the co - pending and commonly assigned u . s . patent application entitled efficient mapping to optimized code for processor embedded run - time optimizer ser . no . 09 / 252 , 367 , filed feb . 18 , 1999 , and issued feb . 6 , 2001 , as u . s . pat . no . 6 , 185 , 669 ; which is hereby incorporated by reference . the trace 202 will be stored in the trace memory 203 . there is a mapping from the trace starting instruction bundle in the original binary to the trace in the trace memory . when the trace starting bundle is executed , the mapping will automatically lead the execution to the trace stored in the trace memory 203 . typically , an executed branch instruction has its target in the trace memory . this is discussed in the co - pending and commonly assigned u . s . patent application entitled system and method using a hardware embedded run - time optimizer ser . no . 09 / 252 , 170 , filed feb . 18 , 1999 , and issued sep . 17 , 2002 , as u . s . pat . no . 6 , 453 , 411 , which is hereby incorporated by reference . note that the trace may require more than one cache line . as stated previously , long cache lines are inefficient for original binary . this is because the original binary is loaded sequentially , i . e . bundle 1 , 2 , 3 , 4 , 5 , 6 , etc ., and branches taken within the bundles may result in many of the loaded bundles not being used . for example , suppose bundle 6 has a branch taken to bundle 50 , then the loading of bundles 7 - 49 represent wasted time and cache space as they are not going to be used . however , when the trace is loaded into the cache , the entire trace is almost certain to be used . thus , the long cache lines are much more efficient , because of the sequential locality , as the bundles of the trace will ( almost always ) fall through to the next bundle of the trace . note that a trace usually spans several cache lines . it may not end at the end of a cache line . in this case , the remaining part of the cache line can be the start of another trace . note that since traces are also brought into the icache , the profiling and trace selection may end up generating a trace on top of an existing trace . traces can be identified since their addresses are preserved addresses in physical memory . if their participation in subsequent trace selection is not desired , then when the trace is moved into the icache , the counters associated with the trace will not be initialized to the threshold value , and instead are set to a null value . thus , the trace will not participate in profiling . however , subsequent profiling and trace selection could be used to determine whether the trace is considered \u201c good .\u201d for example , if a trace has frequent early exits , then the trace may need to be regenerated . note that more bits of branch history will allow for more accurate predictions to be made by the trace selector . however , this will require more cache space . alternatively , a multi - tiered system may be used such that the trace selector would not to select a trace when a bundle traps for the first time . instead , the trace selector may record the branch history information of the bundle in another location of memory , and then set the threshold back to a second value , which could be smaller , larger or the same as the original threshold value , and return to execution . when this bundle traps again , the trace selector can accumulate the current branch history with the branch history from the first trap to make more accurate branch predictions . although the present invention and its advantages have been described in detail , it should be understood that various changes , substitutions and alterations can be made herein without departing from the spirit and scope of the invention as defined by the appended claims .", "category": "Fixed Constructions"}
|
Does the patent belong in this category?
| 0.25 |
fdcc34483564ad9019b21f93490188a72a8ae33eb03472de41376137951476c1
| 0.012817 | 0.018555 | 0.028931 | 0.084961 | 0.119141 | 0.10498 |
null |
{"category": "Physics", "patent": "fig1 a depicts the inventive instruction cache ( icache ) 100 of a processor , and includes long cache lines 101 , 102 , 103 , and 104 . only four lines are depicted for simplicity , icache 100 size is implementation dependent . each cache line includes a tag 105 , which is used to tell a cache hit or miss , a plurality of instruction bundles 106 , and counter / branch information 107 . fig1 b depicts the contents of instruction bundles 106 . each bundle is comprised of a group of instructions 108 that can be issued in the same cycle , for example , bundle 0 includes a load instruction , an add instruction , and a compare - branch instruction . note that each bundle has a fixed number of instructions , however , some of the instructions may be nops . fig1 c depicts the counter / branch information associated with each instruction bundle . each instruction bundle has counter information 109 , which is used to determine whether the code within the bundle is hot code . when the bundle is brought into the icache , the counter is initialized to a threshold value . depending on the threshold value desired , the counter can be as small as 8 to 10 bits . the counter is updated when the instruction bundle is retired from the execution pipeline . each update decrements the counter by 1 . note that the counter could initially be set to zero and increment with each retirement . however , this would require a comparison with a non - zero threshold number , e . g . 100 , which requires more work than comparing with a zero threshold number . each instruction bundle 106 in the icache 100 also maintains a branch history 110 , 111 for each instruction within the bundle . this history describes whether the comparisons in the branch instructions have resulted in a fall through to the next instruction or a branch taken to another instruction . branch history 110 is associated with bundle 0 , including slots a , b , c , which correspond to the instructions within the bundle 0 . thus , it appears one slot in the history is allocated for each instruction in the bundle , whether the instruction is a branch instruction or not . when the instructions from the original binary are brought into the icache , the branch history is cleared . the branch history information is updated when the instruction bundle is retired from fie pipeline . note that the number of instructions ( and thus the number of slots ) is by way of example only , as each bundle could have more or fewer instructions . since the third instruction in bundle 0 is a branch instruction , then slot 110 c has branch information . binary zeros indicate a fall through , and binary ones indicate a branch taken . thus , the information in 110 c , i . e . 00100 , indicates that of the last five times that this instruction has been executed , that the instruction br 1 has fallen through , fallen through , been taken , fallen through , and fallen through . note that the number of bits in the history is by way of example only , and more bits could be used to provide a more detailed history ( while requiring more space ), while fewer bits could be used to save space ( while providing less history ). note that either the most significant bit or the last significant bit may represent the most urgent execution instruction . similarly , the information in 111 b and 111 c describe the histories of instruction br 2 and br 3 respectively . note that br 2 has not recently branched , whereas the previous four executions of br 3 have resulted in the branch taken . in operation , once the counter of a bundle reaches zero , a software component known as the trace selector 201 is invoked , via a special trap , to select a trace . diagnose instructions ( special instructions to diagnose hardware ) are used by the trace selector to examine the icache and the branch history information to form a trace . regular instructions cannot read i - cache contents since i - cache is not part of the architecture states . each processor has a set of diagnose instructions defined ( not visible to application programmer ) which can be used to examine i - cache contents . fig2 a and 2b depict trace formation . fig2 a depicts instruction bundles 106 and their associated branch information 11110 . assume that the counter of bundle 1 ( not shown ) has reached zero , and that bundles 5 - 9 and 14 - 99 are not shown for reason of simplicity . note that bundles 1 - 101 may be in one or more cache lines of icache 100 . the trace selector 201 begins building the trace 202 from the hot code , in this case bundle 1 . the trace selector 201 examines the branch information ( if any ) in bundle 1 to predict whether the branch will be taken or fall through . if there are no branch instructions in the bundle , then bundle will fall through to the next sequential bundle . if the trace selector determines that the branch is most likely to fall through , then the next sequential bundle is added to the trace 202 , in this case it would be bundle 2 . note that if a branch instruction that is in the middle of bundle is assumed to be taken , the remaining instructions of the bundle are not included in the trace . the trace 202 is stored in the trace memory 203 . if the trace selector determines that the branch is most likely to be taken , then the target bundle of the branch is added to the trace 202 , in this case it would be bundle 30 . after examining the branch history 112 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 30 will not be taken , and will add the next sequential bundle , bundle 2 , to the trace 202 , and then will examine bundle 2 . after examining the branch history 113 of bundle 2 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken four times and fallen through once . therefore , the trace selector will predict that the branch to bundle 10 will be taken , and will add the target bundle , bundle 10 , to the trace 202 , and then will examine bundle 10 . after examining the branch history 114 of bundle 10 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 20 will not be taken , and will add the next sequential bundle , bundle 11 , to the trace 202 , and then will examine bundle 11 . bundle 11 does not contain any branch instructions , and therefore will not have a branch history , thus the trace selector 201 will add the next sequential bundle , bundle 12 , to the trace 202 , and then will examine bundle 12 . after examining the branch history 115 of bundle 12 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 24 will not be taken , and will add the next sequential bundle , bundle 13 , to the trace 202 , and then will examine bundle 13 . after examining the branch history 116 of bundle 13 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken four times and fallen through once . therefore , the trace selector will predict that the branch to bundle 101 will be taken , and will add the target bundle , bundle 101 , to the trace 202 , and then will examine bundle 101 . after examining the branch history 117 of bundle 101 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken five times . therefore , the trace selector will predict that the branch to bundle 1 will be taken . the trace selector notes that bundle 1 is already part of the trace 202 in trace memory 203 , via the trace of a sequence of bundles , by examining the address of a backward branch , it can be detected whether the target bundle is already part of a trace . the trace selector then ends the trace or passes the formed trace to the optimizer . the branch to bundle 1 from bundle 101 is known as a backward branch , which forms a loop . at this point , the trace may be stopped , as the trace would merely repeat bundles that are already present in the trace . the trace selector may also end the trace based on other criteria from a set of heuristics including the length of the trace , the number of conditional branches encountered , the probability of accumulated branch predictions and other considerations . thus , a trace may end when its length is a multiple of a cache line size . this would make cache operations easier , as the entire line could be loaded or overwritten without having to be concerned about starting and stopping points in the middle of a cache line . the trace could also end after a certain , predetermined number of conditional branches has been encountered . note that branch histories 113 and 116 do indicate that branch falls through occasionally , and thus the trace would be inaccurate as the trace predicts that the branch will be taken . the predetermined number could be based on the probability of error of the trace . for example , the predetermined number would be low if many of the branches have histories of 00011 or 00111 . on the other hand , the predetermined number would be high if many of the branches have histories of 00000 or 11111 . note that a trace may terminate at an indirect branch since the target address is not known . an indirect branch is different from an ip - relative ( or pc - relative ) branch in that the branch target address cannot be computed directly from the branch instruction . its target is stored either in a register or in a memory location . so the target address is unknown unless the instruction is actually executed . for example , however , the trace selector may decide to grow the trace by predicting its most recent target from the target address cache ( tac ), which is a structure commonly used to predict branch target address . for a return branch which is an indirect branch , with its target being dependent on the call site , the trace selector would know the return address if the call instruction is in the trace , if the call instruction is not in the trace , the trace selector can predict the call site using the top address of the return stack buffer ( rsb ), which is a commonly used data structure to predict return branches . the tac and the rsb are discussed in the co - pending and commonly assigned u . s . patent application entitled efficient mapping to optimized code for processor embedded run - time optimizer ser . no . 09 / 252 , 367 , filed feb . 18 , 1999 , and issued feb . 6 , 2001 , as u . s . pat . no . 6 , 185 , 669 ; which is hereby incorporated by reference . the trace 202 will be stored in the trace memory 203 . there is a mapping from the trace starting instruction bundle in the original binary to the trace in the trace memory . when the trace starting bundle is executed , the mapping will automatically lead the execution to the trace stored in the trace memory 203 . typically , an executed branch instruction has its target in the trace memory . this is discussed in the co - pending and commonly assigned u . s . patent application entitled system and method using a hardware embedded run - time optimizer ser . no . 09 / 252 , 170 , filed feb . 18 , 1999 , and issued sep . 17 , 2002 , as u . s . pat . no . 6 , 453 , 411 , which is hereby incorporated by reference . note that the trace may require more than one cache line . as stated previously , long cache lines are inefficient for original binary . this is because the original binary is loaded sequentially , i . e . bundle 1 , 2 , 3 , 4 , 5 , 6 , etc ., and branches taken within the bundles may result in many of the loaded bundles not being used . for example , suppose bundle 6 has a branch taken to bundle 50 , then the loading of bundles 7 - 49 represent wasted time and cache space as they are not going to be used . however , when the trace is loaded into the cache , the entire trace is almost certain to be used . thus , the long cache lines are much more efficient , because of the sequential locality , as the bundles of the trace will ( almost always ) fall through to the next bundle of the trace . note that a trace usually spans several cache lines . it may not end at the end of a cache line . in this case , the remaining part of the cache line can be the start of another trace . note that since traces are also brought into the icache , the profiling and trace selection may end up generating a trace on top of an existing trace . traces can be identified since their addresses are preserved addresses in physical memory . if their participation in subsequent trace selection is not desired , then when the trace is moved into the icache , the counters associated with the trace will not be initialized to the threshold value , and instead are set to a null value . thus , the trace will not participate in profiling . however , subsequent profiling and trace selection could be used to determine whether the trace is considered \u201c good .\u201d for example , if a trace has frequent early exits , then the trace may need to be regenerated . note that more bits of branch history will allow for more accurate predictions to be made by the trace selector . however , this will require more cache space . alternatively , a multi - tiered system may be used such that the trace selector would not to select a trace when a bundle traps for the first time . instead , the trace selector may record the branch history information of the bundle in another location of memory , and then set the threshold back to a second value , which could be smaller , larger or the same as the original threshold value , and return to execution . when this bundle traps again , the trace selector can accumulate the current branch history with the branch history from the first trap to make more accurate branch predictions . although the present invention and its advantages have been described in detail , it should be understood that various changes , substitutions and alterations can be made herein without departing from the spirit and scope of the invention as defined by the appended claims ."}
|
{"category": "Mechanical Engineering; Lightning; Heating; Weapons; Blasting", "patent": "fig1 a depicts the inventive instruction cache ( icache ) 100 of a processor , and includes long cache lines 101 , 102 , 103 , and 104 . only four lines are depicted for simplicity , icache 100 size is implementation dependent . each cache line includes a tag 105 , which is used to tell a cache hit or miss , a plurality of instruction bundles 106 , and counter / branch information 107 . fig1 b depicts the contents of instruction bundles 106 . each bundle is comprised of a group of instructions 108 that can be issued in the same cycle , for example , bundle 0 includes a load instruction , an add instruction , and a compare - branch instruction . note that each bundle has a fixed number of instructions , however , some of the instructions may be nops . fig1 c depicts the counter / branch information associated with each instruction bundle . each instruction bundle has counter information 109 , which is used to determine whether the code within the bundle is hot code . when the bundle is brought into the icache , the counter is initialized to a threshold value . depending on the threshold value desired , the counter can be as small as 8 to 10 bits . the counter is updated when the instruction bundle is retired from the execution pipeline . each update decrements the counter by 1 . note that the counter could initially be set to zero and increment with each retirement . however , this would require a comparison with a non - zero threshold number , e . g . 100 , which requires more work than comparing with a zero threshold number . each instruction bundle 106 in the icache 100 also maintains a branch history 110 , 111 for each instruction within the bundle . this history describes whether the comparisons in the branch instructions have resulted in a fall through to the next instruction or a branch taken to another instruction . branch history 110 is associated with bundle 0 , including slots a , b , c , which correspond to the instructions within the bundle 0 . thus , it appears one slot in the history is allocated for each instruction in the bundle , whether the instruction is a branch instruction or not . when the instructions from the original binary are brought into the icache , the branch history is cleared . the branch history information is updated when the instruction bundle is retired from fie pipeline . note that the number of instructions ( and thus the number of slots ) is by way of example only , as each bundle could have more or fewer instructions . since the third instruction in bundle 0 is a branch instruction , then slot 110 c has branch information . binary zeros indicate a fall through , and binary ones indicate a branch taken . thus , the information in 110 c , i . e . 00100 , indicates that of the last five times that this instruction has been executed , that the instruction br 1 has fallen through , fallen through , been taken , fallen through , and fallen through . note that the number of bits in the history is by way of example only , and more bits could be used to provide a more detailed history ( while requiring more space ), while fewer bits could be used to save space ( while providing less history ). note that either the most significant bit or the last significant bit may represent the most urgent execution instruction . similarly , the information in 111 b and 111 c describe the histories of instruction br 2 and br 3 respectively . note that br 2 has not recently branched , whereas the previous four executions of br 3 have resulted in the branch taken . in operation , once the counter of a bundle reaches zero , a software component known as the trace selector 201 is invoked , via a special trap , to select a trace . diagnose instructions ( special instructions to diagnose hardware ) are used by the trace selector to examine the icache and the branch history information to form a trace . regular instructions cannot read i - cache contents since i - cache is not part of the architecture states . each processor has a set of diagnose instructions defined ( not visible to application programmer ) which can be used to examine i - cache contents . fig2 a and 2b depict trace formation . fig2 a depicts instruction bundles 106 and their associated branch information 11110 . assume that the counter of bundle 1 ( not shown ) has reached zero , and that bundles 5 - 9 and 14 - 99 are not shown for reason of simplicity . note that bundles 1 - 101 may be in one or more cache lines of icache 100 . the trace selector 201 begins building the trace 202 from the hot code , in this case bundle 1 . the trace selector 201 examines the branch information ( if any ) in bundle 1 to predict whether the branch will be taken or fall through . if there are no branch instructions in the bundle , then bundle will fall through to the next sequential bundle . if the trace selector determines that the branch is most likely to fall through , then the next sequential bundle is added to the trace 202 , in this case it would be bundle 2 . note that if a branch instruction that is in the middle of bundle is assumed to be taken , the remaining instructions of the bundle are not included in the trace . the trace 202 is stored in the trace memory 203 . if the trace selector determines that the branch is most likely to be taken , then the target bundle of the branch is added to the trace 202 , in this case it would be bundle 30 . after examining the branch history 112 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 30 will not be taken , and will add the next sequential bundle , bundle 2 , to the trace 202 , and then will examine bundle 2 . after examining the branch history 113 of bundle 2 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken four times and fallen through once . therefore , the trace selector will predict that the branch to bundle 10 will be taken , and will add the target bundle , bundle 10 , to the trace 202 , and then will examine bundle 10 . after examining the branch history 114 of bundle 10 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 20 will not be taken , and will add the next sequential bundle , bundle 11 , to the trace 202 , and then will examine bundle 11 . bundle 11 does not contain any branch instructions , and therefore will not have a branch history , thus the trace selector 201 will add the next sequential bundle , bundle 12 , to the trace 202 , and then will examine bundle 12 . after examining the branch history 115 of bundle 12 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 24 will not be taken , and will add the next sequential bundle , bundle 13 , to the trace 202 , and then will examine bundle 13 . after examining the branch history 116 of bundle 13 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken four times and fallen through once . therefore , the trace selector will predict that the branch to bundle 101 will be taken , and will add the target bundle , bundle 101 , to the trace 202 , and then will examine bundle 101 . after examining the branch history 117 of bundle 101 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken five times . therefore , the trace selector will predict that the branch to bundle 1 will be taken . the trace selector notes that bundle 1 is already part of the trace 202 in trace memory 203 , via the trace of a sequence of bundles , by examining the address of a backward branch , it can be detected whether the target bundle is already part of a trace . the trace selector then ends the trace or passes the formed trace to the optimizer . the branch to bundle 1 from bundle 101 is known as a backward branch , which forms a loop . at this point , the trace may be stopped , as the trace would merely repeat bundles that are already present in the trace . the trace selector may also end the trace based on other criteria from a set of heuristics including the length of the trace , the number of conditional branches encountered , the probability of accumulated branch predictions and other considerations . thus , a trace may end when its length is a multiple of a cache line size . this would make cache operations easier , as the entire line could be loaded or overwritten without having to be concerned about starting and stopping points in the middle of a cache line . the trace could also end after a certain , predetermined number of conditional branches has been encountered . note that branch histories 113 and 116 do indicate that branch falls through occasionally , and thus the trace would be inaccurate as the trace predicts that the branch will be taken . the predetermined number could be based on the probability of error of the trace . for example , the predetermined number would be low if many of the branches have histories of 00011 or 00111 . on the other hand , the predetermined number would be high if many of the branches have histories of 00000 or 11111 . note that a trace may terminate at an indirect branch since the target address is not known . an indirect branch is different from an ip - relative ( or pc - relative ) branch in that the branch target address cannot be computed directly from the branch instruction . its target is stored either in a register or in a memory location . so the target address is unknown unless the instruction is actually executed . for example , however , the trace selector may decide to grow the trace by predicting its most recent target from the target address cache ( tac ), which is a structure commonly used to predict branch target address . for a return branch which is an indirect branch , with its target being dependent on the call site , the trace selector would know the return address if the call instruction is in the trace , if the call instruction is not in the trace , the trace selector can predict the call site using the top address of the return stack buffer ( rsb ), which is a commonly used data structure to predict return branches . the tac and the rsb are discussed in the co - pending and commonly assigned u . s . patent application entitled efficient mapping to optimized code for processor embedded run - time optimizer ser . no . 09 / 252 , 367 , filed feb . 18 , 1999 , and issued feb . 6 , 2001 , as u . s . pat . no . 6 , 185 , 669 ; which is hereby incorporated by reference . the trace 202 will be stored in the trace memory 203 . there is a mapping from the trace starting instruction bundle in the original binary to the trace in the trace memory . when the trace starting bundle is executed , the mapping will automatically lead the execution to the trace stored in the trace memory 203 . typically , an executed branch instruction has its target in the trace memory . this is discussed in the co - pending and commonly assigned u . s . patent application entitled system and method using a hardware embedded run - time optimizer ser . no . 09 / 252 , 170 , filed feb . 18 , 1999 , and issued sep . 17 , 2002 , as u . s . pat . no . 6 , 453 , 411 , which is hereby incorporated by reference . note that the trace may require more than one cache line . as stated previously , long cache lines are inefficient for original binary . this is because the original binary is loaded sequentially , i . e . bundle 1 , 2 , 3 , 4 , 5 , 6 , etc ., and branches taken within the bundles may result in many of the loaded bundles not being used . for example , suppose bundle 6 has a branch taken to bundle 50 , then the loading of bundles 7 - 49 represent wasted time and cache space as they are not going to be used . however , when the trace is loaded into the cache , the entire trace is almost certain to be used . thus , the long cache lines are much more efficient , because of the sequential locality , as the bundles of the trace will ( almost always ) fall through to the next bundle of the trace . note that a trace usually spans several cache lines . it may not end at the end of a cache line . in this case , the remaining part of the cache line can be the start of another trace . note that since traces are also brought into the icache , the profiling and trace selection may end up generating a trace on top of an existing trace . traces can be identified since their addresses are preserved addresses in physical memory . if their participation in subsequent trace selection is not desired , then when the trace is moved into the icache , the counters associated with the trace will not be initialized to the threshold value , and instead are set to a null value . thus , the trace will not participate in profiling . however , subsequent profiling and trace selection could be used to determine whether the trace is considered \u201c good .\u201d for example , if a trace has frequent early exits , then the trace may need to be regenerated . note that more bits of branch history will allow for more accurate predictions to be made by the trace selector . however , this will require more cache space . alternatively , a multi - tiered system may be used such that the trace selector would not to select a trace when a bundle traps for the first time . instead , the trace selector may record the branch history information of the bundle in another location of memory , and then set the threshold back to a second value , which could be smaller , larger or the same as the original threshold value , and return to execution . when this bundle traps again , the trace selector can accumulate the current branch history with the branch history from the first trap to make more accurate branch predictions . although the present invention and its advantages have been described in detail , it should be understood that various changes , substitutions and alterations can be made herein without departing from the spirit and scope of the invention as defined by the appended claims ."}
|
Does the category match the content of the patent?
| 0.25 |
fdcc34483564ad9019b21f93490188a72a8ae33eb03472de41376137951476c1
| 0.021606 | 0.001503 | 0.05835 | 0.000969 | 0.03418 | 0.004608 |
null |
{"category": "Physics", "patent": "fig1 a depicts the inventive instruction cache ( icache ) 100 of a processor , and includes long cache lines 101 , 102 , 103 , and 104 . only four lines are depicted for simplicity , icache 100 size is implementation dependent . each cache line includes a tag 105 , which is used to tell a cache hit or miss , a plurality of instruction bundles 106 , and counter / branch information 107 . fig1 b depicts the contents of instruction bundles 106 . each bundle is comprised of a group of instructions 108 that can be issued in the same cycle , for example , bundle 0 includes a load instruction , an add instruction , and a compare - branch instruction . note that each bundle has a fixed number of instructions , however , some of the instructions may be nops . fig1 c depicts the counter / branch information associated with each instruction bundle . each instruction bundle has counter information 109 , which is used to determine whether the code within the bundle is hot code . when the bundle is brought into the icache , the counter is initialized to a threshold value . depending on the threshold value desired , the counter can be as small as 8 to 10 bits . the counter is updated when the instruction bundle is retired from the execution pipeline . each update decrements the counter by 1 . note that the counter could initially be set to zero and increment with each retirement . however , this would require a comparison with a non - zero threshold number , e . g . 100 , which requires more work than comparing with a zero threshold number . each instruction bundle 106 in the icache 100 also maintains a branch history 110 , 111 for each instruction within the bundle . this history describes whether the comparisons in the branch instructions have resulted in a fall through to the next instruction or a branch taken to another instruction . branch history 110 is associated with bundle 0 , including slots a , b , c , which correspond to the instructions within the bundle 0 . thus , it appears one slot in the history is allocated for each instruction in the bundle , whether the instruction is a branch instruction or not . when the instructions from the original binary are brought into the icache , the branch history is cleared . the branch history information is updated when the instruction bundle is retired from fie pipeline . note that the number of instructions ( and thus the number of slots ) is by way of example only , as each bundle could have more or fewer instructions . since the third instruction in bundle 0 is a branch instruction , then slot 110 c has branch information . binary zeros indicate a fall through , and binary ones indicate a branch taken . thus , the information in 110 c , i . e . 00100 , indicates that of the last five times that this instruction has been executed , that the instruction br 1 has fallen through , fallen through , been taken , fallen through , and fallen through . note that the number of bits in the history is by way of example only , and more bits could be used to provide a more detailed history ( while requiring more space ), while fewer bits could be used to save space ( while providing less history ). note that either the most significant bit or the last significant bit may represent the most urgent execution instruction . similarly , the information in 111 b and 111 c describe the histories of instruction br 2 and br 3 respectively . note that br 2 has not recently branched , whereas the previous four executions of br 3 have resulted in the branch taken . in operation , once the counter of a bundle reaches zero , a software component known as the trace selector 201 is invoked , via a special trap , to select a trace . diagnose instructions ( special instructions to diagnose hardware ) are used by the trace selector to examine the icache and the branch history information to form a trace . regular instructions cannot read i - cache contents since i - cache is not part of the architecture states . each processor has a set of diagnose instructions defined ( not visible to application programmer ) which can be used to examine i - cache contents . fig2 a and 2b depict trace formation . fig2 a depicts instruction bundles 106 and their associated branch information 11110 . assume that the counter of bundle 1 ( not shown ) has reached zero , and that bundles 5 - 9 and 14 - 99 are not shown for reason of simplicity . note that bundles 1 - 101 may be in one or more cache lines of icache 100 . the trace selector 201 begins building the trace 202 from the hot code , in this case bundle 1 . the trace selector 201 examines the branch information ( if any ) in bundle 1 to predict whether the branch will be taken or fall through . if there are no branch instructions in the bundle , then bundle will fall through to the next sequential bundle . if the trace selector determines that the branch is most likely to fall through , then the next sequential bundle is added to the trace 202 , in this case it would be bundle 2 . note that if a branch instruction that is in the middle of bundle is assumed to be taken , the remaining instructions of the bundle are not included in the trace . the trace 202 is stored in the trace memory 203 . if the trace selector determines that the branch is most likely to be taken , then the target bundle of the branch is added to the trace 202 , in this case it would be bundle 30 . after examining the branch history 112 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 30 will not be taken , and will add the next sequential bundle , bundle 2 , to the trace 202 , and then will examine bundle 2 . after examining the branch history 113 of bundle 2 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken four times and fallen through once . therefore , the trace selector will predict that the branch to bundle 10 will be taken , and will add the target bundle , bundle 10 , to the trace 202 , and then will examine bundle 10 . after examining the branch history 114 of bundle 10 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 20 will not be taken , and will add the next sequential bundle , bundle 11 , to the trace 202 , and then will examine bundle 11 . bundle 11 does not contain any branch instructions , and therefore will not have a branch history , thus the trace selector 201 will add the next sequential bundle , bundle 12 , to the trace 202 , and then will examine bundle 12 . after examining the branch history 115 of bundle 12 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 24 will not be taken , and will add the next sequential bundle , bundle 13 , to the trace 202 , and then will examine bundle 13 . after examining the branch history 116 of bundle 13 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken four times and fallen through once . therefore , the trace selector will predict that the branch to bundle 101 will be taken , and will add the target bundle , bundle 101 , to the trace 202 , and then will examine bundle 101 . after examining the branch history 117 of bundle 101 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken five times . therefore , the trace selector will predict that the branch to bundle 1 will be taken . the trace selector notes that bundle 1 is already part of the trace 202 in trace memory 203 , via the trace of a sequence of bundles , by examining the address of a backward branch , it can be detected whether the target bundle is already part of a trace . the trace selector then ends the trace or passes the formed trace to the optimizer . the branch to bundle 1 from bundle 101 is known as a backward branch , which forms a loop . at this point , the trace may be stopped , as the trace would merely repeat bundles that are already present in the trace . the trace selector may also end the trace based on other criteria from a set of heuristics including the length of the trace , the number of conditional branches encountered , the probability of accumulated branch predictions and other considerations . thus , a trace may end when its length is a multiple of a cache line size . this would make cache operations easier , as the entire line could be loaded or overwritten without having to be concerned about starting and stopping points in the middle of a cache line . the trace could also end after a certain , predetermined number of conditional branches has been encountered . note that branch histories 113 and 116 do indicate that branch falls through occasionally , and thus the trace would be inaccurate as the trace predicts that the branch will be taken . the predetermined number could be based on the probability of error of the trace . for example , the predetermined number would be low if many of the branches have histories of 00011 or 00111 . on the other hand , the predetermined number would be high if many of the branches have histories of 00000 or 11111 . note that a trace may terminate at an indirect branch since the target address is not known . an indirect branch is different from an ip - relative ( or pc - relative ) branch in that the branch target address cannot be computed directly from the branch instruction . its target is stored either in a register or in a memory location . so the target address is unknown unless the instruction is actually executed . for example , however , the trace selector may decide to grow the trace by predicting its most recent target from the target address cache ( tac ), which is a structure commonly used to predict branch target address . for a return branch which is an indirect branch , with its target being dependent on the call site , the trace selector would know the return address if the call instruction is in the trace , if the call instruction is not in the trace , the trace selector can predict the call site using the top address of the return stack buffer ( rsb ), which is a commonly used data structure to predict return branches . the tac and the rsb are discussed in the co - pending and commonly assigned u . s . patent application entitled efficient mapping to optimized code for processor embedded run - time optimizer ser . no . 09 / 252 , 367 , filed feb . 18 , 1999 , and issued feb . 6 , 2001 , as u . s . pat . no . 6 , 185 , 669 ; which is hereby incorporated by reference . the trace 202 will be stored in the trace memory 203 . there is a mapping from the trace starting instruction bundle in the original binary to the trace in the trace memory . when the trace starting bundle is executed , the mapping will automatically lead the execution to the trace stored in the trace memory 203 . typically , an executed branch instruction has its target in the trace memory . this is discussed in the co - pending and commonly assigned u . s . patent application entitled system and method using a hardware embedded run - time optimizer ser . no . 09 / 252 , 170 , filed feb . 18 , 1999 , and issued sep . 17 , 2002 , as u . s . pat . no . 6 , 453 , 411 , which is hereby incorporated by reference . note that the trace may require more than one cache line . as stated previously , long cache lines are inefficient for original binary . this is because the original binary is loaded sequentially , i . e . bundle 1 , 2 , 3 , 4 , 5 , 6 , etc ., and branches taken within the bundles may result in many of the loaded bundles not being used . for example , suppose bundle 6 has a branch taken to bundle 50 , then the loading of bundles 7 - 49 represent wasted time and cache space as they are not going to be used . however , when the trace is loaded into the cache , the entire trace is almost certain to be used . thus , the long cache lines are much more efficient , because of the sequential locality , as the bundles of the trace will ( almost always ) fall through to the next bundle of the trace . note that a trace usually spans several cache lines . it may not end at the end of a cache line . in this case , the remaining part of the cache line can be the start of another trace . note that since traces are also brought into the icache , the profiling and trace selection may end up generating a trace on top of an existing trace . traces can be identified since their addresses are preserved addresses in physical memory . if their participation in subsequent trace selection is not desired , then when the trace is moved into the icache , the counters associated with the trace will not be initialized to the threshold value , and instead are set to a null value . thus , the trace will not participate in profiling . however , subsequent profiling and trace selection could be used to determine whether the trace is considered \u201c good .\u201d for example , if a trace has frequent early exits , then the trace may need to be regenerated . note that more bits of branch history will allow for more accurate predictions to be made by the trace selector . however , this will require more cache space . alternatively , a multi - tiered system may be used such that the trace selector would not to select a trace when a bundle traps for the first time . instead , the trace selector may record the branch history information of the bundle in another location of memory , and then set the threshold back to a second value , which could be smaller , larger or the same as the original threshold value , and return to execution . when this bundle traps again , the trace selector can accumulate the current branch history with the branch history from the first trap to make more accurate branch predictions . although the present invention and its advantages have been described in detail , it should be understood that various changes , substitutions and alterations can be made herein without departing from the spirit and scope of the invention as defined by the appended claims ."}
|
{"patent": "fig1 a depicts the inventive instruction cache ( icache ) 100 of a processor , and includes long cache lines 101 , 102 , 103 , and 104 . only four lines are depicted for simplicity , icache 100 size is implementation dependent . each cache line includes a tag 105 , which is used to tell a cache hit or miss , a plurality of instruction bundles 106 , and counter / branch information 107 . fig1 b depicts the contents of instruction bundles 106 . each bundle is comprised of a group of instructions 108 that can be issued in the same cycle , for example , bundle 0 includes a load instruction , an add instruction , and a compare - branch instruction . note that each bundle has a fixed number of instructions , however , some of the instructions may be nops . fig1 c depicts the counter / branch information associated with each instruction bundle . each instruction bundle has counter information 109 , which is used to determine whether the code within the bundle is hot code . when the bundle is brought into the icache , the counter is initialized to a threshold value . depending on the threshold value desired , the counter can be as small as 8 to 10 bits . the counter is updated when the instruction bundle is retired from the execution pipeline . each update decrements the counter by 1 . note that the counter could initially be set to zero and increment with each retirement . however , this would require a comparison with a non - zero threshold number , e . g . 100 , which requires more work than comparing with a zero threshold number . each instruction bundle 106 in the icache 100 also maintains a branch history 110 , 111 for each instruction within the bundle . this history describes whether the comparisons in the branch instructions have resulted in a fall through to the next instruction or a branch taken to another instruction . branch history 110 is associated with bundle 0 , including slots a , b , c , which correspond to the instructions within the bundle 0 . thus , it appears one slot in the history is allocated for each instruction in the bundle , whether the instruction is a branch instruction or not . when the instructions from the original binary are brought into the icache , the branch history is cleared . the branch history information is updated when the instruction bundle is retired from fie pipeline . note that the number of instructions ( and thus the number of slots ) is by way of example only , as each bundle could have more or fewer instructions . since the third instruction in bundle 0 is a branch instruction , then slot 110 c has branch information . binary zeros indicate a fall through , and binary ones indicate a branch taken . thus , the information in 110 c , i . e . 00100 , indicates that of the last five times that this instruction has been executed , that the instruction br 1 has fallen through , fallen through , been taken , fallen through , and fallen through . note that the number of bits in the history is by way of example only , and more bits could be used to provide a more detailed history ( while requiring more space ), while fewer bits could be used to save space ( while providing less history ). note that either the most significant bit or the last significant bit may represent the most urgent execution instruction . similarly , the information in 111 b and 111 c describe the histories of instruction br 2 and br 3 respectively . note that br 2 has not recently branched , whereas the previous four executions of br 3 have resulted in the branch taken . in operation , once the counter of a bundle reaches zero , a software component known as the trace selector 201 is invoked , via a special trap , to select a trace . diagnose instructions ( special instructions to diagnose hardware ) are used by the trace selector to examine the icache and the branch history information to form a trace . regular instructions cannot read i - cache contents since i - cache is not part of the architecture states . each processor has a set of diagnose instructions defined ( not visible to application programmer ) which can be used to examine i - cache contents . fig2 a and 2b depict trace formation . fig2 a depicts instruction bundles 106 and their associated branch information 11110 . assume that the counter of bundle 1 ( not shown ) has reached zero , and that bundles 5 - 9 and 14 - 99 are not shown for reason of simplicity . note that bundles 1 - 101 may be in one or more cache lines of icache 100 . the trace selector 201 begins building the trace 202 from the hot code , in this case bundle 1 . the trace selector 201 examines the branch information ( if any ) in bundle 1 to predict whether the branch will be taken or fall through . if there are no branch instructions in the bundle , then bundle will fall through to the next sequential bundle . if the trace selector determines that the branch is most likely to fall through , then the next sequential bundle is added to the trace 202 , in this case it would be bundle 2 . note that if a branch instruction that is in the middle of bundle is assumed to be taken , the remaining instructions of the bundle are not included in the trace . the trace 202 is stored in the trace memory 203 . if the trace selector determines that the branch is most likely to be taken , then the target bundle of the branch is added to the trace 202 , in this case it would be bundle 30 . after examining the branch history 112 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 30 will not be taken , and will add the next sequential bundle , bundle 2 , to the trace 202 , and then will examine bundle 2 . after examining the branch history 113 of bundle 2 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken four times and fallen through once . therefore , the trace selector will predict that the branch to bundle 10 will be taken , and will add the target bundle , bundle 10 , to the trace 202 , and then will examine bundle 10 . after examining the branch history 114 of bundle 10 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 20 will not be taken , and will add the next sequential bundle , bundle 11 , to the trace 202 , and then will examine bundle 11 . bundle 11 does not contain any branch instructions , and therefore will not have a branch history , thus the trace selector 201 will add the next sequential bundle , bundle 12 , to the trace 202 , and then will examine bundle 12 . after examining the branch history 115 of bundle 12 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 24 will not be taken , and will add the next sequential bundle , bundle 13 , to the trace 202 , and then will examine bundle 13 . after examining the branch history 116 of bundle 13 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken four times and fallen through once . therefore , the trace selector will predict that the branch to bundle 101 will be taken , and will add the target bundle , bundle 101 , to the trace 202 , and then will examine bundle 101 . after examining the branch history 117 of bundle 101 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken five times . therefore , the trace selector will predict that the branch to bundle 1 will be taken . the trace selector notes that bundle 1 is already part of the trace 202 in trace memory 203 , via the trace of a sequence of bundles , by examining the address of a backward branch , it can be detected whether the target bundle is already part of a trace . the trace selector then ends the trace or passes the formed trace to the optimizer . the branch to bundle 1 from bundle 101 is known as a backward branch , which forms a loop . at this point , the trace may be stopped , as the trace would merely repeat bundles that are already present in the trace . the trace selector may also end the trace based on other criteria from a set of heuristics including the length of the trace , the number of conditional branches encountered , the probability of accumulated branch predictions and other considerations . thus , a trace may end when its length is a multiple of a cache line size . this would make cache operations easier , as the entire line could be loaded or overwritten without having to be concerned about starting and stopping points in the middle of a cache line . the trace could also end after a certain , predetermined number of conditional branches has been encountered . note that branch histories 113 and 116 do indicate that branch falls through occasionally , and thus the trace would be inaccurate as the trace predicts that the branch will be taken . the predetermined number could be based on the probability of error of the trace . for example , the predetermined number would be low if many of the branches have histories of 00011 or 00111 . on the other hand , the predetermined number would be high if many of the branches have histories of 00000 or 11111 . note that a trace may terminate at an indirect branch since the target address is not known . an indirect branch is different from an ip - relative ( or pc - relative ) branch in that the branch target address cannot be computed directly from the branch instruction . its target is stored either in a register or in a memory location . so the target address is unknown unless the instruction is actually executed . for example , however , the trace selector may decide to grow the trace by predicting its most recent target from the target address cache ( tac ), which is a structure commonly used to predict branch target address . for a return branch which is an indirect branch , with its target being dependent on the call site , the trace selector would know the return address if the call instruction is in the trace , if the call instruction is not in the trace , the trace selector can predict the call site using the top address of the return stack buffer ( rsb ), which is a commonly used data structure to predict return branches . the tac and the rsb are discussed in the co - pending and commonly assigned u . s . patent application entitled efficient mapping to optimized code for processor embedded run - time optimizer ser . no . 09 / 252 , 367 , filed feb . 18 , 1999 , and issued feb . 6 , 2001 , as u . s . pat . no . 6 , 185 , 669 ; which is hereby incorporated by reference . the trace 202 will be stored in the trace memory 203 . there is a mapping from the trace starting instruction bundle in the original binary to the trace in the trace memory . when the trace starting bundle is executed , the mapping will automatically lead the execution to the trace stored in the trace memory 203 . typically , an executed branch instruction has its target in the trace memory . this is discussed in the co - pending and commonly assigned u . s . patent application entitled system and method using a hardware embedded run - time optimizer ser . no . 09 / 252 , 170 , filed feb . 18 , 1999 , and issued sep . 17 , 2002 , as u . s . pat . no . 6 , 453 , 411 , which is hereby incorporated by reference . note that the trace may require more than one cache line . as stated previously , long cache lines are inefficient for original binary . this is because the original binary is loaded sequentially , i . e . bundle 1 , 2 , 3 , 4 , 5 , 6 , etc ., and branches taken within the bundles may result in many of the loaded bundles not being used . for example , suppose bundle 6 has a branch taken to bundle 50 , then the loading of bundles 7 - 49 represent wasted time and cache space as they are not going to be used . however , when the trace is loaded into the cache , the entire trace is almost certain to be used . thus , the long cache lines are much more efficient , because of the sequential locality , as the bundles of the trace will ( almost always ) fall through to the next bundle of the trace . note that a trace usually spans several cache lines . it may not end at the end of a cache line . in this case , the remaining part of the cache line can be the start of another trace . note that since traces are also brought into the icache , the profiling and trace selection may end up generating a trace on top of an existing trace . traces can be identified since their addresses are preserved addresses in physical memory . if their participation in subsequent trace selection is not desired , then when the trace is moved into the icache , the counters associated with the trace will not be initialized to the threshold value , and instead are set to a null value . thus , the trace will not participate in profiling . however , subsequent profiling and trace selection could be used to determine whether the trace is considered \u201c good .\u201d for example , if a trace has frequent early exits , then the trace may need to be regenerated . note that more bits of branch history will allow for more accurate predictions to be made by the trace selector . however , this will require more cache space . alternatively , a multi - tiered system may be used such that the trace selector would not to select a trace when a bundle traps for the first time . instead , the trace selector may record the branch history information of the bundle in another location of memory , and then set the threshold back to a second value , which could be smaller , larger or the same as the original threshold value , and return to execution . when this bundle traps again , the trace selector can accumulate the current branch history with the branch history from the first trap to make more accurate branch predictions . although the present invention and its advantages have been described in detail , it should be understood that various changes , substitutions and alterations can be made herein without departing from the spirit and scope of the invention as defined by the appended claims .", "category": "Electricity"}
|
Is the categorization of this patent accurate?
| 0.25 |
fdcc34483564ad9019b21f93490188a72a8ae33eb03472de41376137951476c1
| 0.033691 | 0.000203 | 0.033691 | 0.005219 | 0.035156 | 0.012451 |
null |
{"category": "Physics", "patent": "fig1 a depicts the inventive instruction cache ( icache ) 100 of a processor , and includes long cache lines 101 , 102 , 103 , and 104 . only four lines are depicted for simplicity , icache 100 size is implementation dependent . each cache line includes a tag 105 , which is used to tell a cache hit or miss , a plurality of instruction bundles 106 , and counter / branch information 107 . fig1 b depicts the contents of instruction bundles 106 . each bundle is comprised of a group of instructions 108 that can be issued in the same cycle , for example , bundle 0 includes a load instruction , an add instruction , and a compare - branch instruction . note that each bundle has a fixed number of instructions , however , some of the instructions may be nops . fig1 c depicts the counter / branch information associated with each instruction bundle . each instruction bundle has counter information 109 , which is used to determine whether the code within the bundle is hot code . when the bundle is brought into the icache , the counter is initialized to a threshold value . depending on the threshold value desired , the counter can be as small as 8 to 10 bits . the counter is updated when the instruction bundle is retired from the execution pipeline . each update decrements the counter by 1 . note that the counter could initially be set to zero and increment with each retirement . however , this would require a comparison with a non - zero threshold number , e . g . 100 , which requires more work than comparing with a zero threshold number . each instruction bundle 106 in the icache 100 also maintains a branch history 110 , 111 for each instruction within the bundle . this history describes whether the comparisons in the branch instructions have resulted in a fall through to the next instruction or a branch taken to another instruction . branch history 110 is associated with bundle 0 , including slots a , b , c , which correspond to the instructions within the bundle 0 . thus , it appears one slot in the history is allocated for each instruction in the bundle , whether the instruction is a branch instruction or not . when the instructions from the original binary are brought into the icache , the branch history is cleared . the branch history information is updated when the instruction bundle is retired from fie pipeline . note that the number of instructions ( and thus the number of slots ) is by way of example only , as each bundle could have more or fewer instructions . since the third instruction in bundle 0 is a branch instruction , then slot 110 c has branch information . binary zeros indicate a fall through , and binary ones indicate a branch taken . thus , the information in 110 c , i . e . 00100 , indicates that of the last five times that this instruction has been executed , that the instruction br 1 has fallen through , fallen through , been taken , fallen through , and fallen through . note that the number of bits in the history is by way of example only , and more bits could be used to provide a more detailed history ( while requiring more space ), while fewer bits could be used to save space ( while providing less history ). note that either the most significant bit or the last significant bit may represent the most urgent execution instruction . similarly , the information in 111 b and 111 c describe the histories of instruction br 2 and br 3 respectively . note that br 2 has not recently branched , whereas the previous four executions of br 3 have resulted in the branch taken . in operation , once the counter of a bundle reaches zero , a software component known as the trace selector 201 is invoked , via a special trap , to select a trace . diagnose instructions ( special instructions to diagnose hardware ) are used by the trace selector to examine the icache and the branch history information to form a trace . regular instructions cannot read i - cache contents since i - cache is not part of the architecture states . each processor has a set of diagnose instructions defined ( not visible to application programmer ) which can be used to examine i - cache contents . fig2 a and 2b depict trace formation . fig2 a depicts instruction bundles 106 and their associated branch information 11110 . assume that the counter of bundle 1 ( not shown ) has reached zero , and that bundles 5 - 9 and 14 - 99 are not shown for reason of simplicity . note that bundles 1 - 101 may be in one or more cache lines of icache 100 . the trace selector 201 begins building the trace 202 from the hot code , in this case bundle 1 . the trace selector 201 examines the branch information ( if any ) in bundle 1 to predict whether the branch will be taken or fall through . if there are no branch instructions in the bundle , then bundle will fall through to the next sequential bundle . if the trace selector determines that the branch is most likely to fall through , then the next sequential bundle is added to the trace 202 , in this case it would be bundle 2 . note that if a branch instruction that is in the middle of bundle is assumed to be taken , the remaining instructions of the bundle are not included in the trace . the trace 202 is stored in the trace memory 203 . if the trace selector determines that the branch is most likely to be taken , then the target bundle of the branch is added to the trace 202 , in this case it would be bundle 30 . after examining the branch history 112 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 30 will not be taken , and will add the next sequential bundle , bundle 2 , to the trace 202 , and then will examine bundle 2 . after examining the branch history 113 of bundle 2 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken four times and fallen through once . therefore , the trace selector will predict that the branch to bundle 10 will be taken , and will add the target bundle , bundle 10 , to the trace 202 , and then will examine bundle 10 . after examining the branch history 114 of bundle 10 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 20 will not be taken , and will add the next sequential bundle , bundle 11 , to the trace 202 , and then will examine bundle 11 . bundle 11 does not contain any branch instructions , and therefore will not have a branch history , thus the trace selector 201 will add the next sequential bundle , bundle 12 , to the trace 202 , and then will examine bundle 12 . after examining the branch history 115 of bundle 12 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 24 will not be taken , and will add the next sequential bundle , bundle 13 , to the trace 202 , and then will examine bundle 13 . after examining the branch history 116 of bundle 13 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken four times and fallen through once . therefore , the trace selector will predict that the branch to bundle 101 will be taken , and will add the target bundle , bundle 101 , to the trace 202 , and then will examine bundle 101 . after examining the branch history 117 of bundle 101 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken five times . therefore , the trace selector will predict that the branch to bundle 1 will be taken . the trace selector notes that bundle 1 is already part of the trace 202 in trace memory 203 , via the trace of a sequence of bundles , by examining the address of a backward branch , it can be detected whether the target bundle is already part of a trace . the trace selector then ends the trace or passes the formed trace to the optimizer . the branch to bundle 1 from bundle 101 is known as a backward branch , which forms a loop . at this point , the trace may be stopped , as the trace would merely repeat bundles that are already present in the trace . the trace selector may also end the trace based on other criteria from a set of heuristics including the length of the trace , the number of conditional branches encountered , the probability of accumulated branch predictions and other considerations . thus , a trace may end when its length is a multiple of a cache line size . this would make cache operations easier , as the entire line could be loaded or overwritten without having to be concerned about starting and stopping points in the middle of a cache line . the trace could also end after a certain , predetermined number of conditional branches has been encountered . note that branch histories 113 and 116 do indicate that branch falls through occasionally , and thus the trace would be inaccurate as the trace predicts that the branch will be taken . the predetermined number could be based on the probability of error of the trace . for example , the predetermined number would be low if many of the branches have histories of 00011 or 00111 . on the other hand , the predetermined number would be high if many of the branches have histories of 00000 or 11111 . note that a trace may terminate at an indirect branch since the target address is not known . an indirect branch is different from an ip - relative ( or pc - relative ) branch in that the branch target address cannot be computed directly from the branch instruction . its target is stored either in a register or in a memory location . so the target address is unknown unless the instruction is actually executed . for example , however , the trace selector may decide to grow the trace by predicting its most recent target from the target address cache ( tac ), which is a structure commonly used to predict branch target address . for a return branch which is an indirect branch , with its target being dependent on the call site , the trace selector would know the return address if the call instruction is in the trace , if the call instruction is not in the trace , the trace selector can predict the call site using the top address of the return stack buffer ( rsb ), which is a commonly used data structure to predict return branches . the tac and the rsb are discussed in the co - pending and commonly assigned u . s . patent application entitled efficient mapping to optimized code for processor embedded run - time optimizer ser . no . 09 / 252 , 367 , filed feb . 18 , 1999 , and issued feb . 6 , 2001 , as u . s . pat . no . 6 , 185 , 669 ; which is hereby incorporated by reference . the trace 202 will be stored in the trace memory 203 . there is a mapping from the trace starting instruction bundle in the original binary to the trace in the trace memory . when the trace starting bundle is executed , the mapping will automatically lead the execution to the trace stored in the trace memory 203 . typically , an executed branch instruction has its target in the trace memory . this is discussed in the co - pending and commonly assigned u . s . patent application entitled system and method using a hardware embedded run - time optimizer ser . no . 09 / 252 , 170 , filed feb . 18 , 1999 , and issued sep . 17 , 2002 , as u . s . pat . no . 6 , 453 , 411 , which is hereby incorporated by reference . note that the trace may require more than one cache line . as stated previously , long cache lines are inefficient for original binary . this is because the original binary is loaded sequentially , i . e . bundle 1 , 2 , 3 , 4 , 5 , 6 , etc ., and branches taken within the bundles may result in many of the loaded bundles not being used . for example , suppose bundle 6 has a branch taken to bundle 50 , then the loading of bundles 7 - 49 represent wasted time and cache space as they are not going to be used . however , when the trace is loaded into the cache , the entire trace is almost certain to be used . thus , the long cache lines are much more efficient , because of the sequential locality , as the bundles of the trace will ( almost always ) fall through to the next bundle of the trace . note that a trace usually spans several cache lines . it may not end at the end of a cache line . in this case , the remaining part of the cache line can be the start of another trace . note that since traces are also brought into the icache , the profiling and trace selection may end up generating a trace on top of an existing trace . traces can be identified since their addresses are preserved addresses in physical memory . if their participation in subsequent trace selection is not desired , then when the trace is moved into the icache , the counters associated with the trace will not be initialized to the threshold value , and instead are set to a null value . thus , the trace will not participate in profiling . however , subsequent profiling and trace selection could be used to determine whether the trace is considered \u201c good .\u201d for example , if a trace has frequent early exits , then the trace may need to be regenerated . note that more bits of branch history will allow for more accurate predictions to be made by the trace selector . however , this will require more cache space . alternatively , a multi - tiered system may be used such that the trace selector would not to select a trace when a bundle traps for the first time . instead , the trace selector may record the branch history information of the bundle in another location of memory , and then set the threshold back to a second value , which could be smaller , larger or the same as the original threshold value , and return to execution . when this bundle traps again , the trace selector can accumulate the current branch history with the branch history from the first trap to make more accurate branch predictions . although the present invention and its advantages have been described in detail , it should be understood that various changes , substitutions and alterations can be made herein without departing from the spirit and scope of the invention as defined by the appended claims ."}
|
{"category": "General tagging of new or cross-sectional technology", "patent": "fig1 a depicts the inventive instruction cache ( icache ) 100 of a processor , and includes long cache lines 101 , 102 , 103 , and 104 . only four lines are depicted for simplicity , icache 100 size is implementation dependent . each cache line includes a tag 105 , which is used to tell a cache hit or miss , a plurality of instruction bundles 106 , and counter / branch information 107 . fig1 b depicts the contents of instruction bundles 106 . each bundle is comprised of a group of instructions 108 that can be issued in the same cycle , for example , bundle 0 includes a load instruction , an add instruction , and a compare - branch instruction . note that each bundle has a fixed number of instructions , however , some of the instructions may be nops . fig1 c depicts the counter / branch information associated with each instruction bundle . each instruction bundle has counter information 109 , which is used to determine whether the code within the bundle is hot code . when the bundle is brought into the icache , the counter is initialized to a threshold value . depending on the threshold value desired , the counter can be as small as 8 to 10 bits . the counter is updated when the instruction bundle is retired from the execution pipeline . each update decrements the counter by 1 . note that the counter could initially be set to zero and increment with each retirement . however , this would require a comparison with a non - zero threshold number , e . g . 100 , which requires more work than comparing with a zero threshold number . each instruction bundle 106 in the icache 100 also maintains a branch history 110 , 111 for each instruction within the bundle . this history describes whether the comparisons in the branch instructions have resulted in a fall through to the next instruction or a branch taken to another instruction . branch history 110 is associated with bundle 0 , including slots a , b , c , which correspond to the instructions within the bundle 0 . thus , it appears one slot in the history is allocated for each instruction in the bundle , whether the instruction is a branch instruction or not . when the instructions from the original binary are brought into the icache , the branch history is cleared . the branch history information is updated when the instruction bundle is retired from fie pipeline . note that the number of instructions ( and thus the number of slots ) is by way of example only , as each bundle could have more or fewer instructions . since the third instruction in bundle 0 is a branch instruction , then slot 110 c has branch information . binary zeros indicate a fall through , and binary ones indicate a branch taken . thus , the information in 110 c , i . e . 00100 , indicates that of the last five times that this instruction has been executed , that the instruction br 1 has fallen through , fallen through , been taken , fallen through , and fallen through . note that the number of bits in the history is by way of example only , and more bits could be used to provide a more detailed history ( while requiring more space ), while fewer bits could be used to save space ( while providing less history ). note that either the most significant bit or the last significant bit may represent the most urgent execution instruction . similarly , the information in 111 b and 111 c describe the histories of instruction br 2 and br 3 respectively . note that br 2 has not recently branched , whereas the previous four executions of br 3 have resulted in the branch taken . in operation , once the counter of a bundle reaches zero , a software component known as the trace selector 201 is invoked , via a special trap , to select a trace . diagnose instructions ( special instructions to diagnose hardware ) are used by the trace selector to examine the icache and the branch history information to form a trace . regular instructions cannot read i - cache contents since i - cache is not part of the architecture states . each processor has a set of diagnose instructions defined ( not visible to application programmer ) which can be used to examine i - cache contents . fig2 a and 2b depict trace formation . fig2 a depicts instruction bundles 106 and their associated branch information 11110 . assume that the counter of bundle 1 ( not shown ) has reached zero , and that bundles 5 - 9 and 14 - 99 are not shown for reason of simplicity . note that bundles 1 - 101 may be in one or more cache lines of icache 100 . the trace selector 201 begins building the trace 202 from the hot code , in this case bundle 1 . the trace selector 201 examines the branch information ( if any ) in bundle 1 to predict whether the branch will be taken or fall through . if there are no branch instructions in the bundle , then bundle will fall through to the next sequential bundle . if the trace selector determines that the branch is most likely to fall through , then the next sequential bundle is added to the trace 202 , in this case it would be bundle 2 . note that if a branch instruction that is in the middle of bundle is assumed to be taken , the remaining instructions of the bundle are not included in the trace . the trace 202 is stored in the trace memory 203 . if the trace selector determines that the branch is most likely to be taken , then the target bundle of the branch is added to the trace 202 , in this case it would be bundle 30 . after examining the branch history 112 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 30 will not be taken , and will add the next sequential bundle , bundle 2 , to the trace 202 , and then will examine bundle 2 . after examining the branch history 113 of bundle 2 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken four times and fallen through once . therefore , the trace selector will predict that the branch to bundle 10 will be taken , and will add the target bundle , bundle 10 , to the trace 202 , and then will examine bundle 10 . after examining the branch history 114 of bundle 10 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 20 will not be taken , and will add the next sequential bundle , bundle 11 , to the trace 202 , and then will examine bundle 11 . bundle 11 does not contain any branch instructions , and therefore will not have a branch history , thus the trace selector 201 will add the next sequential bundle , bundle 12 , to the trace 202 , and then will examine bundle 12 . after examining the branch history 115 of bundle 12 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has not been taken and has fallen through . therefore , the trace selector will predict that the branch to bundle 24 will not be taken , and will add the next sequential bundle , bundle 13 , to the trace 202 , and then will examine bundle 13 . after examining the branch history 116 of bundle 13 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken four times and fallen through once . therefore , the trace selector will predict that the branch to bundle 101 will be taken , and will add the target bundle , bundle 101 , to the trace 202 , and then will examine bundle 101 . after examining the branch history 117 of bundle 101 , the trace selector 201 will note that in the previous five executions of the branch instruction , the branch has been taken five times . therefore , the trace selector will predict that the branch to bundle 1 will be taken . the trace selector notes that bundle 1 is already part of the trace 202 in trace memory 203 , via the trace of a sequence of bundles , by examining the address of a backward branch , it can be detected whether the target bundle is already part of a trace . the trace selector then ends the trace or passes the formed trace to the optimizer . the branch to bundle 1 from bundle 101 is known as a backward branch , which forms a loop . at this point , the trace may be stopped , as the trace would merely repeat bundles that are already present in the trace . the trace selector may also end the trace based on other criteria from a set of heuristics including the length of the trace , the number of conditional branches encountered , the probability of accumulated branch predictions and other considerations . thus , a trace may end when its length is a multiple of a cache line size . this would make cache operations easier , as the entire line could be loaded or overwritten without having to be concerned about starting and stopping points in the middle of a cache line . the trace could also end after a certain , predetermined number of conditional branches has been encountered . note that branch histories 113 and 116 do indicate that branch falls through occasionally , and thus the trace would be inaccurate as the trace predicts that the branch will be taken . the predetermined number could be based on the probability of error of the trace . for example , the predetermined number would be low if many of the branches have histories of 00011 or 00111 . on the other hand , the predetermined number would be high if many of the branches have histories of 00000 or 11111 . note that a trace may terminate at an indirect branch since the target address is not known . an indirect branch is different from an ip - relative ( or pc - relative ) branch in that the branch target address cannot be computed directly from the branch instruction . its target is stored either in a register or in a memory location . so the target address is unknown unless the instruction is actually executed . for example , however , the trace selector may decide to grow the trace by predicting its most recent target from the target address cache ( tac ), which is a structure commonly used to predict branch target address . for a return branch which is an indirect branch , with its target being dependent on the call site , the trace selector would know the return address if the call instruction is in the trace , if the call instruction is not in the trace , the trace selector can predict the call site using the top address of the return stack buffer ( rsb ), which is a commonly used data structure to predict return branches . the tac and the rsb are discussed in the co - pending and commonly assigned u . s . patent application entitled efficient mapping to optimized code for processor embedded run - time optimizer ser . no . 09 / 252 , 367 , filed feb . 18 , 1999 , and issued feb . 6 , 2001 , as u . s . pat . no . 6 , 185 , 669 ; which is hereby incorporated by reference . the trace 202 will be stored in the trace memory 203 . there is a mapping from the trace starting instruction bundle in the original binary to the trace in the trace memory . when the trace starting bundle is executed , the mapping will automatically lead the execution to the trace stored in the trace memory 203 . typically , an executed branch instruction has its target in the trace memory . this is discussed in the co - pending and commonly assigned u . s . patent application entitled system and method using a hardware embedded run - time optimizer ser . no . 09 / 252 , 170 , filed feb . 18 , 1999 , and issued sep . 17 , 2002 , as u . s . pat . no . 6 , 453 , 411 , which is hereby incorporated by reference . note that the trace may require more than one cache line . as stated previously , long cache lines are inefficient for original binary . this is because the original binary is loaded sequentially , i . e . bundle 1 , 2 , 3 , 4 , 5 , 6 , etc ., and branches taken within the bundles may result in many of the loaded bundles not being used . for example , suppose bundle 6 has a branch taken to bundle 50 , then the loading of bundles 7 - 49 represent wasted time and cache space as they are not going to be used . however , when the trace is loaded into the cache , the entire trace is almost certain to be used . thus , the long cache lines are much more efficient , because of the sequential locality , as the bundles of the trace will ( almost always ) fall through to the next bundle of the trace . note that a trace usually spans several cache lines . it may not end at the end of a cache line . in this case , the remaining part of the cache line can be the start of another trace . note that since traces are also brought into the icache , the profiling and trace selection may end up generating a trace on top of an existing trace . traces can be identified since their addresses are preserved addresses in physical memory . if their participation in subsequent trace selection is not desired , then when the trace is moved into the icache , the counters associated with the trace will not be initialized to the threshold value , and instead are set to a null value . thus , the trace will not participate in profiling . however , subsequent profiling and trace selection could be used to determine whether the trace is considered \u201c good .\u201d for example , if a trace has frequent early exits , then the trace may need to be regenerated . note that more bits of branch history will allow for more accurate predictions to be made by the trace selector . however , this will require more cache space . alternatively , a multi - tiered system may be used such that the trace selector would not to select a trace when a bundle traps for the first time . instead , the trace selector may record the branch history information of the bundle in another location of memory , and then set the threshold back to a second value , which could be smaller , larger or the same as the original threshold value , and return to execution . when this bundle traps again , the trace selector can accumulate the current branch history with the branch history from the first trap to make more accurate branch predictions . although the present invention and its advantages have been described in detail , it should be understood that various changes , substitutions and alterations can be made herein without departing from the spirit and scope of the invention as defined by the appended claims ."}
|
Is the patent correctly categorized?
| 0.25 |
fdcc34483564ad9019b21f93490188a72a8ae33eb03472de41376137951476c1
| 0.078125 | 0.289063 | 0.084961 | 0.351563 | 0.279297 | 0.511719 |
null |
{"category": "Human Necessities", "patent": "the present invention is directed towards improving the quality and functionality of both implantable cardioverter defibrillators ( icds ) and implantable atrial defibrillators ( iads ). all such devices need to have a method and mechanism for accurately distinguishing sinus tachycardia ( st ) form non - st , such as atrial fibrillation and atrial flutter . this method and mechanism should be conservative , biased towards not initiating therapy unless non - st is clearly indicated . it should also require as few machine computational cycles as possible to reduce icd and iad battery drain . the principle underlying the present invention is that st and non - st can be distinguished by studying and measuring the atrial electrogram morphology to detect aberrant conduction in the atria . thus , by comparing normal st waveforms to a series of newly - occurring waveforms , st is indicated by abnormal atrial waveform morphology which signals aberrant conduction within the atria improper electrical patterns of depolarization . this will show up in the signal potentials captured by a bipolar atrial lead . with reference to fig1 the present invention can be implemented in an icd 100 ( which could be an iad ) that is implanted in the chest of a patient who suffers from unpleasant and possibly life - threatening irregularities in heart operation due to improper electrical stimulation of the heart muscle cells . during normal heart action , the heart &# 39 ; s electrical impulses originate in the sino - atrial node as an action potential that is transmitted smoothly to all portions of the atria , causing contraction of the atrial chambers . the electrical impulse continues in its path to a cluster of conduction fibrils known as the atrioventricular node . after a delay of about one - tenth of a second , an action potential flows over the ventricles and causes them to contract in synchronism with and following shortly after the atrial contraction . in this manner , the heart pumps blood to the lungs and from the lungs to the body . a bipolar lead 102 is implanted within the atria 104 of a heart 103 to measure the electrical potential between nearby cells of the atria . as the action potential actively passes from cell to cell past the two electrodes of the lead 102 , an oscillating signal potential is developed across the two electrodes and is conveyed by the bipolar lead 102 to the icd 100 where the signal is sampled about 1000 times per second and is digitized , with the sequential samples stored within a ram memory within the icd 100 for further analysis . another bipolar lead 105 may extend to the ventricle ( not shown ) of the heart 103 to measure the action potentials generated within the ventricles . these same leads , or other leads , may be used by the icd for administering various types of therapy , such as pacing or defibrillation shock therapy . the data samples collected from within the atria are now analyzed . first , with reference to data gathered from the ventricles , if some event in the ventricle , such as a mis - timed depolarization event that partially overlaps the atria &# 39 ; s p wave , could leak across to the atria and distort the measurement of the potentials for a given atrial depolarization , then the data for that particular heartbeat is discarded and is not used in further analyses . such distortion can also be found simply because a given heartbeat set of data gathered from the atria is itself badly distorted , and this approach must be taken in the case of an iad having no ventricular lead . the remaining data is retained for further processing . in preparation for further processing , the data for individual heartbeats is identified and gathered . the position within the gathered data of the negative spike to the atrial p depolarization waveform is determined and is selected as the center of the waveform for each beat . then , since data close to this p - wave is to be analyzed , and since data further away may be distorted to some degree by ventricular activity or by variations in the p - to - p interval , 32 sequential data samples are selected for each heartbeat such that the negative spike of the p - wave becomes signal potential sample number 16 of the 32 sequential samples selected for further analyses . these samples for a normal heartbeat appear at 302 in fig3 and at 502 in fig5 where the amplitude of the atrial signal potential is plotted against time over the 32 sampled values , with the negative p - wave spike positioned as signal sample number 16 . likewise , the samples for an abnormal heartbeat appear at 504 in fig5 . straightforward correlation ( cwa ) of the 32 samples against a template containing a typical or normal pattern could be performed at this juncture , as taught on page 563 ( equation ( 1 )) of the article by thorne , cited in the introductory portion of this specification . but the performance of such a cross correlation at repeated regular intervals would generate numerous computations and would drain the icd &# 39 ; s battery more rapidly than is desirable . in addition , provision would have to be made for a full template waveform that can be stored within the icd to be used as a comparison reference . in addition , such templates generally need to be patient specific , and they are more susceptible to normal changes in shape . accordingly , to reduce the mathematical complexity of the cross correlation computations , it is desirable to reduce the number of data points from 32 down to some much lower number through the use of some form of transformation into a different set of values . for example , the data could be transformed using the karhunen - loeve transformation described in the morris article cited at the beginning of this specification . but that transformation is fairly computationally intensive . computationally , a fast fourier transformation would be more efficient , but any such transformation into the frequency domain , where variable frequency sine and cosine wave amplitudes result from the transformation , does not match itself well to the morphology of atrial heart waveforms , which tend to be formed from large , ringing , spike - like representations of p - wave depolarization events travelling as moving action potential fronts past the pair of electrodes attached to the bipolar lead , and not as steady harmonics extending across time . accordingly , fourier - class transforms do not adequately reduce the number of significant data values that must be considered . and in addition , when the \u201c window \u201d size for a fourier analysis is selected , if it is wide , then the frequency values are not time - specific , but represent average signal harmonic content over the entire window time duration . and if the window is narrow , then the harmonics are too spread out in frequency , and the frequency - domain data generated by the transform does not resolve things finely enough in the frequency domain . for all of the above reasons , the present invention analyzes the waveforms using a discrete wavelet transform . this has the advantage that while low frequency wavelets are broad in time , high frequency wavelets have a much narrower timeslot focus . thus , for example , with 32 data samples taken sequentially over time , a wavelet transformation generates a d . c . wavelet that spans the entire set of 32 points ; a first reversing wavelet that also spans the entire set of points ; two second harmonic wavelets that each focus upon only one - half of the 32 data points ; four third harmonic wavelets that each focus upon only one - fourth of the 32 data points ; eight fourth harmonic wavelets that each focus upon four of the 32 data points ; and 16 fifth harmonic wavelets that each focuses upon only two of the 32 data points . accordingly , the higher - frequency wavelet amplitudes are quite time specific , unlike fourier sinusoidal amplitudes , while the lower - frequency and d . c . wavelet amplitudes give one broad information about the whole set of 32 points . and like the fourier transform , the discrete wavelet transform preserves all of the information of the original atrial signal , such that the transformation is fully reversible . 32 discrete wavelet transform values may be reverse transformed back into 32 values indicating the atrial signal potentials at 32 sequential points in time . in essence , the haar function is performing multiple digital filtering operations , at variable positions and utilizing varying - width windowing functions , to develop component values that represent varying types of heartbeat activity , at varying frequencies , spread over varying widths , and located at varying positions . these component values can be studied to see which subset of these component values are good at distinguishing st from non - st , or ( more generally ) which subset of these component values are good at distinguishing one type of heartbeat waveform from another . further study of such a selected subset can further determine which of the subset of components are relatively invariant from one individual &# 39 ; s heart to another . one preferably selects component values that are suitable from both of these two perspectives . the particular wavelet transformation to choose can be tailored to the nature of the data , with the wavelets chosen to resemble somewhat the values to be found within the data so that some transformed values are of much larger amplitude than others , and such that the low amplitude transformed values may be disregarded . of particular importance with an atrial p waveform of the type being analyzed here are wavelets having frequency values roughly comparable to and timed to coincide with the ringing of the atrial depolarization . this tailoring can minimize the number of transformed data values that are significant and that are candidates for full participation in the correlation analyses to determine if there has been a substantial change in waveform morphology . another factor in selecting a particular wavelet transformation is reducing the number of computations that must be performed to carry out the transformation . yet another factor is selecting a wavelet transformation where some of the transformed component values vary from normal to abnormal heart rhythms in much the same over a population of individuals so that a generic template of these component values may be derived that can be used with many individuals , rather than with just one individual . these factors suggest that the haar transformation would be a suitable candidate for use in analysis of atrial waveforms and possibly other waveforms as well . with reference to fig3 in the case of 32 voltage values sampled over time , the 32 applicable haar discrete transform wavelets appear as shown in this figure . other discrete wavelet transforms based upon wavelets having triangular or other shapes may also prove usable . the haar transform wavelets , in particular , have the shape of a square waveform , as will be described , and this can simplify the computations required . in fig3 time increases from left to right . a scale 304 indicates the 32 points in time at which the atrial waveform is sampled , and an analog atrial waveform ( before digitization ) is shown at 302 . the p wave negative notch 303 is shown centered at the 16 th timeslot so that the signal potential of the atria waveform sampled at this point in time becomes the 16 th data value in the set of 32 data values that are to be subjected to the haar transformation . a haar wavelet is simply one cycle of a square waveform that starts at zero , then swings positive ( to \u201c+ 1 \u201d), then swings negative ( to \u201c\u2212 1 \u201d), and then swings back to zero . the wavelet w 2 , shown at 308 in fig3 for example , is at \u201c+ 1 \u201d at points in time 1 to 16 and is at \u201c\u2212 1 \u201d at points in time 17 to 32 . the shorter waveform w 6 , shown at 316 in fig3 is at \u201c+ 1 \u201d at points in time 9 to 12 and is at \u201c\u2212 1 \u201d at points in time 13 to 16 , and is at \u201c 0 \u201d at all other points in time . the waveform w 1 , shown at 306 in fig3 is so slow to fluctuate in time that its \u201c\u2212 1 \u201d portion is off of the chart ( in fig3 ) to the right , and is ignored ; and accordingly , it has the value \u201c+ 1 \u201d for all 32 of the points in time shown in fig3 . it thus represents the average , or d . c ., component of the atrial signal potential . the lowest frequency wavelets w 1 and w 2 encompass the entire set of 32 points in time . the higher frequency wavelets w 3 and w 4 each encompasses only half of the points in time , and the two wavelets taken together encompass all the points in time . the four wavelets w 5 , w 6 , w 7 , and w 8 each encompasses only eight points in time , and the four wavelets taken together encompass all points in time . the eight wavelets w 9 . . . w 16 each encompasses only four points in time , and together all eight wavelets encompass all points in time . and finally , the wavelets w 17 . . . w 32 each encompass only two points in time , but the sixteen wavelets together encompass all points in time . thus , each wavelet has a characteristic time span and position as well as a characteristic frequency , with higher - frequency wavelets having a narrower and more specific time span and position than lower - frequency wavelets . this is advantageous with an impulse - type , ringing signal such as the atrial depolarization waveform considered here , since many higher - frequency transform values that correspond to haar wavelets positioned in time away from the p waveform or that do not correspond to its frequency of ringing may be low in amplitude such that they may safely be ignored during the analysis steps , thereby reducing the data that must be processed during the correlation steps . ( the above , while presented in the context of the haar wavelet , is also applicable to other wavelet shapes that may used to perform a discrete wavelet transform ( dwt )). each of the 32 haar transformed values is computed as follows : multiply each of the 32 sampled and digitized voltage values representing the analog atrial waveform 302 by the correspondingly - positioned -( in - time ) amplitude values of one of the 32 haar wavelets ( shown in fig3 ); then sum the resulting products ; and then multiply the resulting sums by a discrete wavelet transform scaling factor 2 \u2212 j / 2 , where j equals 1 for the wavelets w 17 to w 32 , 2 for w 9 to w 16 , 3 for w 5 to w 8 , 4 for w 3 to w 4 , and 5 for w 1 and w 2 . for example , and referring to fig3 : the first wavelet w 1 is always + 1 , so the atrial signal potential values are simply summed and then multiplied by 2 \u2212 5 / 2 ; the second wavelet is + 1 for time points 1 to 16 and \u2212 1 for time points 17 to 32 , so the first 16 values are summed , the second 16 values are summed , the difference between the first and second sums is computed , and the result is multiplied by 2 \u2212 5 / 2 ; and so on until all the transformed values have been computed , one for each haar wavelet . the 32 time - sequential atrial signal potential values are thus transformed into 32 haar wavelet amplitude values which may be called \u201c transformed values \u201d and which may be assigned the numbers 1 to 32 corresponding to the subscript numbers of the haar wavelets to which they each correspond and whose amplitudes they represent . the above description of how to compute the haar wavelet transformed values is accurate , but it is not the most efficient way to proceed in the case of haar wavelets . like the fourier transform , which has a corresponding fast fourier transform that saves intermediate results and thereby avoids re - computing them and thus reduces substantially the number of computations , the haar transformation also may be carried out in a manner that saves intermediate sums and re - uses them to reduce substantially the number of computations . this is described below at the point where fig4 is described , since fig4 illustrates graphically how this can be done . the illustrative program listing presented below is also an implementation of the fast haar wavelet transformation algorithm . the haar transformation can readily be carried out by any digital computer . as an example of how the haar transformation can be carried out on 32 values , the following program , written in the language c , is illustrative of many possible programs that may be written . in the exemplary program that follows , a 32 - element array yy contains 32 data values representing the 32 sampled atrial signal potential values representing the fluctuations over time of the signal supplied by the bipolar lead 102 . the analog atrial signal is sampled , digitized , broken into separate beat data sets , pre - processed ( to remove waveforms distorted by ventricular activity ), centered ( with the negative depolarization spike at data point 16 ), and fed into the subroutine presented below . this subroutine is compiled ( or assembled , if rewritten in assembly language ) and installed in the rom of the icd &# 39 ; s embedded microprocessor . this subroutine returns , contained within the same array yy , the 32 transformed haar wavelet amplitude values described above . ( the constant value \u201c recipsqrttwo \u201d is the reciprocal of the square root of two ). an illustrative version of the subroutine for computing all 32 of the haar transformed values in an efficient manner is presented here : void haar ( double yy []) { int i , j , l ; double zz [ 32 ]; for ( i = 5 ; i & gt ; 0 ; i \u2212\u2212) { l = 1 ; for ( j = 1 ; j & lt ;= i ; ++ j ) l = 2 * l ; for ( j = 0 ; j & lt ; l ; ++ j ) zz [ j ] = yy [ j ]; for ( j = 0 ; j & lt ; l \u2212 1 ; j = j + 2 ) ( yy [ j / 2 ] = recipsqrttwo * ( zz [ j ] + zz [ j + 1 ]); yy [( j + l )/ 2 ] = rcipsqrttwo * ( zz [ j ] \u2212 zz [ j + 1 ]); } } return ; } the above program computes the 32 haar transformed values 1 to 32 from the 32 atrial signal potential time - sequenced input values . it does so with only 31 additions , 31 subtractions , and 62 multiplications . it is carefully designed to compute each value in a systematic way , making multiple use of intermediate results . first , the program computes sixteen sums of and sixteen differences between eight adjacent pairs of the 16 time sequenced atrial signal potential values , and the sixteen difference values become the haar transformed values 17 to 32 , reflecting the strength of the highest frequency haar wavelets positioned at 16 different positions in time , corresponding to the wavelets w 17 to w 32 shown in fig3 . all the sum and difference values are scaled by multiplication by 2 \u2212 1 / 2 . next , taking the 16 sums of atrial signal potential values as an intermediate result , the above algorithm generates eight sums of and eight differences between adjacent pairs of these sixteen intermediate values that resulted from the first sixteen additions , and the eight newly - computed difference values become the haar transformed values 9 to 16 , reflecting the strength of the second to the highest frequency values at eight different points in time , corresponding to wavelets w 9 to w 16 in fig3 . all of these eight sum and eight difference values are again scaled by 2 \u2212 1 / 2 so that the wavelets w 9 to w 16 are scaled by \u00bd ( 2 \u2212 2 / 2 or 2 \u2212 1 / 2 times 2 \u2212 1 / 2 ). next , taking these eight sums of sums of adjacent atrial signal potential values as an intermediate result , the above algorithm generates four sum and four difference values , again scaling by 2 \u2212 1 / 2 , and the four difference values become haar transformed values 5 to 8 , reflecting the strength of the middle frequency wavelets at four different points in time , and corresponding to wavelets w 5 to w 8 in fig3 . next , taking these four sums of sums of sums of adjacent atrial signal potential values , the above algorithm generates two sum and two difference values , again scaling by 2 \u2212 1 / 2 , and the two difference values become haar transformed values 3 and 4 , reflecting the strength of the second to the lowest frequency wavelets only two of which encompass all the time domain data , corresponding to the wavelets w 3 and w 4 shown in fig3 . and finally , the algorithm generates the sum of and the difference between the final remaining two sums of sums of sums of sums of atrial signal potential values , again scaling by 2 \u2212 1 / 2 . the difference value is then the haar transformed value 2 , representing the strength of the lowest - frequency wavelet , the one corresponding to the wavelet w 2 in fig3 the wavelet that extends the full length of the time scale . the sum value is then the first , or d . c ., transformed value , representing the average signal potential level over the 32 sampled points in time , which corresponds to the wavelet w 1 in fig3 . another way of viewing this computation is illustrated in fig4 a and 4b . the 32 atrial signal potential values are shown at the top of fig4 a and are identified as c 0 5 through c 31 5 . the superscript \u201c 5 \u201d indicates that these values are processed when the index value \u201c n \u201d in the above computer program is equal to \u201c 5 \u201d\u2014 that is , during the first pass through the data generating intermediary sums and differences . during subsequent passes , the value of n is decremented to 4 , 3 , 2 , and finally to 1 . during each pass through the data , the computer generates sums c n - 1 and differences d n - 1 , as shown in fig4 b , between adjacent pairs of the values c 0 n and c 1 n ; c 2 n and c 3 n ; and so on , so that the number of newly - generated c n - 1 and d n - 1 terms is reduced by half with each computer pass through the intermediary results . at the end of all these computations , the value c 0 0 is the first , or d . c ., haar transformed value ; and the values d 0 0 ; d 0 1 and d 1 1 ; d 0 2 , d 1 2 , d 2 2 , and d 3 2 ; d 0 3 , d 1 3 , . . . and d 7 3 ; and d 0 4 , d 1 4 , and d 15 4 are , respectively , the remaining haar transform values 2 through 32 . fig4 thus illustrates quite succinctly how all the transform computations are carried out , and how the intermediary \u201c c \u201d sum values , such as c 0 4 , are used to compute multiple haar values , such as the values d 0 3 , d 0 2 , d 0 1 , d 0 0 , and c 0 0 all of which are computed from the intermediary value c 0 4 . saving and reusing these intermediary \u201c c \u201d values saves much computational time . fig4 b indicates the precise addition and subtraction operations that are carried out at each level to compute the values in the next lower level , proceeding down through the chart presented in fig4 a . but in any given application to heart waveform analysis , all of these computations may not be needed , and accordingly the number of computations may be reduced much further . since the present invention teaches that only a small number of these haar transformed values need actually be considered , a far less computationally intensive transform can be developed which only generates the intermediate and final transformed values that are actually needed to generate the specific haar transformed values which have proved to be significant in discriminating between sinus tachycardia ( or st ) and non - st conditions . all others need not be computed , and the above program may be reduced to a special algorithm that omits as many sums , differences , and multiplications as possible . for example , if only the transformed values 1 , 5 , 9 , and 24 are significant , then 31 additions are required to compute the first transformed value ( the d . c . value \u2014 the simple sum of all the time domain signal values ); six additions and one subtraction are required to compute the fifth transformed value ( in fig3 the difference between the sums of time domain values under each half of the wavelet w 5 at 314 in fig3 ); two additions and one subtraction are required to compute the ninth transformed value ( in fig3 the difference between the sums of the time domain values under each half of the wavelet w9 at 322 in fig3 ); and only one subtraction is required to compute the 24 th transformed value ( in fig3 the difference between the two time domain values under the respective halves of the wavelet w 24 ( not shown ) which is the same size as , but differently positioned in time than , the wavelet w 18 at 330 ( fig3 ). accordingly , if only the transformed values 1 , 5 , 9 , and 24 actually evaluated , then only 42 additions and subtractions are required . but even this number can be reduced further . if the computational algorithm set forth in the above illustrative computer program is followed as a guide , then the intermediary sums used in computing the haar first , or d . c ., transformed value can be re - used to compute the sums for the haar transformed values 5 and 9 , assuming these intermediary results are saved in the manner described in the above program example . then 31 additions are still required to compute the first transformed value , but only one subtraction is required to compute each of the transformed values 5 , 9 , and 24 , giving a total of additions and subtraction of only 34 operations . and if only transformed values 1 , 5 , and 9 are computed , the number of additions and subtractions is reduced to 33 . and if the first transformed value is omitted and if transformed values 5 , 9 , and 24 are selected , then the 31 additions needed to compute the first transformed value are not required . then the number of computations for transformed value 5 is 8 additions and one subtraction ; the number of computations for transformed value 9 is 2 additions and one subtraction ; and the number of computations for transformed value 24 is still just one subtraction . so the total number of additions and subtractions is just 13 . but even this number can be reduced to 11 when it is realized that the two additions done for the transformed value 9 are also done ( in the above computational algorithm ) when computing the transformed value 5 . so the number of additions and subtractions can be reduced to 11 . also , when computing such a small number of transformed values , the number of multiplications can be reduced as well by postponing the multiplications until a transformed value is actually computed . with reference to fig4 a , instead of dividing every sum and difference by the square root of two ( as shown in fig4 b ), several vertical sums of \u201c c \u201d values in different rows of the fig4 a table can be formed , and then a \u201c d \u201d value can be computed by subtraction , and then the resulting unscaled transformed value can be scaled with a single multiplication by 2 \u2212 j / 2 where j is the number of rows in the table of fig4 a traversed by the one or more sum and the one difference computations . thus , the number of scaling multiplications can sometimes be equal to the number of transformed values that are computed , typically 3 or 4 . accordingly , by using wavelet analysis and transformation , instead of fourier or discrete cosine or other sinusoid analysis and transformation , a highly useful and highly frequency - specific result can be achieved with a very small number of computations , thereby conserving power and battery life , and yet achieving a high degree of precision in recognizing changes in morphology . in our tests , we have selected certain transformed values as being much more significant than other values in distinguishing between st and non - st . we focused upon those transform values whose \u201c+ 1 \u201d and \u201c\u2212 1 \u201d scope included the middle 8 time points most significant to analysis of the peak of the atrial depolarization . working across test data samples obtained from patients who had dual chamber icds , we computed the variance of each transformed value for st and non - st events using the standard statistical formula for computing variance . we then calculated the ratio of the variance of non - st events to the variance of st events , and selected those terms with the highest variance ratios for further consideration . in this manner , we reduced the number of transformed values that were included in the cwa or correlation process while always maintaining or improving the performance achieved . ultimately we settled upon the three or four transformed values having the highest ratios . these were the transform values 1 , 5 , 9 , and 24 . we tested 1 , 5 , and 24 together ; 5 , 9 , and 24 together ; and 1 , 5 , 9 , and 24 together . these selected haar transform values , when used in these combinations , required minimal computations and thereby produced a savings in battery power , and that also gave better performance at distinguishing st from non - st than did all of the transformed values used together . so an increase in accuracy was achieved as well as a decrease in computational complexity by this approach to atrial waveform analysis . [ 0054 ] fig5 for example , illustrates actual plots of digital information illustrating the effect of using only the three coefficients 1 , 5 , and 24 . at 502 , a normal waveform is shown , with amplitude plotted against sample numbers from 1 to 32 . at 504 , an abnormal waveform is shown , again with amplitude plotted against sample numbers . after the haar transformation , at 506 and 508 , the wavelet coefficient amplitude values are indicated for the coefficients 1 to 32 . at 506 , the coefficients generated by the normal waveform 502 are shown , and at 508 , the coefficients generated by the abnormal waveform 504 are shown . one may directly compare these component values and verify that the coefficients 1 , 5 , and 24 at 506 and at 508 vary significantly in amplitude . comparison of this same information among different patients ( not shown in this figure ) also indicated that this variation is relatively constant from one patient to another . at 510 and 512 , using only the coefficients 1 , 5 , and 24 , the heartbeat waveforms are reconstructed by a reverse transformation , the normal reconstructed waveform shown at 510 and the abnormal reconstructed waveform shown at 512 . the marked differences between these two reconstructed waveforms highlights the way in which these three coefficients can signal an abnormal condition with less computation . in a practical system , a small number of transform values are selected , in the manner just described . the algorithm for computing these values is then refined , as explained above , to reduce the number of additions , subtractions , and multiplications to the minimum possible while preserving the accuracy of the computations . we first compute a value \u03c1 , which is the correlation between these values with the corresponding values in a template that represents the values for an average normal population . we then compare this value \u03c1 to a threshold correlation value \u03b2 that is chosen to give optimal results on experimental data . ( see the examples in the tables presented below .) for each waveform analyzed , a test is made to determine whether \u03c1 is greater than \u03b2 . if so , then this waveform is placed into the buffer marked \u201c st \u201d at step 210 . if not , then this waveform is placed into the buffer marked \u201c non - st \u201d at step 211 . finally , after ten waveforms have been analyzed , a count of st and non - st marked waveforms is made to see if the count of non - st waveforms is greater than some threshold value x ( at step 212 ), where x can be , for example , 7 . if more than seven waveforms are non - st , then therapy is delivered at step 214 . the buffers at steps 210 and 211 in fig2 are buffers that hold the last ten decision values . these buffers can be pictured as sliding windows revealing the most recent ten values to permit the decision at 212 to be continuously updated . the buffers thus function as a nonlinear filter preventing irregular values from delivering therapy improperly . in developing a working prototype of this system , we used a development data set consisting of 20 episodes of st taken from 4 patients and 18 episodes of af taken from 7 patients . patient number episodes of st episodes of af 1 2 6 2 12 0 13 0 1 4 0 2 5 5 1 6 0 3 7 0 3 8 0 2 9 1 0 total 20 18 we used this data to optimize our choice of \u03c1 and \u03b2 and also the particular coefficients that we chose to examine . next , we tested our prototype using a set of 16 episodes of st obtained from 4 patients together with 17 episodes of af obtained from 7 patients . patient number episodes of st episodes of af 1 8 0 2 3 1 3 3 0 4 2 0 5 0 4 6 0 4 7 0 3 8 0 3 9 0 1 10 0 1 total 16 17 x out threshold coefficients of 10 ( \u03b2 ) sensitivity specificity 1 , 5 , 9 , & amp ; 24 3 0 . 808 76 % 94 % 1 , 5 , 9 , & amp ; 24 5 0 . 975 88 % 94 % 1 , 5 , 9 , & amp ; 24 6 0 . 976 82 % 94 % 1 , 5 , 9 , & amp ; 24 7 0 . 983 88 % 94 % 1 , 5 , & amp ; 24 3 0 . 815 82 % 94 % 1 , 5 , & amp ; 24 4 0 . 943 88 % 94 % 1 , 5 , & amp ; 24 5 0 . 956 82 % 94 % 1 , 5 , & amp ; 24 7 0 . 991 94 % 94 % 5 , 9 , & amp ; 24 8 0 . 9993 82 % 94 % 5 , 9 , & amp ; 24 9 0 . 9997 76 % 81 % the quality of the selected subset of transformed values for characterizing changes in morphology may be demonstrated graphically , as indicated in fig5 by performing a reverse haar transform using only the three or four or so selected transformed values . as can be seen ( at 510 and 512 in fig5 ), the inverse transformations produce distorted representations of the original atrial waveforms which illustrate how the choice of transform values can emphasize the changes . this is another useful way to assist one in selecting which transformed values are most useful . one feature of the invention is its ability to use a template derived from an average normal population , rather than deriving a patient customized average normal template for each individual patient . prior systems have required each new patient to be monitored while in a normal state , and the captured data was then averaged to form a patient specific normal template . contrary to this , the present invention achieved the results shown above using the same template for all patients . accordingly , the invention may be used with a new patient without the necessity of such preliminary testing of each patient and without a new customized template necessarily having to be created for each patient . one reason why the present invention can function using a template that represents average values for a normal population is because the specific coefficients selected in the transform domain , in addition to having been selected to emphasize the difference between normal and abnormal values , are also preferably chosen to minimize the differences between different patients . in some cases , values that were good at distinguishing between normal and abnormal rhythms for one patient were not as good at doing so for some other patient . these values are preferably not selected for use in implementing the present invention . for example , the transform coefficients may be selected such that all ( or most of ) the normal values of a particular coefficient gathered from many patients had the same sign ( positive or negative ) and similar amplitudes ( or if they varied in sign , they were near to zero ). in addition , the abnormal coefficients are significantly different in value from the normal coefficients for each particular patient . [ 0068 ] fig6 at 600 , illustrates one way in which this may be done . first , at step 602 , one gathers a large number of st and non - st data samples . next , all of this data is transformed , generating coefficients for every waveform set of data ( step 604 ). the variance of each coefficient is then computed first for the st samples ( step 606 ) and then for the non - st samples . then , for each coefficient , the ratio of the variance of the non - st samples to that of the st samples is computed ( step 608 ). finally , those coefficients whose variance ratio was large may be selected a coefficients for use in creating a population template and in testing patients ( step 612 ). in addition , at step 614 , the number of coefficients retained may be further reduced to reduce the number of computations that need to be performed , as explained above . in the version of the invention that was used to generate the data shown above , the population template was computed as follows , using the 20 st episodes in the development data set ( also used in the first of the two tables presented above ). this is illustrated at 700 in fig7 . first , the coefficients were selected as was explained above and as shown in fig6 . next , for each patient episode of st ( step 702 ), the recorded heartbeats were broken up into groups of ten consecutive heartbeats ( step 704 ). for each group ( step 706 ), the selected coefficients of the haar transform were computed ( step 708 ) for each of the ten heartbeats and the median value was then selected ( step 710 ) for each coefficient to form a proto - template . if these median values correlated well with the coefficient values for each of the ten heartbeats , the proto - template was retained ( step 714 ). otherwise , it was discarded . in this manner , a whole bunch of proto - templates were obtained from each patient episode . next , median values of the coefficients from all of the proto - templates ( step 716 ) were computed ( step 718 ). these values were then used as the population template for distinguishing st from non - st events ( step 720 ). while the preferred embodiment of the invention has been described above , it is to be understood that numerous modifications and changes will occur to those who are skilled in the art to which the invention pertains . accordingly , the following claims annexed to and forming a part of this specification are intended to define the true spirit and scope of the invention \u2014 that is , what is new and what is desired to be secured by letters patent of the united states ."}
|
{"patent": "the present invention is directed towards improving the quality and functionality of both implantable cardioverter defibrillators ( icds ) and implantable atrial defibrillators ( iads ). all such devices need to have a method and mechanism for accurately distinguishing sinus tachycardia ( st ) form non - st , such as atrial fibrillation and atrial flutter . this method and mechanism should be conservative , biased towards not initiating therapy unless non - st is clearly indicated . it should also require as few machine computational cycles as possible to reduce icd and iad battery drain . the principle underlying the present invention is that st and non - st can be distinguished by studying and measuring the atrial electrogram morphology to detect aberrant conduction in the atria . thus , by comparing normal st waveforms to a series of newly - occurring waveforms , st is indicated by abnormal atrial waveform morphology which signals aberrant conduction within the atria improper electrical patterns of depolarization . this will show up in the signal potentials captured by a bipolar atrial lead . with reference to fig1 the present invention can be implemented in an icd 100 ( which could be an iad ) that is implanted in the chest of a patient who suffers from unpleasant and possibly life - threatening irregularities in heart operation due to improper electrical stimulation of the heart muscle cells . during normal heart action , the heart &# 39 ; s electrical impulses originate in the sino - atrial node as an action potential that is transmitted smoothly to all portions of the atria , causing contraction of the atrial chambers . the electrical impulse continues in its path to a cluster of conduction fibrils known as the atrioventricular node . after a delay of about one - tenth of a second , an action potential flows over the ventricles and causes them to contract in synchronism with and following shortly after the atrial contraction . in this manner , the heart pumps blood to the lungs and from the lungs to the body . a bipolar lead 102 is implanted within the atria 104 of a heart 103 to measure the electrical potential between nearby cells of the atria . as the action potential actively passes from cell to cell past the two electrodes of the lead 102 , an oscillating signal potential is developed across the two electrodes and is conveyed by the bipolar lead 102 to the icd 100 where the signal is sampled about 1000 times per second and is digitized , with the sequential samples stored within a ram memory within the icd 100 for further analysis . another bipolar lead 105 may extend to the ventricle ( not shown ) of the heart 103 to measure the action potentials generated within the ventricles . these same leads , or other leads , may be used by the icd for administering various types of therapy , such as pacing or defibrillation shock therapy . the data samples collected from within the atria are now analyzed . first , with reference to data gathered from the ventricles , if some event in the ventricle , such as a mis - timed depolarization event that partially overlaps the atria &# 39 ; s p wave , could leak across to the atria and distort the measurement of the potentials for a given atrial depolarization , then the data for that particular heartbeat is discarded and is not used in further analyses . such distortion can also be found simply because a given heartbeat set of data gathered from the atria is itself badly distorted , and this approach must be taken in the case of an iad having no ventricular lead . the remaining data is retained for further processing . in preparation for further processing , the data for individual heartbeats is identified and gathered . the position within the gathered data of the negative spike to the atrial p depolarization waveform is determined and is selected as the center of the waveform for each beat . then , since data close to this p - wave is to be analyzed , and since data further away may be distorted to some degree by ventricular activity or by variations in the p - to - p interval , 32 sequential data samples are selected for each heartbeat such that the negative spike of the p - wave becomes signal potential sample number 16 of the 32 sequential samples selected for further analyses . these samples for a normal heartbeat appear at 302 in fig3 and at 502 in fig5 where the amplitude of the atrial signal potential is plotted against time over the 32 sampled values , with the negative p - wave spike positioned as signal sample number 16 . likewise , the samples for an abnormal heartbeat appear at 504 in fig5 . straightforward correlation ( cwa ) of the 32 samples against a template containing a typical or normal pattern could be performed at this juncture , as taught on page 563 ( equation ( 1 )) of the article by thorne , cited in the introductory portion of this specification . but the performance of such a cross correlation at repeated regular intervals would generate numerous computations and would drain the icd &# 39 ; s battery more rapidly than is desirable . in addition , provision would have to be made for a full template waveform that can be stored within the icd to be used as a comparison reference . in addition , such templates generally need to be patient specific , and they are more susceptible to normal changes in shape . accordingly , to reduce the mathematical complexity of the cross correlation computations , it is desirable to reduce the number of data points from 32 down to some much lower number through the use of some form of transformation into a different set of values . for example , the data could be transformed using the karhunen - loeve transformation described in the morris article cited at the beginning of this specification . but that transformation is fairly computationally intensive . computationally , a fast fourier transformation would be more efficient , but any such transformation into the frequency domain , where variable frequency sine and cosine wave amplitudes result from the transformation , does not match itself well to the morphology of atrial heart waveforms , which tend to be formed from large , ringing , spike - like representations of p - wave depolarization events travelling as moving action potential fronts past the pair of electrodes attached to the bipolar lead , and not as steady harmonics extending across time . accordingly , fourier - class transforms do not adequately reduce the number of significant data values that must be considered . and in addition , when the \u201c window \u201d size for a fourier analysis is selected , if it is wide , then the frequency values are not time - specific , but represent average signal harmonic content over the entire window time duration . and if the window is narrow , then the harmonics are too spread out in frequency , and the frequency - domain data generated by the transform does not resolve things finely enough in the frequency domain . for all of the above reasons , the present invention analyzes the waveforms using a discrete wavelet transform . this has the advantage that while low frequency wavelets are broad in time , high frequency wavelets have a much narrower timeslot focus . thus , for example , with 32 data samples taken sequentially over time , a wavelet transformation generates a d . c . wavelet that spans the entire set of 32 points ; a first reversing wavelet that also spans the entire set of points ; two second harmonic wavelets that each focus upon only one - half of the 32 data points ; four third harmonic wavelets that each focus upon only one - fourth of the 32 data points ; eight fourth harmonic wavelets that each focus upon four of the 32 data points ; and 16 fifth harmonic wavelets that each focuses upon only two of the 32 data points . accordingly , the higher - frequency wavelet amplitudes are quite time specific , unlike fourier sinusoidal amplitudes , while the lower - frequency and d . c . wavelet amplitudes give one broad information about the whole set of 32 points . and like the fourier transform , the discrete wavelet transform preserves all of the information of the original atrial signal , such that the transformation is fully reversible . 32 discrete wavelet transform values may be reverse transformed back into 32 values indicating the atrial signal potentials at 32 sequential points in time . in essence , the haar function is performing multiple digital filtering operations , at variable positions and utilizing varying - width windowing functions , to develop component values that represent varying types of heartbeat activity , at varying frequencies , spread over varying widths , and located at varying positions . these component values can be studied to see which subset of these component values are good at distinguishing st from non - st , or ( more generally ) which subset of these component values are good at distinguishing one type of heartbeat waveform from another . further study of such a selected subset can further determine which of the subset of components are relatively invariant from one individual &# 39 ; s heart to another . one preferably selects component values that are suitable from both of these two perspectives . the particular wavelet transformation to choose can be tailored to the nature of the data , with the wavelets chosen to resemble somewhat the values to be found within the data so that some transformed values are of much larger amplitude than others , and such that the low amplitude transformed values may be disregarded . of particular importance with an atrial p waveform of the type being analyzed here are wavelets having frequency values roughly comparable to and timed to coincide with the ringing of the atrial depolarization . this tailoring can minimize the number of transformed data values that are significant and that are candidates for full participation in the correlation analyses to determine if there has been a substantial change in waveform morphology . another factor in selecting a particular wavelet transformation is reducing the number of computations that must be performed to carry out the transformation . yet another factor is selecting a wavelet transformation where some of the transformed component values vary from normal to abnormal heart rhythms in much the same over a population of individuals so that a generic template of these component values may be derived that can be used with many individuals , rather than with just one individual . these factors suggest that the haar transformation would be a suitable candidate for use in analysis of atrial waveforms and possibly other waveforms as well . with reference to fig3 in the case of 32 voltage values sampled over time , the 32 applicable haar discrete transform wavelets appear as shown in this figure . other discrete wavelet transforms based upon wavelets having triangular or other shapes may also prove usable . the haar transform wavelets , in particular , have the shape of a square waveform , as will be described , and this can simplify the computations required . in fig3 time increases from left to right . a scale 304 indicates the 32 points in time at which the atrial waveform is sampled , and an analog atrial waveform ( before digitization ) is shown at 302 . the p wave negative notch 303 is shown centered at the 16 th timeslot so that the signal potential of the atria waveform sampled at this point in time becomes the 16 th data value in the set of 32 data values that are to be subjected to the haar transformation . a haar wavelet is simply one cycle of a square waveform that starts at zero , then swings positive ( to \u201c+ 1 \u201d), then swings negative ( to \u201c\u2212 1 \u201d), and then swings back to zero . the wavelet w 2 , shown at 308 in fig3 for example , is at \u201c+ 1 \u201d at points in time 1 to 16 and is at \u201c\u2212 1 \u201d at points in time 17 to 32 . the shorter waveform w 6 , shown at 316 in fig3 is at \u201c+ 1 \u201d at points in time 9 to 12 and is at \u201c\u2212 1 \u201d at points in time 13 to 16 , and is at \u201c 0 \u201d at all other points in time . the waveform w 1 , shown at 306 in fig3 is so slow to fluctuate in time that its \u201c\u2212 1 \u201d portion is off of the chart ( in fig3 ) to the right , and is ignored ; and accordingly , it has the value \u201c+ 1 \u201d for all 32 of the points in time shown in fig3 . it thus represents the average , or d . c ., component of the atrial signal potential . the lowest frequency wavelets w 1 and w 2 encompass the entire set of 32 points in time . the higher frequency wavelets w 3 and w 4 each encompasses only half of the points in time , and the two wavelets taken together encompass all the points in time . the four wavelets w 5 , w 6 , w 7 , and w 8 each encompasses only eight points in time , and the four wavelets taken together encompass all points in time . the eight wavelets w 9 . . . w 16 each encompasses only four points in time , and together all eight wavelets encompass all points in time . and finally , the wavelets w 17 . . . w 32 each encompass only two points in time , but the sixteen wavelets together encompass all points in time . thus , each wavelet has a characteristic time span and position as well as a characteristic frequency , with higher - frequency wavelets having a narrower and more specific time span and position than lower - frequency wavelets . this is advantageous with an impulse - type , ringing signal such as the atrial depolarization waveform considered here , since many higher - frequency transform values that correspond to haar wavelets positioned in time away from the p waveform or that do not correspond to its frequency of ringing may be low in amplitude such that they may safely be ignored during the analysis steps , thereby reducing the data that must be processed during the correlation steps . ( the above , while presented in the context of the haar wavelet , is also applicable to other wavelet shapes that may used to perform a discrete wavelet transform ( dwt )). each of the 32 haar transformed values is computed as follows : multiply each of the 32 sampled and digitized voltage values representing the analog atrial waveform 302 by the correspondingly - positioned -( in - time ) amplitude values of one of the 32 haar wavelets ( shown in fig3 ); then sum the resulting products ; and then multiply the resulting sums by a discrete wavelet transform scaling factor 2 \u2212 j / 2 , where j equals 1 for the wavelets w 17 to w 32 , 2 for w 9 to w 16 , 3 for w 5 to w 8 , 4 for w 3 to w 4 , and 5 for w 1 and w 2 . for example , and referring to fig3 : the first wavelet w 1 is always + 1 , so the atrial signal potential values are simply summed and then multiplied by 2 \u2212 5 / 2 ; the second wavelet is + 1 for time points 1 to 16 and \u2212 1 for time points 17 to 32 , so the first 16 values are summed , the second 16 values are summed , the difference between the first and second sums is computed , and the result is multiplied by 2 \u2212 5 / 2 ; and so on until all the transformed values have been computed , one for each haar wavelet . the 32 time - sequential atrial signal potential values are thus transformed into 32 haar wavelet amplitude values which may be called \u201c transformed values \u201d and which may be assigned the numbers 1 to 32 corresponding to the subscript numbers of the haar wavelets to which they each correspond and whose amplitudes they represent . the above description of how to compute the haar wavelet transformed values is accurate , but it is not the most efficient way to proceed in the case of haar wavelets . like the fourier transform , which has a corresponding fast fourier transform that saves intermediate results and thereby avoids re - computing them and thus reduces substantially the number of computations , the haar transformation also may be carried out in a manner that saves intermediate sums and re - uses them to reduce substantially the number of computations . this is described below at the point where fig4 is described , since fig4 illustrates graphically how this can be done . the illustrative program listing presented below is also an implementation of the fast haar wavelet transformation algorithm . the haar transformation can readily be carried out by any digital computer . as an example of how the haar transformation can be carried out on 32 values , the following program , written in the language c , is illustrative of many possible programs that may be written . in the exemplary program that follows , a 32 - element array yy contains 32 data values representing the 32 sampled atrial signal potential values representing the fluctuations over time of the signal supplied by the bipolar lead 102 . the analog atrial signal is sampled , digitized , broken into separate beat data sets , pre - processed ( to remove waveforms distorted by ventricular activity ), centered ( with the negative depolarization spike at data point 16 ), and fed into the subroutine presented below . this subroutine is compiled ( or assembled , if rewritten in assembly language ) and installed in the rom of the icd &# 39 ; s embedded microprocessor . this subroutine returns , contained within the same array yy , the 32 transformed haar wavelet amplitude values described above . ( the constant value \u201c recipsqrttwo \u201d is the reciprocal of the square root of two ). an illustrative version of the subroutine for computing all 32 of the haar transformed values in an efficient manner is presented here : void haar ( double yy []) { int i , j , l ; double zz [ 32 ]; for ( i = 5 ; i & gt ; 0 ; i \u2212\u2212) { l = 1 ; for ( j = 1 ; j & lt ;= i ; ++ j ) l = 2 * l ; for ( j = 0 ; j & lt ; l ; ++ j ) zz [ j ] = yy [ j ]; for ( j = 0 ; j & lt ; l \u2212 1 ; j = j + 2 ) ( yy [ j / 2 ] = recipsqrttwo * ( zz [ j ] + zz [ j + 1 ]); yy [( j + l )/ 2 ] = rcipsqrttwo * ( zz [ j ] \u2212 zz [ j + 1 ]); } } return ; } the above program computes the 32 haar transformed values 1 to 32 from the 32 atrial signal potential time - sequenced input values . it does so with only 31 additions , 31 subtractions , and 62 multiplications . it is carefully designed to compute each value in a systematic way , making multiple use of intermediate results . first , the program computes sixteen sums of and sixteen differences between eight adjacent pairs of the 16 time sequenced atrial signal potential values , and the sixteen difference values become the haar transformed values 17 to 32 , reflecting the strength of the highest frequency haar wavelets positioned at 16 different positions in time , corresponding to the wavelets w 17 to w 32 shown in fig3 . all the sum and difference values are scaled by multiplication by 2 \u2212 1 / 2 . next , taking the 16 sums of atrial signal potential values as an intermediate result , the above algorithm generates eight sums of and eight differences between adjacent pairs of these sixteen intermediate values that resulted from the first sixteen additions , and the eight newly - computed difference values become the haar transformed values 9 to 16 , reflecting the strength of the second to the highest frequency values at eight different points in time , corresponding to wavelets w 9 to w 16 in fig3 . all of these eight sum and eight difference values are again scaled by 2 \u2212 1 / 2 so that the wavelets w 9 to w 16 are scaled by \u00bd ( 2 \u2212 2 / 2 or 2 \u2212 1 / 2 times 2 \u2212 1 / 2 ). next , taking these eight sums of sums of adjacent atrial signal potential values as an intermediate result , the above algorithm generates four sum and four difference values , again scaling by 2 \u2212 1 / 2 , and the four difference values become haar transformed values 5 to 8 , reflecting the strength of the middle frequency wavelets at four different points in time , and corresponding to wavelets w 5 to w 8 in fig3 . next , taking these four sums of sums of sums of adjacent atrial signal potential values , the above algorithm generates two sum and two difference values , again scaling by 2 \u2212 1 / 2 , and the two difference values become haar transformed values 3 and 4 , reflecting the strength of the second to the lowest frequency wavelets only two of which encompass all the time domain data , corresponding to the wavelets w 3 and w 4 shown in fig3 . and finally , the algorithm generates the sum of and the difference between the final remaining two sums of sums of sums of sums of atrial signal potential values , again scaling by 2 \u2212 1 / 2 . the difference value is then the haar transformed value 2 , representing the strength of the lowest - frequency wavelet , the one corresponding to the wavelet w 2 in fig3 the wavelet that extends the full length of the time scale . the sum value is then the first , or d . c ., transformed value , representing the average signal potential level over the 32 sampled points in time , which corresponds to the wavelet w 1 in fig3 . another way of viewing this computation is illustrated in fig4 a and 4b . the 32 atrial signal potential values are shown at the top of fig4 a and are identified as c 0 5 through c 31 5 . the superscript \u201c 5 \u201d indicates that these values are processed when the index value \u201c n \u201d in the above computer program is equal to \u201c 5 \u201d\u2014 that is , during the first pass through the data generating intermediary sums and differences . during subsequent passes , the value of n is decremented to 4 , 3 , 2 , and finally to 1 . during each pass through the data , the computer generates sums c n - 1 and differences d n - 1 , as shown in fig4 b , between adjacent pairs of the values c 0 n and c 1 n ; c 2 n and c 3 n ; and so on , so that the number of newly - generated c n - 1 and d n - 1 terms is reduced by half with each computer pass through the intermediary results . at the end of all these computations , the value c 0 0 is the first , or d . c ., haar transformed value ; and the values d 0 0 ; d 0 1 and d 1 1 ; d 0 2 , d 1 2 , d 2 2 , and d 3 2 ; d 0 3 , d 1 3 , . . . and d 7 3 ; and d 0 4 , d 1 4 , and d 15 4 are , respectively , the remaining haar transform values 2 through 32 . fig4 thus illustrates quite succinctly how all the transform computations are carried out , and how the intermediary \u201c c \u201d sum values , such as c 0 4 , are used to compute multiple haar values , such as the values d 0 3 , d 0 2 , d 0 1 , d 0 0 , and c 0 0 all of which are computed from the intermediary value c 0 4 . saving and reusing these intermediary \u201c c \u201d values saves much computational time . fig4 b indicates the precise addition and subtraction operations that are carried out at each level to compute the values in the next lower level , proceeding down through the chart presented in fig4 a . but in any given application to heart waveform analysis , all of these computations may not be needed , and accordingly the number of computations may be reduced much further . since the present invention teaches that only a small number of these haar transformed values need actually be considered , a far less computationally intensive transform can be developed which only generates the intermediate and final transformed values that are actually needed to generate the specific haar transformed values which have proved to be significant in discriminating between sinus tachycardia ( or st ) and non - st conditions . all others need not be computed , and the above program may be reduced to a special algorithm that omits as many sums , differences , and multiplications as possible . for example , if only the transformed values 1 , 5 , 9 , and 24 are significant , then 31 additions are required to compute the first transformed value ( the d . c . value \u2014 the simple sum of all the time domain signal values ); six additions and one subtraction are required to compute the fifth transformed value ( in fig3 the difference between the sums of time domain values under each half of the wavelet w 5 at 314 in fig3 ); two additions and one subtraction are required to compute the ninth transformed value ( in fig3 the difference between the sums of the time domain values under each half of the wavelet w9 at 322 in fig3 ); and only one subtraction is required to compute the 24 th transformed value ( in fig3 the difference between the two time domain values under the respective halves of the wavelet w 24 ( not shown ) which is the same size as , but differently positioned in time than , the wavelet w 18 at 330 ( fig3 ). accordingly , if only the transformed values 1 , 5 , 9 , and 24 actually evaluated , then only 42 additions and subtractions are required . but even this number can be reduced further . if the computational algorithm set forth in the above illustrative computer program is followed as a guide , then the intermediary sums used in computing the haar first , or d . c ., transformed value can be re - used to compute the sums for the haar transformed values 5 and 9 , assuming these intermediary results are saved in the manner described in the above program example . then 31 additions are still required to compute the first transformed value , but only one subtraction is required to compute each of the transformed values 5 , 9 , and 24 , giving a total of additions and subtraction of only 34 operations . and if only transformed values 1 , 5 , and 9 are computed , the number of additions and subtractions is reduced to 33 . and if the first transformed value is omitted and if transformed values 5 , 9 , and 24 are selected , then the 31 additions needed to compute the first transformed value are not required . then the number of computations for transformed value 5 is 8 additions and one subtraction ; the number of computations for transformed value 9 is 2 additions and one subtraction ; and the number of computations for transformed value 24 is still just one subtraction . so the total number of additions and subtractions is just 13 . but even this number can be reduced to 11 when it is realized that the two additions done for the transformed value 9 are also done ( in the above computational algorithm ) when computing the transformed value 5 . so the number of additions and subtractions can be reduced to 11 . also , when computing such a small number of transformed values , the number of multiplications can be reduced as well by postponing the multiplications until a transformed value is actually computed . with reference to fig4 a , instead of dividing every sum and difference by the square root of two ( as shown in fig4 b ), several vertical sums of \u201c c \u201d values in different rows of the fig4 a table can be formed , and then a \u201c d \u201d value can be computed by subtraction , and then the resulting unscaled transformed value can be scaled with a single multiplication by 2 \u2212 j / 2 where j is the number of rows in the table of fig4 a traversed by the one or more sum and the one difference computations . thus , the number of scaling multiplications can sometimes be equal to the number of transformed values that are computed , typically 3 or 4 . accordingly , by using wavelet analysis and transformation , instead of fourier or discrete cosine or other sinusoid analysis and transformation , a highly useful and highly frequency - specific result can be achieved with a very small number of computations , thereby conserving power and battery life , and yet achieving a high degree of precision in recognizing changes in morphology . in our tests , we have selected certain transformed values as being much more significant than other values in distinguishing between st and non - st . we focused upon those transform values whose \u201c+ 1 \u201d and \u201c\u2212 1 \u201d scope included the middle 8 time points most significant to analysis of the peak of the atrial depolarization . working across test data samples obtained from patients who had dual chamber icds , we computed the variance of each transformed value for st and non - st events using the standard statistical formula for computing variance . we then calculated the ratio of the variance of non - st events to the variance of st events , and selected those terms with the highest variance ratios for further consideration . in this manner , we reduced the number of transformed values that were included in the cwa or correlation process while always maintaining or improving the performance achieved . ultimately we settled upon the three or four transformed values having the highest ratios . these were the transform values 1 , 5 , 9 , and 24 . we tested 1 , 5 , and 24 together ; 5 , 9 , and 24 together ; and 1 , 5 , 9 , and 24 together . these selected haar transform values , when used in these combinations , required minimal computations and thereby produced a savings in battery power , and that also gave better performance at distinguishing st from non - st than did all of the transformed values used together . so an increase in accuracy was achieved as well as a decrease in computational complexity by this approach to atrial waveform analysis . [ 0054 ] fig5 for example , illustrates actual plots of digital information illustrating the effect of using only the three coefficients 1 , 5 , and 24 . at 502 , a normal waveform is shown , with amplitude plotted against sample numbers from 1 to 32 . at 504 , an abnormal waveform is shown , again with amplitude plotted against sample numbers . after the haar transformation , at 506 and 508 , the wavelet coefficient amplitude values are indicated for the coefficients 1 to 32 . at 506 , the coefficients generated by the normal waveform 502 are shown , and at 508 , the coefficients generated by the abnormal waveform 504 are shown . one may directly compare these component values and verify that the coefficients 1 , 5 , and 24 at 506 and at 508 vary significantly in amplitude . comparison of this same information among different patients ( not shown in this figure ) also indicated that this variation is relatively constant from one patient to another . at 510 and 512 , using only the coefficients 1 , 5 , and 24 , the heartbeat waveforms are reconstructed by a reverse transformation , the normal reconstructed waveform shown at 510 and the abnormal reconstructed waveform shown at 512 . the marked differences between these two reconstructed waveforms highlights the way in which these three coefficients can signal an abnormal condition with less computation . in a practical system , a small number of transform values are selected , in the manner just described . the algorithm for computing these values is then refined , as explained above , to reduce the number of additions , subtractions , and multiplications to the minimum possible while preserving the accuracy of the computations . we first compute a value \u03c1 , which is the correlation between these values with the corresponding values in a template that represents the values for an average normal population . we then compare this value \u03c1 to a threshold correlation value \u03b2 that is chosen to give optimal results on experimental data . ( see the examples in the tables presented below .) for each waveform analyzed , a test is made to determine whether \u03c1 is greater than \u03b2 . if so , then this waveform is placed into the buffer marked \u201c st \u201d at step 210 . if not , then this waveform is placed into the buffer marked \u201c non - st \u201d at step 211 . finally , after ten waveforms have been analyzed , a count of st and non - st marked waveforms is made to see if the count of non - st waveforms is greater than some threshold value x ( at step 212 ), where x can be , for example , 7 . if more than seven waveforms are non - st , then therapy is delivered at step 214 . the buffers at steps 210 and 211 in fig2 are buffers that hold the last ten decision values . these buffers can be pictured as sliding windows revealing the most recent ten values to permit the decision at 212 to be continuously updated . the buffers thus function as a nonlinear filter preventing irregular values from delivering therapy improperly . in developing a working prototype of this system , we used a development data set consisting of 20 episodes of st taken from 4 patients and 18 episodes of af taken from 7 patients . patient number episodes of st episodes of af 1 2 6 2 12 0 13 0 1 4 0 2 5 5 1 6 0 3 7 0 3 8 0 2 9 1 0 total 20 18 we used this data to optimize our choice of \u03c1 and \u03b2 and also the particular coefficients that we chose to examine . next , we tested our prototype using a set of 16 episodes of st obtained from 4 patients together with 17 episodes of af obtained from 7 patients . patient number episodes of st episodes of af 1 8 0 2 3 1 3 3 0 4 2 0 5 0 4 6 0 4 7 0 3 8 0 3 9 0 1 10 0 1 total 16 17 x out threshold coefficients of 10 ( \u03b2 ) sensitivity specificity 1 , 5 , 9 , & amp ; 24 3 0 . 808 76 % 94 % 1 , 5 , 9 , & amp ; 24 5 0 . 975 88 % 94 % 1 , 5 , 9 , & amp ; 24 6 0 . 976 82 % 94 % 1 , 5 , 9 , & amp ; 24 7 0 . 983 88 % 94 % 1 , 5 , & amp ; 24 3 0 . 815 82 % 94 % 1 , 5 , & amp ; 24 4 0 . 943 88 % 94 % 1 , 5 , & amp ; 24 5 0 . 956 82 % 94 % 1 , 5 , & amp ; 24 7 0 . 991 94 % 94 % 5 , 9 , & amp ; 24 8 0 . 9993 82 % 94 % 5 , 9 , & amp ; 24 9 0 . 9997 76 % 81 % the quality of the selected subset of transformed values for characterizing changes in morphology may be demonstrated graphically , as indicated in fig5 by performing a reverse haar transform using only the three or four or so selected transformed values . as can be seen ( at 510 and 512 in fig5 ), the inverse transformations produce distorted representations of the original atrial waveforms which illustrate how the choice of transform values can emphasize the changes . this is another useful way to assist one in selecting which transformed values are most useful . one feature of the invention is its ability to use a template derived from an average normal population , rather than deriving a patient customized average normal template for each individual patient . prior systems have required each new patient to be monitored while in a normal state , and the captured data was then averaged to form a patient specific normal template . contrary to this , the present invention achieved the results shown above using the same template for all patients . accordingly , the invention may be used with a new patient without the necessity of such preliminary testing of each patient and without a new customized template necessarily having to be created for each patient . one reason why the present invention can function using a template that represents average values for a normal population is because the specific coefficients selected in the transform domain , in addition to having been selected to emphasize the difference between normal and abnormal values , are also preferably chosen to minimize the differences between different patients . in some cases , values that were good at distinguishing between normal and abnormal rhythms for one patient were not as good at doing so for some other patient . these values are preferably not selected for use in implementing the present invention . for example , the transform coefficients may be selected such that all ( or most of ) the normal values of a particular coefficient gathered from many patients had the same sign ( positive or negative ) and similar amplitudes ( or if they varied in sign , they were near to zero ). in addition , the abnormal coefficients are significantly different in value from the normal coefficients for each particular patient . [ 0068 ] fig6 at 600 , illustrates one way in which this may be done . first , at step 602 , one gathers a large number of st and non - st data samples . next , all of this data is transformed , generating coefficients for every waveform set of data ( step 604 ). the variance of each coefficient is then computed first for the st samples ( step 606 ) and then for the non - st samples . then , for each coefficient , the ratio of the variance of the non - st samples to that of the st samples is computed ( step 608 ). finally , those coefficients whose variance ratio was large may be selected a coefficients for use in creating a population template and in testing patients ( step 612 ). in addition , at step 614 , the number of coefficients retained may be further reduced to reduce the number of computations that need to be performed , as explained above . in the version of the invention that was used to generate the data shown above , the population template was computed as follows , using the 20 st episodes in the development data set ( also used in the first of the two tables presented above ). this is illustrated at 700 in fig7 . first , the coefficients were selected as was explained above and as shown in fig6 . next , for each patient episode of st ( step 702 ), the recorded heartbeats were broken up into groups of ten consecutive heartbeats ( step 704 ). for each group ( step 706 ), the selected coefficients of the haar transform were computed ( step 708 ) for each of the ten heartbeats and the median value was then selected ( step 710 ) for each coefficient to form a proto - template . if these median values correlated well with the coefficient values for each of the ten heartbeats , the proto - template was retained ( step 714 ). otherwise , it was discarded . in this manner , a whole bunch of proto - templates were obtained from each patient episode . next , median values of the coefficients from all of the proto - templates ( step 716 ) were computed ( step 718 ). these values were then used as the population template for distinguishing st from non - st events ( step 720 ). while the preferred embodiment of the invention has been described above , it is to be understood that numerous modifications and changes will occur to those who are skilled in the art to which the invention pertains . accordingly , the following claims annexed to and forming a part of this specification are intended to define the true spirit and scope of the invention \u2014 that is , what is new and what is desired to be secured by letters patent of the united states .", "category": "Performing Operations; Transporting"}
|
Is the categorization of this patent accurate?
| 0.25 |
1cd5305994f1692ec7ec09368a2a725ff3c3dd15c21bf56c12e04e558cdb5fbc
| 0.123535 | 0.018799 | 0.066406 | 0.185547 | 0.138672 | 0.3125 |
null |
{"category": "Human Necessities", "patent": "the present invention is directed towards improving the quality and functionality of both implantable cardioverter defibrillators ( icds ) and implantable atrial defibrillators ( iads ). all such devices need to have a method and mechanism for accurately distinguishing sinus tachycardia ( st ) form non - st , such as atrial fibrillation and atrial flutter . this method and mechanism should be conservative , biased towards not initiating therapy unless non - st is clearly indicated . it should also require as few machine computational cycles as possible to reduce icd and iad battery drain . the principle underlying the present invention is that st and non - st can be distinguished by studying and measuring the atrial electrogram morphology to detect aberrant conduction in the atria . thus , by comparing normal st waveforms to a series of newly - occurring waveforms , st is indicated by abnormal atrial waveform morphology which signals aberrant conduction within the atria improper electrical patterns of depolarization . this will show up in the signal potentials captured by a bipolar atrial lead . with reference to fig1 the present invention can be implemented in an icd 100 ( which could be an iad ) that is implanted in the chest of a patient who suffers from unpleasant and possibly life - threatening irregularities in heart operation due to improper electrical stimulation of the heart muscle cells . during normal heart action , the heart &# 39 ; s electrical impulses originate in the sino - atrial node as an action potential that is transmitted smoothly to all portions of the atria , causing contraction of the atrial chambers . the electrical impulse continues in its path to a cluster of conduction fibrils known as the atrioventricular node . after a delay of about one - tenth of a second , an action potential flows over the ventricles and causes them to contract in synchronism with and following shortly after the atrial contraction . in this manner , the heart pumps blood to the lungs and from the lungs to the body . a bipolar lead 102 is implanted within the atria 104 of a heart 103 to measure the electrical potential between nearby cells of the atria . as the action potential actively passes from cell to cell past the two electrodes of the lead 102 , an oscillating signal potential is developed across the two electrodes and is conveyed by the bipolar lead 102 to the icd 100 where the signal is sampled about 1000 times per second and is digitized , with the sequential samples stored within a ram memory within the icd 100 for further analysis . another bipolar lead 105 may extend to the ventricle ( not shown ) of the heart 103 to measure the action potentials generated within the ventricles . these same leads , or other leads , may be used by the icd for administering various types of therapy , such as pacing or defibrillation shock therapy . the data samples collected from within the atria are now analyzed . first , with reference to data gathered from the ventricles , if some event in the ventricle , such as a mis - timed depolarization event that partially overlaps the atria &# 39 ; s p wave , could leak across to the atria and distort the measurement of the potentials for a given atrial depolarization , then the data for that particular heartbeat is discarded and is not used in further analyses . such distortion can also be found simply because a given heartbeat set of data gathered from the atria is itself badly distorted , and this approach must be taken in the case of an iad having no ventricular lead . the remaining data is retained for further processing . in preparation for further processing , the data for individual heartbeats is identified and gathered . the position within the gathered data of the negative spike to the atrial p depolarization waveform is determined and is selected as the center of the waveform for each beat . then , since data close to this p - wave is to be analyzed , and since data further away may be distorted to some degree by ventricular activity or by variations in the p - to - p interval , 32 sequential data samples are selected for each heartbeat such that the negative spike of the p - wave becomes signal potential sample number 16 of the 32 sequential samples selected for further analyses . these samples for a normal heartbeat appear at 302 in fig3 and at 502 in fig5 where the amplitude of the atrial signal potential is plotted against time over the 32 sampled values , with the negative p - wave spike positioned as signal sample number 16 . likewise , the samples for an abnormal heartbeat appear at 504 in fig5 . straightforward correlation ( cwa ) of the 32 samples against a template containing a typical or normal pattern could be performed at this juncture , as taught on page 563 ( equation ( 1 )) of the article by thorne , cited in the introductory portion of this specification . but the performance of such a cross correlation at repeated regular intervals would generate numerous computations and would drain the icd &# 39 ; s battery more rapidly than is desirable . in addition , provision would have to be made for a full template waveform that can be stored within the icd to be used as a comparison reference . in addition , such templates generally need to be patient specific , and they are more susceptible to normal changes in shape . accordingly , to reduce the mathematical complexity of the cross correlation computations , it is desirable to reduce the number of data points from 32 down to some much lower number through the use of some form of transformation into a different set of values . for example , the data could be transformed using the karhunen - loeve transformation described in the morris article cited at the beginning of this specification . but that transformation is fairly computationally intensive . computationally , a fast fourier transformation would be more efficient , but any such transformation into the frequency domain , where variable frequency sine and cosine wave amplitudes result from the transformation , does not match itself well to the morphology of atrial heart waveforms , which tend to be formed from large , ringing , spike - like representations of p - wave depolarization events travelling as moving action potential fronts past the pair of electrodes attached to the bipolar lead , and not as steady harmonics extending across time . accordingly , fourier - class transforms do not adequately reduce the number of significant data values that must be considered . and in addition , when the \u201c window \u201d size for a fourier analysis is selected , if it is wide , then the frequency values are not time - specific , but represent average signal harmonic content over the entire window time duration . and if the window is narrow , then the harmonics are too spread out in frequency , and the frequency - domain data generated by the transform does not resolve things finely enough in the frequency domain . for all of the above reasons , the present invention analyzes the waveforms using a discrete wavelet transform . this has the advantage that while low frequency wavelets are broad in time , high frequency wavelets have a much narrower timeslot focus . thus , for example , with 32 data samples taken sequentially over time , a wavelet transformation generates a d . c . wavelet that spans the entire set of 32 points ; a first reversing wavelet that also spans the entire set of points ; two second harmonic wavelets that each focus upon only one - half of the 32 data points ; four third harmonic wavelets that each focus upon only one - fourth of the 32 data points ; eight fourth harmonic wavelets that each focus upon four of the 32 data points ; and 16 fifth harmonic wavelets that each focuses upon only two of the 32 data points . accordingly , the higher - frequency wavelet amplitudes are quite time specific , unlike fourier sinusoidal amplitudes , while the lower - frequency and d . c . wavelet amplitudes give one broad information about the whole set of 32 points . and like the fourier transform , the discrete wavelet transform preserves all of the information of the original atrial signal , such that the transformation is fully reversible . 32 discrete wavelet transform values may be reverse transformed back into 32 values indicating the atrial signal potentials at 32 sequential points in time . in essence , the haar function is performing multiple digital filtering operations , at variable positions and utilizing varying - width windowing functions , to develop component values that represent varying types of heartbeat activity , at varying frequencies , spread over varying widths , and located at varying positions . these component values can be studied to see which subset of these component values are good at distinguishing st from non - st , or ( more generally ) which subset of these component values are good at distinguishing one type of heartbeat waveform from another . further study of such a selected subset can further determine which of the subset of components are relatively invariant from one individual &# 39 ; s heart to another . one preferably selects component values that are suitable from both of these two perspectives . the particular wavelet transformation to choose can be tailored to the nature of the data , with the wavelets chosen to resemble somewhat the values to be found within the data so that some transformed values are of much larger amplitude than others , and such that the low amplitude transformed values may be disregarded . of particular importance with an atrial p waveform of the type being analyzed here are wavelets having frequency values roughly comparable to and timed to coincide with the ringing of the atrial depolarization . this tailoring can minimize the number of transformed data values that are significant and that are candidates for full participation in the correlation analyses to determine if there has been a substantial change in waveform morphology . another factor in selecting a particular wavelet transformation is reducing the number of computations that must be performed to carry out the transformation . yet another factor is selecting a wavelet transformation where some of the transformed component values vary from normal to abnormal heart rhythms in much the same over a population of individuals so that a generic template of these component values may be derived that can be used with many individuals , rather than with just one individual . these factors suggest that the haar transformation would be a suitable candidate for use in analysis of atrial waveforms and possibly other waveforms as well . with reference to fig3 in the case of 32 voltage values sampled over time , the 32 applicable haar discrete transform wavelets appear as shown in this figure . other discrete wavelet transforms based upon wavelets having triangular or other shapes may also prove usable . the haar transform wavelets , in particular , have the shape of a square waveform , as will be described , and this can simplify the computations required . in fig3 time increases from left to right . a scale 304 indicates the 32 points in time at which the atrial waveform is sampled , and an analog atrial waveform ( before digitization ) is shown at 302 . the p wave negative notch 303 is shown centered at the 16 th timeslot so that the signal potential of the atria waveform sampled at this point in time becomes the 16 th data value in the set of 32 data values that are to be subjected to the haar transformation . a haar wavelet is simply one cycle of a square waveform that starts at zero , then swings positive ( to \u201c+ 1 \u201d), then swings negative ( to \u201c\u2212 1 \u201d), and then swings back to zero . the wavelet w 2 , shown at 308 in fig3 for example , is at \u201c+ 1 \u201d at points in time 1 to 16 and is at \u201c\u2212 1 \u201d at points in time 17 to 32 . the shorter waveform w 6 , shown at 316 in fig3 is at \u201c+ 1 \u201d at points in time 9 to 12 and is at \u201c\u2212 1 \u201d at points in time 13 to 16 , and is at \u201c 0 \u201d at all other points in time . the waveform w 1 , shown at 306 in fig3 is so slow to fluctuate in time that its \u201c\u2212 1 \u201d portion is off of the chart ( in fig3 ) to the right , and is ignored ; and accordingly , it has the value \u201c+ 1 \u201d for all 32 of the points in time shown in fig3 . it thus represents the average , or d . c ., component of the atrial signal potential . the lowest frequency wavelets w 1 and w 2 encompass the entire set of 32 points in time . the higher frequency wavelets w 3 and w 4 each encompasses only half of the points in time , and the two wavelets taken together encompass all the points in time . the four wavelets w 5 , w 6 , w 7 , and w 8 each encompasses only eight points in time , and the four wavelets taken together encompass all points in time . the eight wavelets w 9 . . . w 16 each encompasses only four points in time , and together all eight wavelets encompass all points in time . and finally , the wavelets w 17 . . . w 32 each encompass only two points in time , but the sixteen wavelets together encompass all points in time . thus , each wavelet has a characteristic time span and position as well as a characteristic frequency , with higher - frequency wavelets having a narrower and more specific time span and position than lower - frequency wavelets . this is advantageous with an impulse - type , ringing signal such as the atrial depolarization waveform considered here , since many higher - frequency transform values that correspond to haar wavelets positioned in time away from the p waveform or that do not correspond to its frequency of ringing may be low in amplitude such that they may safely be ignored during the analysis steps , thereby reducing the data that must be processed during the correlation steps . ( the above , while presented in the context of the haar wavelet , is also applicable to other wavelet shapes that may used to perform a discrete wavelet transform ( dwt )). each of the 32 haar transformed values is computed as follows : multiply each of the 32 sampled and digitized voltage values representing the analog atrial waveform 302 by the correspondingly - positioned -( in - time ) amplitude values of one of the 32 haar wavelets ( shown in fig3 ); then sum the resulting products ; and then multiply the resulting sums by a discrete wavelet transform scaling factor 2 \u2212 j / 2 , where j equals 1 for the wavelets w 17 to w 32 , 2 for w 9 to w 16 , 3 for w 5 to w 8 , 4 for w 3 to w 4 , and 5 for w 1 and w 2 . for example , and referring to fig3 : the first wavelet w 1 is always + 1 , so the atrial signal potential values are simply summed and then multiplied by 2 \u2212 5 / 2 ; the second wavelet is + 1 for time points 1 to 16 and \u2212 1 for time points 17 to 32 , so the first 16 values are summed , the second 16 values are summed , the difference between the first and second sums is computed , and the result is multiplied by 2 \u2212 5 / 2 ; and so on until all the transformed values have been computed , one for each haar wavelet . the 32 time - sequential atrial signal potential values are thus transformed into 32 haar wavelet amplitude values which may be called \u201c transformed values \u201d and which may be assigned the numbers 1 to 32 corresponding to the subscript numbers of the haar wavelets to which they each correspond and whose amplitudes they represent . the above description of how to compute the haar wavelet transformed values is accurate , but it is not the most efficient way to proceed in the case of haar wavelets . like the fourier transform , which has a corresponding fast fourier transform that saves intermediate results and thereby avoids re - computing them and thus reduces substantially the number of computations , the haar transformation also may be carried out in a manner that saves intermediate sums and re - uses them to reduce substantially the number of computations . this is described below at the point where fig4 is described , since fig4 illustrates graphically how this can be done . the illustrative program listing presented below is also an implementation of the fast haar wavelet transformation algorithm . the haar transformation can readily be carried out by any digital computer . as an example of how the haar transformation can be carried out on 32 values , the following program , written in the language c , is illustrative of many possible programs that may be written . in the exemplary program that follows , a 32 - element array yy contains 32 data values representing the 32 sampled atrial signal potential values representing the fluctuations over time of the signal supplied by the bipolar lead 102 . the analog atrial signal is sampled , digitized , broken into separate beat data sets , pre - processed ( to remove waveforms distorted by ventricular activity ), centered ( with the negative depolarization spike at data point 16 ), and fed into the subroutine presented below . this subroutine is compiled ( or assembled , if rewritten in assembly language ) and installed in the rom of the icd &# 39 ; s embedded microprocessor . this subroutine returns , contained within the same array yy , the 32 transformed haar wavelet amplitude values described above . ( the constant value \u201c recipsqrttwo \u201d is the reciprocal of the square root of two ). an illustrative version of the subroutine for computing all 32 of the haar transformed values in an efficient manner is presented here : void haar ( double yy []) { int i , j , l ; double zz [ 32 ]; for ( i = 5 ; i & gt ; 0 ; i \u2212\u2212) { l = 1 ; for ( j = 1 ; j & lt ;= i ; ++ j ) l = 2 * l ; for ( j = 0 ; j & lt ; l ; ++ j ) zz [ j ] = yy [ j ]; for ( j = 0 ; j & lt ; l \u2212 1 ; j = j + 2 ) ( yy [ j / 2 ] = recipsqrttwo * ( zz [ j ] + zz [ j + 1 ]); yy [( j + l )/ 2 ] = rcipsqrttwo * ( zz [ j ] \u2212 zz [ j + 1 ]); } } return ; } the above program computes the 32 haar transformed values 1 to 32 from the 32 atrial signal potential time - sequenced input values . it does so with only 31 additions , 31 subtractions , and 62 multiplications . it is carefully designed to compute each value in a systematic way , making multiple use of intermediate results . first , the program computes sixteen sums of and sixteen differences between eight adjacent pairs of the 16 time sequenced atrial signal potential values , and the sixteen difference values become the haar transformed values 17 to 32 , reflecting the strength of the highest frequency haar wavelets positioned at 16 different positions in time , corresponding to the wavelets w 17 to w 32 shown in fig3 . all the sum and difference values are scaled by multiplication by 2 \u2212 1 / 2 . next , taking the 16 sums of atrial signal potential values as an intermediate result , the above algorithm generates eight sums of and eight differences between adjacent pairs of these sixteen intermediate values that resulted from the first sixteen additions , and the eight newly - computed difference values become the haar transformed values 9 to 16 , reflecting the strength of the second to the highest frequency values at eight different points in time , corresponding to wavelets w 9 to w 16 in fig3 . all of these eight sum and eight difference values are again scaled by 2 \u2212 1 / 2 so that the wavelets w 9 to w 16 are scaled by \u00bd ( 2 \u2212 2 / 2 or 2 \u2212 1 / 2 times 2 \u2212 1 / 2 ). next , taking these eight sums of sums of adjacent atrial signal potential values as an intermediate result , the above algorithm generates four sum and four difference values , again scaling by 2 \u2212 1 / 2 , and the four difference values become haar transformed values 5 to 8 , reflecting the strength of the middle frequency wavelets at four different points in time , and corresponding to wavelets w 5 to w 8 in fig3 . next , taking these four sums of sums of sums of adjacent atrial signal potential values , the above algorithm generates two sum and two difference values , again scaling by 2 \u2212 1 / 2 , and the two difference values become haar transformed values 3 and 4 , reflecting the strength of the second to the lowest frequency wavelets only two of which encompass all the time domain data , corresponding to the wavelets w 3 and w 4 shown in fig3 . and finally , the algorithm generates the sum of and the difference between the final remaining two sums of sums of sums of sums of atrial signal potential values , again scaling by 2 \u2212 1 / 2 . the difference value is then the haar transformed value 2 , representing the strength of the lowest - frequency wavelet , the one corresponding to the wavelet w 2 in fig3 the wavelet that extends the full length of the time scale . the sum value is then the first , or d . c ., transformed value , representing the average signal potential level over the 32 sampled points in time , which corresponds to the wavelet w 1 in fig3 . another way of viewing this computation is illustrated in fig4 a and 4b . the 32 atrial signal potential values are shown at the top of fig4 a and are identified as c 0 5 through c 31 5 . the superscript \u201c 5 \u201d indicates that these values are processed when the index value \u201c n \u201d in the above computer program is equal to \u201c 5 \u201d\u2014 that is , during the first pass through the data generating intermediary sums and differences . during subsequent passes , the value of n is decremented to 4 , 3 , 2 , and finally to 1 . during each pass through the data , the computer generates sums c n - 1 and differences d n - 1 , as shown in fig4 b , between adjacent pairs of the values c 0 n and c 1 n ; c 2 n and c 3 n ; and so on , so that the number of newly - generated c n - 1 and d n - 1 terms is reduced by half with each computer pass through the intermediary results . at the end of all these computations , the value c 0 0 is the first , or d . c ., haar transformed value ; and the values d 0 0 ; d 0 1 and d 1 1 ; d 0 2 , d 1 2 , d 2 2 , and d 3 2 ; d 0 3 , d 1 3 , . . . and d 7 3 ; and d 0 4 , d 1 4 , and d 15 4 are , respectively , the remaining haar transform values 2 through 32 . fig4 thus illustrates quite succinctly how all the transform computations are carried out , and how the intermediary \u201c c \u201d sum values , such as c 0 4 , are used to compute multiple haar values , such as the values d 0 3 , d 0 2 , d 0 1 , d 0 0 , and c 0 0 all of which are computed from the intermediary value c 0 4 . saving and reusing these intermediary \u201c c \u201d values saves much computational time . fig4 b indicates the precise addition and subtraction operations that are carried out at each level to compute the values in the next lower level , proceeding down through the chart presented in fig4 a . but in any given application to heart waveform analysis , all of these computations may not be needed , and accordingly the number of computations may be reduced much further . since the present invention teaches that only a small number of these haar transformed values need actually be considered , a far less computationally intensive transform can be developed which only generates the intermediate and final transformed values that are actually needed to generate the specific haar transformed values which have proved to be significant in discriminating between sinus tachycardia ( or st ) and non - st conditions . all others need not be computed , and the above program may be reduced to a special algorithm that omits as many sums , differences , and multiplications as possible . for example , if only the transformed values 1 , 5 , 9 , and 24 are significant , then 31 additions are required to compute the first transformed value ( the d . c . value \u2014 the simple sum of all the time domain signal values ); six additions and one subtraction are required to compute the fifth transformed value ( in fig3 the difference between the sums of time domain values under each half of the wavelet w 5 at 314 in fig3 ); two additions and one subtraction are required to compute the ninth transformed value ( in fig3 the difference between the sums of the time domain values under each half of the wavelet w9 at 322 in fig3 ); and only one subtraction is required to compute the 24 th transformed value ( in fig3 the difference between the two time domain values under the respective halves of the wavelet w 24 ( not shown ) which is the same size as , but differently positioned in time than , the wavelet w 18 at 330 ( fig3 ). accordingly , if only the transformed values 1 , 5 , 9 , and 24 actually evaluated , then only 42 additions and subtractions are required . but even this number can be reduced further . if the computational algorithm set forth in the above illustrative computer program is followed as a guide , then the intermediary sums used in computing the haar first , or d . c ., transformed value can be re - used to compute the sums for the haar transformed values 5 and 9 , assuming these intermediary results are saved in the manner described in the above program example . then 31 additions are still required to compute the first transformed value , but only one subtraction is required to compute each of the transformed values 5 , 9 , and 24 , giving a total of additions and subtraction of only 34 operations . and if only transformed values 1 , 5 , and 9 are computed , the number of additions and subtractions is reduced to 33 . and if the first transformed value is omitted and if transformed values 5 , 9 , and 24 are selected , then the 31 additions needed to compute the first transformed value are not required . then the number of computations for transformed value 5 is 8 additions and one subtraction ; the number of computations for transformed value 9 is 2 additions and one subtraction ; and the number of computations for transformed value 24 is still just one subtraction . so the total number of additions and subtractions is just 13 . but even this number can be reduced to 11 when it is realized that the two additions done for the transformed value 9 are also done ( in the above computational algorithm ) when computing the transformed value 5 . so the number of additions and subtractions can be reduced to 11 . also , when computing such a small number of transformed values , the number of multiplications can be reduced as well by postponing the multiplications until a transformed value is actually computed . with reference to fig4 a , instead of dividing every sum and difference by the square root of two ( as shown in fig4 b ), several vertical sums of \u201c c \u201d values in different rows of the fig4 a table can be formed , and then a \u201c d \u201d value can be computed by subtraction , and then the resulting unscaled transformed value can be scaled with a single multiplication by 2 \u2212 j / 2 where j is the number of rows in the table of fig4 a traversed by the one or more sum and the one difference computations . thus , the number of scaling multiplications can sometimes be equal to the number of transformed values that are computed , typically 3 or 4 . accordingly , by using wavelet analysis and transformation , instead of fourier or discrete cosine or other sinusoid analysis and transformation , a highly useful and highly frequency - specific result can be achieved with a very small number of computations , thereby conserving power and battery life , and yet achieving a high degree of precision in recognizing changes in morphology . in our tests , we have selected certain transformed values as being much more significant than other values in distinguishing between st and non - st . we focused upon those transform values whose \u201c+ 1 \u201d and \u201c\u2212 1 \u201d scope included the middle 8 time points most significant to analysis of the peak of the atrial depolarization . working across test data samples obtained from patients who had dual chamber icds , we computed the variance of each transformed value for st and non - st events using the standard statistical formula for computing variance . we then calculated the ratio of the variance of non - st events to the variance of st events , and selected those terms with the highest variance ratios for further consideration . in this manner , we reduced the number of transformed values that were included in the cwa or correlation process while always maintaining or improving the performance achieved . ultimately we settled upon the three or four transformed values having the highest ratios . these were the transform values 1 , 5 , 9 , and 24 . we tested 1 , 5 , and 24 together ; 5 , 9 , and 24 together ; and 1 , 5 , 9 , and 24 together . these selected haar transform values , when used in these combinations , required minimal computations and thereby produced a savings in battery power , and that also gave better performance at distinguishing st from non - st than did all of the transformed values used together . so an increase in accuracy was achieved as well as a decrease in computational complexity by this approach to atrial waveform analysis . [ 0054 ] fig5 for example , illustrates actual plots of digital information illustrating the effect of using only the three coefficients 1 , 5 , and 24 . at 502 , a normal waveform is shown , with amplitude plotted against sample numbers from 1 to 32 . at 504 , an abnormal waveform is shown , again with amplitude plotted against sample numbers . after the haar transformation , at 506 and 508 , the wavelet coefficient amplitude values are indicated for the coefficients 1 to 32 . at 506 , the coefficients generated by the normal waveform 502 are shown , and at 508 , the coefficients generated by the abnormal waveform 504 are shown . one may directly compare these component values and verify that the coefficients 1 , 5 , and 24 at 506 and at 508 vary significantly in amplitude . comparison of this same information among different patients ( not shown in this figure ) also indicated that this variation is relatively constant from one patient to another . at 510 and 512 , using only the coefficients 1 , 5 , and 24 , the heartbeat waveforms are reconstructed by a reverse transformation , the normal reconstructed waveform shown at 510 and the abnormal reconstructed waveform shown at 512 . the marked differences between these two reconstructed waveforms highlights the way in which these three coefficients can signal an abnormal condition with less computation . in a practical system , a small number of transform values are selected , in the manner just described . the algorithm for computing these values is then refined , as explained above , to reduce the number of additions , subtractions , and multiplications to the minimum possible while preserving the accuracy of the computations . we first compute a value \u03c1 , which is the correlation between these values with the corresponding values in a template that represents the values for an average normal population . we then compare this value \u03c1 to a threshold correlation value \u03b2 that is chosen to give optimal results on experimental data . ( see the examples in the tables presented below .) for each waveform analyzed , a test is made to determine whether \u03c1 is greater than \u03b2 . if so , then this waveform is placed into the buffer marked \u201c st \u201d at step 210 . if not , then this waveform is placed into the buffer marked \u201c non - st \u201d at step 211 . finally , after ten waveforms have been analyzed , a count of st and non - st marked waveforms is made to see if the count of non - st waveforms is greater than some threshold value x ( at step 212 ), where x can be , for example , 7 . if more than seven waveforms are non - st , then therapy is delivered at step 214 . the buffers at steps 210 and 211 in fig2 are buffers that hold the last ten decision values . these buffers can be pictured as sliding windows revealing the most recent ten values to permit the decision at 212 to be continuously updated . the buffers thus function as a nonlinear filter preventing irregular values from delivering therapy improperly . in developing a working prototype of this system , we used a development data set consisting of 20 episodes of st taken from 4 patients and 18 episodes of af taken from 7 patients . patient number episodes of st episodes of af 1 2 6 2 12 0 13 0 1 4 0 2 5 5 1 6 0 3 7 0 3 8 0 2 9 1 0 total 20 18 we used this data to optimize our choice of \u03c1 and \u03b2 and also the particular coefficients that we chose to examine . next , we tested our prototype using a set of 16 episodes of st obtained from 4 patients together with 17 episodes of af obtained from 7 patients . patient number episodes of st episodes of af 1 8 0 2 3 1 3 3 0 4 2 0 5 0 4 6 0 4 7 0 3 8 0 3 9 0 1 10 0 1 total 16 17 x out threshold coefficients of 10 ( \u03b2 ) sensitivity specificity 1 , 5 , 9 , & amp ; 24 3 0 . 808 76 % 94 % 1 , 5 , 9 , & amp ; 24 5 0 . 975 88 % 94 % 1 , 5 , 9 , & amp ; 24 6 0 . 976 82 % 94 % 1 , 5 , 9 , & amp ; 24 7 0 . 983 88 % 94 % 1 , 5 , & amp ; 24 3 0 . 815 82 % 94 % 1 , 5 , & amp ; 24 4 0 . 943 88 % 94 % 1 , 5 , & amp ; 24 5 0 . 956 82 % 94 % 1 , 5 , & amp ; 24 7 0 . 991 94 % 94 % 5 , 9 , & amp ; 24 8 0 . 9993 82 % 94 % 5 , 9 , & amp ; 24 9 0 . 9997 76 % 81 % the quality of the selected subset of transformed values for characterizing changes in morphology may be demonstrated graphically , as indicated in fig5 by performing a reverse haar transform using only the three or four or so selected transformed values . as can be seen ( at 510 and 512 in fig5 ), the inverse transformations produce distorted representations of the original atrial waveforms which illustrate how the choice of transform values can emphasize the changes . this is another useful way to assist one in selecting which transformed values are most useful . one feature of the invention is its ability to use a template derived from an average normal population , rather than deriving a patient customized average normal template for each individual patient . prior systems have required each new patient to be monitored while in a normal state , and the captured data was then averaged to form a patient specific normal template . contrary to this , the present invention achieved the results shown above using the same template for all patients . accordingly , the invention may be used with a new patient without the necessity of such preliminary testing of each patient and without a new customized template necessarily having to be created for each patient . one reason why the present invention can function using a template that represents average values for a normal population is because the specific coefficients selected in the transform domain , in addition to having been selected to emphasize the difference between normal and abnormal values , are also preferably chosen to minimize the differences between different patients . in some cases , values that were good at distinguishing between normal and abnormal rhythms for one patient were not as good at doing so for some other patient . these values are preferably not selected for use in implementing the present invention . for example , the transform coefficients may be selected such that all ( or most of ) the normal values of a particular coefficient gathered from many patients had the same sign ( positive or negative ) and similar amplitudes ( or if they varied in sign , they were near to zero ). in addition , the abnormal coefficients are significantly different in value from the normal coefficients for each particular patient . [ 0068 ] fig6 at 600 , illustrates one way in which this may be done . first , at step 602 , one gathers a large number of st and non - st data samples . next , all of this data is transformed , generating coefficients for every waveform set of data ( step 604 ). the variance of each coefficient is then computed first for the st samples ( step 606 ) and then for the non - st samples . then , for each coefficient , the ratio of the variance of the non - st samples to that of the st samples is computed ( step 608 ). finally , those coefficients whose variance ratio was large may be selected a coefficients for use in creating a population template and in testing patients ( step 612 ). in addition , at step 614 , the number of coefficients retained may be further reduced to reduce the number of computations that need to be performed , as explained above . in the version of the invention that was used to generate the data shown above , the population template was computed as follows , using the 20 st episodes in the development data set ( also used in the first of the two tables presented above ). this is illustrated at 700 in fig7 . first , the coefficients were selected as was explained above and as shown in fig6 . next , for each patient episode of st ( step 702 ), the recorded heartbeats were broken up into groups of ten consecutive heartbeats ( step 704 ). for each group ( step 706 ), the selected coefficients of the haar transform were computed ( step 708 ) for each of the ten heartbeats and the median value was then selected ( step 710 ) for each coefficient to form a proto - template . if these median values correlated well with the coefficient values for each of the ten heartbeats , the proto - template was retained ( step 714 ). otherwise , it was discarded . in this manner , a whole bunch of proto - templates were obtained from each patient episode . next , median values of the coefficients from all of the proto - templates ( step 716 ) were computed ( step 718 ). these values were then used as the population template for distinguishing st from non - st events ( step 720 ). while the preferred embodiment of the invention has been described above , it is to be understood that numerous modifications and changes will occur to those who are skilled in the art to which the invention pertains . accordingly , the following claims annexed to and forming a part of this specification are intended to define the true spirit and scope of the invention \u2014 that is , what is new and what is desired to be secured by letters patent of the united states ."}
|
{"category": "Chemistry; Metallurgy", "patent": "the present invention is directed towards improving the quality and functionality of both implantable cardioverter defibrillators ( icds ) and implantable atrial defibrillators ( iads ). all such devices need to have a method and mechanism for accurately distinguishing sinus tachycardia ( st ) form non - st , such as atrial fibrillation and atrial flutter . this method and mechanism should be conservative , biased towards not initiating therapy unless non - st is clearly indicated . it should also require as few machine computational cycles as possible to reduce icd and iad battery drain . the principle underlying the present invention is that st and non - st can be distinguished by studying and measuring the atrial electrogram morphology to detect aberrant conduction in the atria . thus , by comparing normal st waveforms to a series of newly - occurring waveforms , st is indicated by abnormal atrial waveform morphology which signals aberrant conduction within the atria improper electrical patterns of depolarization . this will show up in the signal potentials captured by a bipolar atrial lead . with reference to fig1 the present invention can be implemented in an icd 100 ( which could be an iad ) that is implanted in the chest of a patient who suffers from unpleasant and possibly life - threatening irregularities in heart operation due to improper electrical stimulation of the heart muscle cells . during normal heart action , the heart &# 39 ; s electrical impulses originate in the sino - atrial node as an action potential that is transmitted smoothly to all portions of the atria , causing contraction of the atrial chambers . the electrical impulse continues in its path to a cluster of conduction fibrils known as the atrioventricular node . after a delay of about one - tenth of a second , an action potential flows over the ventricles and causes them to contract in synchronism with and following shortly after the atrial contraction . in this manner , the heart pumps blood to the lungs and from the lungs to the body . a bipolar lead 102 is implanted within the atria 104 of a heart 103 to measure the electrical potential between nearby cells of the atria . as the action potential actively passes from cell to cell past the two electrodes of the lead 102 , an oscillating signal potential is developed across the two electrodes and is conveyed by the bipolar lead 102 to the icd 100 where the signal is sampled about 1000 times per second and is digitized , with the sequential samples stored within a ram memory within the icd 100 for further analysis . another bipolar lead 105 may extend to the ventricle ( not shown ) of the heart 103 to measure the action potentials generated within the ventricles . these same leads , or other leads , may be used by the icd for administering various types of therapy , such as pacing or defibrillation shock therapy . the data samples collected from within the atria are now analyzed . first , with reference to data gathered from the ventricles , if some event in the ventricle , such as a mis - timed depolarization event that partially overlaps the atria &# 39 ; s p wave , could leak across to the atria and distort the measurement of the potentials for a given atrial depolarization , then the data for that particular heartbeat is discarded and is not used in further analyses . such distortion can also be found simply because a given heartbeat set of data gathered from the atria is itself badly distorted , and this approach must be taken in the case of an iad having no ventricular lead . the remaining data is retained for further processing . in preparation for further processing , the data for individual heartbeats is identified and gathered . the position within the gathered data of the negative spike to the atrial p depolarization waveform is determined and is selected as the center of the waveform for each beat . then , since data close to this p - wave is to be analyzed , and since data further away may be distorted to some degree by ventricular activity or by variations in the p - to - p interval , 32 sequential data samples are selected for each heartbeat such that the negative spike of the p - wave becomes signal potential sample number 16 of the 32 sequential samples selected for further analyses . these samples for a normal heartbeat appear at 302 in fig3 and at 502 in fig5 where the amplitude of the atrial signal potential is plotted against time over the 32 sampled values , with the negative p - wave spike positioned as signal sample number 16 . likewise , the samples for an abnormal heartbeat appear at 504 in fig5 . straightforward correlation ( cwa ) of the 32 samples against a template containing a typical or normal pattern could be performed at this juncture , as taught on page 563 ( equation ( 1 )) of the article by thorne , cited in the introductory portion of this specification . but the performance of such a cross correlation at repeated regular intervals would generate numerous computations and would drain the icd &# 39 ; s battery more rapidly than is desirable . in addition , provision would have to be made for a full template waveform that can be stored within the icd to be used as a comparison reference . in addition , such templates generally need to be patient specific , and they are more susceptible to normal changes in shape . accordingly , to reduce the mathematical complexity of the cross correlation computations , it is desirable to reduce the number of data points from 32 down to some much lower number through the use of some form of transformation into a different set of values . for example , the data could be transformed using the karhunen - loeve transformation described in the morris article cited at the beginning of this specification . but that transformation is fairly computationally intensive . computationally , a fast fourier transformation would be more efficient , but any such transformation into the frequency domain , where variable frequency sine and cosine wave amplitudes result from the transformation , does not match itself well to the morphology of atrial heart waveforms , which tend to be formed from large , ringing , spike - like representations of p - wave depolarization events travelling as moving action potential fronts past the pair of electrodes attached to the bipolar lead , and not as steady harmonics extending across time . accordingly , fourier - class transforms do not adequately reduce the number of significant data values that must be considered . and in addition , when the \u201c window \u201d size for a fourier analysis is selected , if it is wide , then the frequency values are not time - specific , but represent average signal harmonic content over the entire window time duration . and if the window is narrow , then the harmonics are too spread out in frequency , and the frequency - domain data generated by the transform does not resolve things finely enough in the frequency domain . for all of the above reasons , the present invention analyzes the waveforms using a discrete wavelet transform . this has the advantage that while low frequency wavelets are broad in time , high frequency wavelets have a much narrower timeslot focus . thus , for example , with 32 data samples taken sequentially over time , a wavelet transformation generates a d . c . wavelet that spans the entire set of 32 points ; a first reversing wavelet that also spans the entire set of points ; two second harmonic wavelets that each focus upon only one - half of the 32 data points ; four third harmonic wavelets that each focus upon only one - fourth of the 32 data points ; eight fourth harmonic wavelets that each focus upon four of the 32 data points ; and 16 fifth harmonic wavelets that each focuses upon only two of the 32 data points . accordingly , the higher - frequency wavelet amplitudes are quite time specific , unlike fourier sinusoidal amplitudes , while the lower - frequency and d . c . wavelet amplitudes give one broad information about the whole set of 32 points . and like the fourier transform , the discrete wavelet transform preserves all of the information of the original atrial signal , such that the transformation is fully reversible . 32 discrete wavelet transform values may be reverse transformed back into 32 values indicating the atrial signal potentials at 32 sequential points in time . in essence , the haar function is performing multiple digital filtering operations , at variable positions and utilizing varying - width windowing functions , to develop component values that represent varying types of heartbeat activity , at varying frequencies , spread over varying widths , and located at varying positions . these component values can be studied to see which subset of these component values are good at distinguishing st from non - st , or ( more generally ) which subset of these component values are good at distinguishing one type of heartbeat waveform from another . further study of such a selected subset can further determine which of the subset of components are relatively invariant from one individual &# 39 ; s heart to another . one preferably selects component values that are suitable from both of these two perspectives . the particular wavelet transformation to choose can be tailored to the nature of the data , with the wavelets chosen to resemble somewhat the values to be found within the data so that some transformed values are of much larger amplitude than others , and such that the low amplitude transformed values may be disregarded . of particular importance with an atrial p waveform of the type being analyzed here are wavelets having frequency values roughly comparable to and timed to coincide with the ringing of the atrial depolarization . this tailoring can minimize the number of transformed data values that are significant and that are candidates for full participation in the correlation analyses to determine if there has been a substantial change in waveform morphology . another factor in selecting a particular wavelet transformation is reducing the number of computations that must be performed to carry out the transformation . yet another factor is selecting a wavelet transformation where some of the transformed component values vary from normal to abnormal heart rhythms in much the same over a population of individuals so that a generic template of these component values may be derived that can be used with many individuals , rather than with just one individual . these factors suggest that the haar transformation would be a suitable candidate for use in analysis of atrial waveforms and possibly other waveforms as well . with reference to fig3 in the case of 32 voltage values sampled over time , the 32 applicable haar discrete transform wavelets appear as shown in this figure . other discrete wavelet transforms based upon wavelets having triangular or other shapes may also prove usable . the haar transform wavelets , in particular , have the shape of a square waveform , as will be described , and this can simplify the computations required . in fig3 time increases from left to right . a scale 304 indicates the 32 points in time at which the atrial waveform is sampled , and an analog atrial waveform ( before digitization ) is shown at 302 . the p wave negative notch 303 is shown centered at the 16 th timeslot so that the signal potential of the atria waveform sampled at this point in time becomes the 16 th data value in the set of 32 data values that are to be subjected to the haar transformation . a haar wavelet is simply one cycle of a square waveform that starts at zero , then swings positive ( to \u201c+ 1 \u201d), then swings negative ( to \u201c\u2212 1 \u201d), and then swings back to zero . the wavelet w 2 , shown at 308 in fig3 for example , is at \u201c+ 1 \u201d at points in time 1 to 16 and is at \u201c\u2212 1 \u201d at points in time 17 to 32 . the shorter waveform w 6 , shown at 316 in fig3 is at \u201c+ 1 \u201d at points in time 9 to 12 and is at \u201c\u2212 1 \u201d at points in time 13 to 16 , and is at \u201c 0 \u201d at all other points in time . the waveform w 1 , shown at 306 in fig3 is so slow to fluctuate in time that its \u201c\u2212 1 \u201d portion is off of the chart ( in fig3 ) to the right , and is ignored ; and accordingly , it has the value \u201c+ 1 \u201d for all 32 of the points in time shown in fig3 . it thus represents the average , or d . c ., component of the atrial signal potential . the lowest frequency wavelets w 1 and w 2 encompass the entire set of 32 points in time . the higher frequency wavelets w 3 and w 4 each encompasses only half of the points in time , and the two wavelets taken together encompass all the points in time . the four wavelets w 5 , w 6 , w 7 , and w 8 each encompasses only eight points in time , and the four wavelets taken together encompass all points in time . the eight wavelets w 9 . . . w 16 each encompasses only four points in time , and together all eight wavelets encompass all points in time . and finally , the wavelets w 17 . . . w 32 each encompass only two points in time , but the sixteen wavelets together encompass all points in time . thus , each wavelet has a characteristic time span and position as well as a characteristic frequency , with higher - frequency wavelets having a narrower and more specific time span and position than lower - frequency wavelets . this is advantageous with an impulse - type , ringing signal such as the atrial depolarization waveform considered here , since many higher - frequency transform values that correspond to haar wavelets positioned in time away from the p waveform or that do not correspond to its frequency of ringing may be low in amplitude such that they may safely be ignored during the analysis steps , thereby reducing the data that must be processed during the correlation steps . ( the above , while presented in the context of the haar wavelet , is also applicable to other wavelet shapes that may used to perform a discrete wavelet transform ( dwt )). each of the 32 haar transformed values is computed as follows : multiply each of the 32 sampled and digitized voltage values representing the analog atrial waveform 302 by the correspondingly - positioned -( in - time ) amplitude values of one of the 32 haar wavelets ( shown in fig3 ); then sum the resulting products ; and then multiply the resulting sums by a discrete wavelet transform scaling factor 2 \u2212 j / 2 , where j equals 1 for the wavelets w 17 to w 32 , 2 for w 9 to w 16 , 3 for w 5 to w 8 , 4 for w 3 to w 4 , and 5 for w 1 and w 2 . for example , and referring to fig3 : the first wavelet w 1 is always + 1 , so the atrial signal potential values are simply summed and then multiplied by 2 \u2212 5 / 2 ; the second wavelet is + 1 for time points 1 to 16 and \u2212 1 for time points 17 to 32 , so the first 16 values are summed , the second 16 values are summed , the difference between the first and second sums is computed , and the result is multiplied by 2 \u2212 5 / 2 ; and so on until all the transformed values have been computed , one for each haar wavelet . the 32 time - sequential atrial signal potential values are thus transformed into 32 haar wavelet amplitude values which may be called \u201c transformed values \u201d and which may be assigned the numbers 1 to 32 corresponding to the subscript numbers of the haar wavelets to which they each correspond and whose amplitudes they represent . the above description of how to compute the haar wavelet transformed values is accurate , but it is not the most efficient way to proceed in the case of haar wavelets . like the fourier transform , which has a corresponding fast fourier transform that saves intermediate results and thereby avoids re - computing them and thus reduces substantially the number of computations , the haar transformation also may be carried out in a manner that saves intermediate sums and re - uses them to reduce substantially the number of computations . this is described below at the point where fig4 is described , since fig4 illustrates graphically how this can be done . the illustrative program listing presented below is also an implementation of the fast haar wavelet transformation algorithm . the haar transformation can readily be carried out by any digital computer . as an example of how the haar transformation can be carried out on 32 values , the following program , written in the language c , is illustrative of many possible programs that may be written . in the exemplary program that follows , a 32 - element array yy contains 32 data values representing the 32 sampled atrial signal potential values representing the fluctuations over time of the signal supplied by the bipolar lead 102 . the analog atrial signal is sampled , digitized , broken into separate beat data sets , pre - processed ( to remove waveforms distorted by ventricular activity ), centered ( with the negative depolarization spike at data point 16 ), and fed into the subroutine presented below . this subroutine is compiled ( or assembled , if rewritten in assembly language ) and installed in the rom of the icd &# 39 ; s embedded microprocessor . this subroutine returns , contained within the same array yy , the 32 transformed haar wavelet amplitude values described above . ( the constant value \u201c recipsqrttwo \u201d is the reciprocal of the square root of two ). an illustrative version of the subroutine for computing all 32 of the haar transformed values in an efficient manner is presented here : void haar ( double yy []) { int i , j , l ; double zz [ 32 ]; for ( i = 5 ; i & gt ; 0 ; i \u2212\u2212) { l = 1 ; for ( j = 1 ; j & lt ;= i ; ++ j ) l = 2 * l ; for ( j = 0 ; j & lt ; l ; ++ j ) zz [ j ] = yy [ j ]; for ( j = 0 ; j & lt ; l \u2212 1 ; j = j + 2 ) ( yy [ j / 2 ] = recipsqrttwo * ( zz [ j ] + zz [ j + 1 ]); yy [( j + l )/ 2 ] = rcipsqrttwo * ( zz [ j ] \u2212 zz [ j + 1 ]); } } return ; } the above program computes the 32 haar transformed values 1 to 32 from the 32 atrial signal potential time - sequenced input values . it does so with only 31 additions , 31 subtractions , and 62 multiplications . it is carefully designed to compute each value in a systematic way , making multiple use of intermediate results . first , the program computes sixteen sums of and sixteen differences between eight adjacent pairs of the 16 time sequenced atrial signal potential values , and the sixteen difference values become the haar transformed values 17 to 32 , reflecting the strength of the highest frequency haar wavelets positioned at 16 different positions in time , corresponding to the wavelets w 17 to w 32 shown in fig3 . all the sum and difference values are scaled by multiplication by 2 \u2212 1 / 2 . next , taking the 16 sums of atrial signal potential values as an intermediate result , the above algorithm generates eight sums of and eight differences between adjacent pairs of these sixteen intermediate values that resulted from the first sixteen additions , and the eight newly - computed difference values become the haar transformed values 9 to 16 , reflecting the strength of the second to the highest frequency values at eight different points in time , corresponding to wavelets w 9 to w 16 in fig3 . all of these eight sum and eight difference values are again scaled by 2 \u2212 1 / 2 so that the wavelets w 9 to w 16 are scaled by \u00bd ( 2 \u2212 2 / 2 or 2 \u2212 1 / 2 times 2 \u2212 1 / 2 ). next , taking these eight sums of sums of adjacent atrial signal potential values as an intermediate result , the above algorithm generates four sum and four difference values , again scaling by 2 \u2212 1 / 2 , and the four difference values become haar transformed values 5 to 8 , reflecting the strength of the middle frequency wavelets at four different points in time , and corresponding to wavelets w 5 to w 8 in fig3 . next , taking these four sums of sums of sums of adjacent atrial signal potential values , the above algorithm generates two sum and two difference values , again scaling by 2 \u2212 1 / 2 , and the two difference values become haar transformed values 3 and 4 , reflecting the strength of the second to the lowest frequency wavelets only two of which encompass all the time domain data , corresponding to the wavelets w 3 and w 4 shown in fig3 . and finally , the algorithm generates the sum of and the difference between the final remaining two sums of sums of sums of sums of atrial signal potential values , again scaling by 2 \u2212 1 / 2 . the difference value is then the haar transformed value 2 , representing the strength of the lowest - frequency wavelet , the one corresponding to the wavelet w 2 in fig3 the wavelet that extends the full length of the time scale . the sum value is then the first , or d . c ., transformed value , representing the average signal potential level over the 32 sampled points in time , which corresponds to the wavelet w 1 in fig3 . another way of viewing this computation is illustrated in fig4 a and 4b . the 32 atrial signal potential values are shown at the top of fig4 a and are identified as c 0 5 through c 31 5 . the superscript \u201c 5 \u201d indicates that these values are processed when the index value \u201c n \u201d in the above computer program is equal to \u201c 5 \u201d\u2014 that is , during the first pass through the data generating intermediary sums and differences . during subsequent passes , the value of n is decremented to 4 , 3 , 2 , and finally to 1 . during each pass through the data , the computer generates sums c n - 1 and differences d n - 1 , as shown in fig4 b , between adjacent pairs of the values c 0 n and c 1 n ; c 2 n and c 3 n ; and so on , so that the number of newly - generated c n - 1 and d n - 1 terms is reduced by half with each computer pass through the intermediary results . at the end of all these computations , the value c 0 0 is the first , or d . c ., haar transformed value ; and the values d 0 0 ; d 0 1 and d 1 1 ; d 0 2 , d 1 2 , d 2 2 , and d 3 2 ; d 0 3 , d 1 3 , . . . and d 7 3 ; and d 0 4 , d 1 4 , and d 15 4 are , respectively , the remaining haar transform values 2 through 32 . fig4 thus illustrates quite succinctly how all the transform computations are carried out , and how the intermediary \u201c c \u201d sum values , such as c 0 4 , are used to compute multiple haar values , such as the values d 0 3 , d 0 2 , d 0 1 , d 0 0 , and c 0 0 all of which are computed from the intermediary value c 0 4 . saving and reusing these intermediary \u201c c \u201d values saves much computational time . fig4 b indicates the precise addition and subtraction operations that are carried out at each level to compute the values in the next lower level , proceeding down through the chart presented in fig4 a . but in any given application to heart waveform analysis , all of these computations may not be needed , and accordingly the number of computations may be reduced much further . since the present invention teaches that only a small number of these haar transformed values need actually be considered , a far less computationally intensive transform can be developed which only generates the intermediate and final transformed values that are actually needed to generate the specific haar transformed values which have proved to be significant in discriminating between sinus tachycardia ( or st ) and non - st conditions . all others need not be computed , and the above program may be reduced to a special algorithm that omits as many sums , differences , and multiplications as possible . for example , if only the transformed values 1 , 5 , 9 , and 24 are significant , then 31 additions are required to compute the first transformed value ( the d . c . value \u2014 the simple sum of all the time domain signal values ); six additions and one subtraction are required to compute the fifth transformed value ( in fig3 the difference between the sums of time domain values under each half of the wavelet w 5 at 314 in fig3 ); two additions and one subtraction are required to compute the ninth transformed value ( in fig3 the difference between the sums of the time domain values under each half of the wavelet w9 at 322 in fig3 ); and only one subtraction is required to compute the 24 th transformed value ( in fig3 the difference between the two time domain values under the respective halves of the wavelet w 24 ( not shown ) which is the same size as , but differently positioned in time than , the wavelet w 18 at 330 ( fig3 ). accordingly , if only the transformed values 1 , 5 , 9 , and 24 actually evaluated , then only 42 additions and subtractions are required . but even this number can be reduced further . if the computational algorithm set forth in the above illustrative computer program is followed as a guide , then the intermediary sums used in computing the haar first , or d . c ., transformed value can be re - used to compute the sums for the haar transformed values 5 and 9 , assuming these intermediary results are saved in the manner described in the above program example . then 31 additions are still required to compute the first transformed value , but only one subtraction is required to compute each of the transformed values 5 , 9 , and 24 , giving a total of additions and subtraction of only 34 operations . and if only transformed values 1 , 5 , and 9 are computed , the number of additions and subtractions is reduced to 33 . and if the first transformed value is omitted and if transformed values 5 , 9 , and 24 are selected , then the 31 additions needed to compute the first transformed value are not required . then the number of computations for transformed value 5 is 8 additions and one subtraction ; the number of computations for transformed value 9 is 2 additions and one subtraction ; and the number of computations for transformed value 24 is still just one subtraction . so the total number of additions and subtractions is just 13 . but even this number can be reduced to 11 when it is realized that the two additions done for the transformed value 9 are also done ( in the above computational algorithm ) when computing the transformed value 5 . so the number of additions and subtractions can be reduced to 11 . also , when computing such a small number of transformed values , the number of multiplications can be reduced as well by postponing the multiplications until a transformed value is actually computed . with reference to fig4 a , instead of dividing every sum and difference by the square root of two ( as shown in fig4 b ), several vertical sums of \u201c c \u201d values in different rows of the fig4 a table can be formed , and then a \u201c d \u201d value can be computed by subtraction , and then the resulting unscaled transformed value can be scaled with a single multiplication by 2 \u2212 j / 2 where j is the number of rows in the table of fig4 a traversed by the one or more sum and the one difference computations . thus , the number of scaling multiplications can sometimes be equal to the number of transformed values that are computed , typically 3 or 4 . accordingly , by using wavelet analysis and transformation , instead of fourier or discrete cosine or other sinusoid analysis and transformation , a highly useful and highly frequency - specific result can be achieved with a very small number of computations , thereby conserving power and battery life , and yet achieving a high degree of precision in recognizing changes in morphology . in our tests , we have selected certain transformed values as being much more significant than other values in distinguishing between st and non - st . we focused upon those transform values whose \u201c+ 1 \u201d and \u201c\u2212 1 \u201d scope included the middle 8 time points most significant to analysis of the peak of the atrial depolarization . working across test data samples obtained from patients who had dual chamber icds , we computed the variance of each transformed value for st and non - st events using the standard statistical formula for computing variance . we then calculated the ratio of the variance of non - st events to the variance of st events , and selected those terms with the highest variance ratios for further consideration . in this manner , we reduced the number of transformed values that were included in the cwa or correlation process while always maintaining or improving the performance achieved . ultimately we settled upon the three or four transformed values having the highest ratios . these were the transform values 1 , 5 , 9 , and 24 . we tested 1 , 5 , and 24 together ; 5 , 9 , and 24 together ; and 1 , 5 , 9 , and 24 together . these selected haar transform values , when used in these combinations , required minimal computations and thereby produced a savings in battery power , and that also gave better performance at distinguishing st from non - st than did all of the transformed values used together . so an increase in accuracy was achieved as well as a decrease in computational complexity by this approach to atrial waveform analysis . [ 0054 ] fig5 for example , illustrates actual plots of digital information illustrating the effect of using only the three coefficients 1 , 5 , and 24 . at 502 , a normal waveform is shown , with amplitude plotted against sample numbers from 1 to 32 . at 504 , an abnormal waveform is shown , again with amplitude plotted against sample numbers . after the haar transformation , at 506 and 508 , the wavelet coefficient amplitude values are indicated for the coefficients 1 to 32 . at 506 , the coefficients generated by the normal waveform 502 are shown , and at 508 , the coefficients generated by the abnormal waveform 504 are shown . one may directly compare these component values and verify that the coefficients 1 , 5 , and 24 at 506 and at 508 vary significantly in amplitude . comparison of this same information among different patients ( not shown in this figure ) also indicated that this variation is relatively constant from one patient to another . at 510 and 512 , using only the coefficients 1 , 5 , and 24 , the heartbeat waveforms are reconstructed by a reverse transformation , the normal reconstructed waveform shown at 510 and the abnormal reconstructed waveform shown at 512 . the marked differences between these two reconstructed waveforms highlights the way in which these three coefficients can signal an abnormal condition with less computation . in a practical system , a small number of transform values are selected , in the manner just described . the algorithm for computing these values is then refined , as explained above , to reduce the number of additions , subtractions , and multiplications to the minimum possible while preserving the accuracy of the computations . we first compute a value \u03c1 , which is the correlation between these values with the corresponding values in a template that represents the values for an average normal population . we then compare this value \u03c1 to a threshold correlation value \u03b2 that is chosen to give optimal results on experimental data . ( see the examples in the tables presented below .) for each waveform analyzed , a test is made to determine whether \u03c1 is greater than \u03b2 . if so , then this waveform is placed into the buffer marked \u201c st \u201d at step 210 . if not , then this waveform is placed into the buffer marked \u201c non - st \u201d at step 211 . finally , after ten waveforms have been analyzed , a count of st and non - st marked waveforms is made to see if the count of non - st waveforms is greater than some threshold value x ( at step 212 ), where x can be , for example , 7 . if more than seven waveforms are non - st , then therapy is delivered at step 214 . the buffers at steps 210 and 211 in fig2 are buffers that hold the last ten decision values . these buffers can be pictured as sliding windows revealing the most recent ten values to permit the decision at 212 to be continuously updated . the buffers thus function as a nonlinear filter preventing irregular values from delivering therapy improperly . in developing a working prototype of this system , we used a development data set consisting of 20 episodes of st taken from 4 patients and 18 episodes of af taken from 7 patients . patient number episodes of st episodes of af 1 2 6 2 12 0 13 0 1 4 0 2 5 5 1 6 0 3 7 0 3 8 0 2 9 1 0 total 20 18 we used this data to optimize our choice of \u03c1 and \u03b2 and also the particular coefficients that we chose to examine . next , we tested our prototype using a set of 16 episodes of st obtained from 4 patients together with 17 episodes of af obtained from 7 patients . patient number episodes of st episodes of af 1 8 0 2 3 1 3 3 0 4 2 0 5 0 4 6 0 4 7 0 3 8 0 3 9 0 1 10 0 1 total 16 17 x out threshold coefficients of 10 ( \u03b2 ) sensitivity specificity 1 , 5 , 9 , & amp ; 24 3 0 . 808 76 % 94 % 1 , 5 , 9 , & amp ; 24 5 0 . 975 88 % 94 % 1 , 5 , 9 , & amp ; 24 6 0 . 976 82 % 94 % 1 , 5 , 9 , & amp ; 24 7 0 . 983 88 % 94 % 1 , 5 , & amp ; 24 3 0 . 815 82 % 94 % 1 , 5 , & amp ; 24 4 0 . 943 88 % 94 % 1 , 5 , & amp ; 24 5 0 . 956 82 % 94 % 1 , 5 , & amp ; 24 7 0 . 991 94 % 94 % 5 , 9 , & amp ; 24 8 0 . 9993 82 % 94 % 5 , 9 , & amp ; 24 9 0 . 9997 76 % 81 % the quality of the selected subset of transformed values for characterizing changes in morphology may be demonstrated graphically , as indicated in fig5 by performing a reverse haar transform using only the three or four or so selected transformed values . as can be seen ( at 510 and 512 in fig5 ), the inverse transformations produce distorted representations of the original atrial waveforms which illustrate how the choice of transform values can emphasize the changes . this is another useful way to assist one in selecting which transformed values are most useful . one feature of the invention is its ability to use a template derived from an average normal population , rather than deriving a patient customized average normal template for each individual patient . prior systems have required each new patient to be monitored while in a normal state , and the captured data was then averaged to form a patient specific normal template . contrary to this , the present invention achieved the results shown above using the same template for all patients . accordingly , the invention may be used with a new patient without the necessity of such preliminary testing of each patient and without a new customized template necessarily having to be created for each patient . one reason why the present invention can function using a template that represents average values for a normal population is because the specific coefficients selected in the transform domain , in addition to having been selected to emphasize the difference between normal and abnormal values , are also preferably chosen to minimize the differences between different patients . in some cases , values that were good at distinguishing between normal and abnormal rhythms for one patient were not as good at doing so for some other patient . these values are preferably not selected for use in implementing the present invention . for example , the transform coefficients may be selected such that all ( or most of ) the normal values of a particular coefficient gathered from many patients had the same sign ( positive or negative ) and similar amplitudes ( or if they varied in sign , they were near to zero ). in addition , the abnormal coefficients are significantly different in value from the normal coefficients for each particular patient . [ 0068 ] fig6 at 600 , illustrates one way in which this may be done . first , at step 602 , one gathers a large number of st and non - st data samples . next , all of this data is transformed , generating coefficients for every waveform set of data ( step 604 ). the variance of each coefficient is then computed first for the st samples ( step 606 ) and then for the non - st samples . then , for each coefficient , the ratio of the variance of the non - st samples to that of the st samples is computed ( step 608 ). finally , those coefficients whose variance ratio was large may be selected a coefficients for use in creating a population template and in testing patients ( step 612 ). in addition , at step 614 , the number of coefficients retained may be further reduced to reduce the number of computations that need to be performed , as explained above . in the version of the invention that was used to generate the data shown above , the population template was computed as follows , using the 20 st episodes in the development data set ( also used in the first of the two tables presented above ). this is illustrated at 700 in fig7 . first , the coefficients were selected as was explained above and as shown in fig6 . next , for each patient episode of st ( step 702 ), the recorded heartbeats were broken up into groups of ten consecutive heartbeats ( step 704 ). for each group ( step 706 ), the selected coefficients of the haar transform were computed ( step 708 ) for each of the ten heartbeats and the median value was then selected ( step 710 ) for each coefficient to form a proto - template . if these median values correlated well with the coefficient values for each of the ten heartbeats , the proto - template was retained ( step 714 ). otherwise , it was discarded . in this manner , a whole bunch of proto - templates were obtained from each patient episode . next , median values of the coefficients from all of the proto - templates ( step 716 ) were computed ( step 718 ). these values were then used as the population template for distinguishing st from non - st events ( step 720 ). while the preferred embodiment of the invention has been described above , it is to be understood that numerous modifications and changes will occur to those who are skilled in the art to which the invention pertains . accordingly , the following claims annexed to and forming a part of this specification are intended to define the true spirit and scope of the invention \u2014 that is , what is new and what is desired to be secured by letters patent of the united states ."}
|
Is the category the most suitable category for the given patent?
| 0.25 |
1cd5305994f1692ec7ec09368a2a725ff3c3dd15c21bf56c12e04e558cdb5fbc
| 0.046631 | 0.003601 | 0.013245 | 0.002472 | 0.148438 | 0.008057 |
null |
{"category": "Human Necessities", "patent": "the present invention is directed towards improving the quality and functionality of both implantable cardioverter defibrillators ( icds ) and implantable atrial defibrillators ( iads ). all such devices need to have a method and mechanism for accurately distinguishing sinus tachycardia ( st ) form non - st , such as atrial fibrillation and atrial flutter . this method and mechanism should be conservative , biased towards not initiating therapy unless non - st is clearly indicated . it should also require as few machine computational cycles as possible to reduce icd and iad battery drain . the principle underlying the present invention is that st and non - st can be distinguished by studying and measuring the atrial electrogram morphology to detect aberrant conduction in the atria . thus , by comparing normal st waveforms to a series of newly - occurring waveforms , st is indicated by abnormal atrial waveform morphology which signals aberrant conduction within the atria improper electrical patterns of depolarization . this will show up in the signal potentials captured by a bipolar atrial lead . with reference to fig1 the present invention can be implemented in an icd 100 ( which could be an iad ) that is implanted in the chest of a patient who suffers from unpleasant and possibly life - threatening irregularities in heart operation due to improper electrical stimulation of the heart muscle cells . during normal heart action , the heart &# 39 ; s electrical impulses originate in the sino - atrial node as an action potential that is transmitted smoothly to all portions of the atria , causing contraction of the atrial chambers . the electrical impulse continues in its path to a cluster of conduction fibrils known as the atrioventricular node . after a delay of about one - tenth of a second , an action potential flows over the ventricles and causes them to contract in synchronism with and following shortly after the atrial contraction . in this manner , the heart pumps blood to the lungs and from the lungs to the body . a bipolar lead 102 is implanted within the atria 104 of a heart 103 to measure the electrical potential between nearby cells of the atria . as the action potential actively passes from cell to cell past the two electrodes of the lead 102 , an oscillating signal potential is developed across the two electrodes and is conveyed by the bipolar lead 102 to the icd 100 where the signal is sampled about 1000 times per second and is digitized , with the sequential samples stored within a ram memory within the icd 100 for further analysis . another bipolar lead 105 may extend to the ventricle ( not shown ) of the heart 103 to measure the action potentials generated within the ventricles . these same leads , or other leads , may be used by the icd for administering various types of therapy , such as pacing or defibrillation shock therapy . the data samples collected from within the atria are now analyzed . first , with reference to data gathered from the ventricles , if some event in the ventricle , such as a mis - timed depolarization event that partially overlaps the atria &# 39 ; s p wave , could leak across to the atria and distort the measurement of the potentials for a given atrial depolarization , then the data for that particular heartbeat is discarded and is not used in further analyses . such distortion can also be found simply because a given heartbeat set of data gathered from the atria is itself badly distorted , and this approach must be taken in the case of an iad having no ventricular lead . the remaining data is retained for further processing . in preparation for further processing , the data for individual heartbeats is identified and gathered . the position within the gathered data of the negative spike to the atrial p depolarization waveform is determined and is selected as the center of the waveform for each beat . then , since data close to this p - wave is to be analyzed , and since data further away may be distorted to some degree by ventricular activity or by variations in the p - to - p interval , 32 sequential data samples are selected for each heartbeat such that the negative spike of the p - wave becomes signal potential sample number 16 of the 32 sequential samples selected for further analyses . these samples for a normal heartbeat appear at 302 in fig3 and at 502 in fig5 where the amplitude of the atrial signal potential is plotted against time over the 32 sampled values , with the negative p - wave spike positioned as signal sample number 16 . likewise , the samples for an abnormal heartbeat appear at 504 in fig5 . straightforward correlation ( cwa ) of the 32 samples against a template containing a typical or normal pattern could be performed at this juncture , as taught on page 563 ( equation ( 1 )) of the article by thorne , cited in the introductory portion of this specification . but the performance of such a cross correlation at repeated regular intervals would generate numerous computations and would drain the icd &# 39 ; s battery more rapidly than is desirable . in addition , provision would have to be made for a full template waveform that can be stored within the icd to be used as a comparison reference . in addition , such templates generally need to be patient specific , and they are more susceptible to normal changes in shape . accordingly , to reduce the mathematical complexity of the cross correlation computations , it is desirable to reduce the number of data points from 32 down to some much lower number through the use of some form of transformation into a different set of values . for example , the data could be transformed using the karhunen - loeve transformation described in the morris article cited at the beginning of this specification . but that transformation is fairly computationally intensive . computationally , a fast fourier transformation would be more efficient , but any such transformation into the frequency domain , where variable frequency sine and cosine wave amplitudes result from the transformation , does not match itself well to the morphology of atrial heart waveforms , which tend to be formed from large , ringing , spike - like representations of p - wave depolarization events travelling as moving action potential fronts past the pair of electrodes attached to the bipolar lead , and not as steady harmonics extending across time . accordingly , fourier - class transforms do not adequately reduce the number of significant data values that must be considered . and in addition , when the \u201c window \u201d size for a fourier analysis is selected , if it is wide , then the frequency values are not time - specific , but represent average signal harmonic content over the entire window time duration . and if the window is narrow , then the harmonics are too spread out in frequency , and the frequency - domain data generated by the transform does not resolve things finely enough in the frequency domain . for all of the above reasons , the present invention analyzes the waveforms using a discrete wavelet transform . this has the advantage that while low frequency wavelets are broad in time , high frequency wavelets have a much narrower timeslot focus . thus , for example , with 32 data samples taken sequentially over time , a wavelet transformation generates a d . c . wavelet that spans the entire set of 32 points ; a first reversing wavelet that also spans the entire set of points ; two second harmonic wavelets that each focus upon only one - half of the 32 data points ; four third harmonic wavelets that each focus upon only one - fourth of the 32 data points ; eight fourth harmonic wavelets that each focus upon four of the 32 data points ; and 16 fifth harmonic wavelets that each focuses upon only two of the 32 data points . accordingly , the higher - frequency wavelet amplitudes are quite time specific , unlike fourier sinusoidal amplitudes , while the lower - frequency and d . c . wavelet amplitudes give one broad information about the whole set of 32 points . and like the fourier transform , the discrete wavelet transform preserves all of the information of the original atrial signal , such that the transformation is fully reversible . 32 discrete wavelet transform values may be reverse transformed back into 32 values indicating the atrial signal potentials at 32 sequential points in time . in essence , the haar function is performing multiple digital filtering operations , at variable positions and utilizing varying - width windowing functions , to develop component values that represent varying types of heartbeat activity , at varying frequencies , spread over varying widths , and located at varying positions . these component values can be studied to see which subset of these component values are good at distinguishing st from non - st , or ( more generally ) which subset of these component values are good at distinguishing one type of heartbeat waveform from another . further study of such a selected subset can further determine which of the subset of components are relatively invariant from one individual &# 39 ; s heart to another . one preferably selects component values that are suitable from both of these two perspectives . the particular wavelet transformation to choose can be tailored to the nature of the data , with the wavelets chosen to resemble somewhat the values to be found within the data so that some transformed values are of much larger amplitude than others , and such that the low amplitude transformed values may be disregarded . of particular importance with an atrial p waveform of the type being analyzed here are wavelets having frequency values roughly comparable to and timed to coincide with the ringing of the atrial depolarization . this tailoring can minimize the number of transformed data values that are significant and that are candidates for full participation in the correlation analyses to determine if there has been a substantial change in waveform morphology . another factor in selecting a particular wavelet transformation is reducing the number of computations that must be performed to carry out the transformation . yet another factor is selecting a wavelet transformation where some of the transformed component values vary from normal to abnormal heart rhythms in much the same over a population of individuals so that a generic template of these component values may be derived that can be used with many individuals , rather than with just one individual . these factors suggest that the haar transformation would be a suitable candidate for use in analysis of atrial waveforms and possibly other waveforms as well . with reference to fig3 in the case of 32 voltage values sampled over time , the 32 applicable haar discrete transform wavelets appear as shown in this figure . other discrete wavelet transforms based upon wavelets having triangular or other shapes may also prove usable . the haar transform wavelets , in particular , have the shape of a square waveform , as will be described , and this can simplify the computations required . in fig3 time increases from left to right . a scale 304 indicates the 32 points in time at which the atrial waveform is sampled , and an analog atrial waveform ( before digitization ) is shown at 302 . the p wave negative notch 303 is shown centered at the 16 th timeslot so that the signal potential of the atria waveform sampled at this point in time becomes the 16 th data value in the set of 32 data values that are to be subjected to the haar transformation . a haar wavelet is simply one cycle of a square waveform that starts at zero , then swings positive ( to \u201c+ 1 \u201d), then swings negative ( to \u201c\u2212 1 \u201d), and then swings back to zero . the wavelet w 2 , shown at 308 in fig3 for example , is at \u201c+ 1 \u201d at points in time 1 to 16 and is at \u201c\u2212 1 \u201d at points in time 17 to 32 . the shorter waveform w 6 , shown at 316 in fig3 is at \u201c+ 1 \u201d at points in time 9 to 12 and is at \u201c\u2212 1 \u201d at points in time 13 to 16 , and is at \u201c 0 \u201d at all other points in time . the waveform w 1 , shown at 306 in fig3 is so slow to fluctuate in time that its \u201c\u2212 1 \u201d portion is off of the chart ( in fig3 ) to the right , and is ignored ; and accordingly , it has the value \u201c+ 1 \u201d for all 32 of the points in time shown in fig3 . it thus represents the average , or d . c ., component of the atrial signal potential . the lowest frequency wavelets w 1 and w 2 encompass the entire set of 32 points in time . the higher frequency wavelets w 3 and w 4 each encompasses only half of the points in time , and the two wavelets taken together encompass all the points in time . the four wavelets w 5 , w 6 , w 7 , and w 8 each encompasses only eight points in time , and the four wavelets taken together encompass all points in time . the eight wavelets w 9 . . . w 16 each encompasses only four points in time , and together all eight wavelets encompass all points in time . and finally , the wavelets w 17 . . . w 32 each encompass only two points in time , but the sixteen wavelets together encompass all points in time . thus , each wavelet has a characteristic time span and position as well as a characteristic frequency , with higher - frequency wavelets having a narrower and more specific time span and position than lower - frequency wavelets . this is advantageous with an impulse - type , ringing signal such as the atrial depolarization waveform considered here , since many higher - frequency transform values that correspond to haar wavelets positioned in time away from the p waveform or that do not correspond to its frequency of ringing may be low in amplitude such that they may safely be ignored during the analysis steps , thereby reducing the data that must be processed during the correlation steps . ( the above , while presented in the context of the haar wavelet , is also applicable to other wavelet shapes that may used to perform a discrete wavelet transform ( dwt )). each of the 32 haar transformed values is computed as follows : multiply each of the 32 sampled and digitized voltage values representing the analog atrial waveform 302 by the correspondingly - positioned -( in - time ) amplitude values of one of the 32 haar wavelets ( shown in fig3 ); then sum the resulting products ; and then multiply the resulting sums by a discrete wavelet transform scaling factor 2 \u2212 j / 2 , where j equals 1 for the wavelets w 17 to w 32 , 2 for w 9 to w 16 , 3 for w 5 to w 8 , 4 for w 3 to w 4 , and 5 for w 1 and w 2 . for example , and referring to fig3 : the first wavelet w 1 is always + 1 , so the atrial signal potential values are simply summed and then multiplied by 2 \u2212 5 / 2 ; the second wavelet is + 1 for time points 1 to 16 and \u2212 1 for time points 17 to 32 , so the first 16 values are summed , the second 16 values are summed , the difference between the first and second sums is computed , and the result is multiplied by 2 \u2212 5 / 2 ; and so on until all the transformed values have been computed , one for each haar wavelet . the 32 time - sequential atrial signal potential values are thus transformed into 32 haar wavelet amplitude values which may be called \u201c transformed values \u201d and which may be assigned the numbers 1 to 32 corresponding to the subscript numbers of the haar wavelets to which they each correspond and whose amplitudes they represent . the above description of how to compute the haar wavelet transformed values is accurate , but it is not the most efficient way to proceed in the case of haar wavelets . like the fourier transform , which has a corresponding fast fourier transform that saves intermediate results and thereby avoids re - computing them and thus reduces substantially the number of computations , the haar transformation also may be carried out in a manner that saves intermediate sums and re - uses them to reduce substantially the number of computations . this is described below at the point where fig4 is described , since fig4 illustrates graphically how this can be done . the illustrative program listing presented below is also an implementation of the fast haar wavelet transformation algorithm . the haar transformation can readily be carried out by any digital computer . as an example of how the haar transformation can be carried out on 32 values , the following program , written in the language c , is illustrative of many possible programs that may be written . in the exemplary program that follows , a 32 - element array yy contains 32 data values representing the 32 sampled atrial signal potential values representing the fluctuations over time of the signal supplied by the bipolar lead 102 . the analog atrial signal is sampled , digitized , broken into separate beat data sets , pre - processed ( to remove waveforms distorted by ventricular activity ), centered ( with the negative depolarization spike at data point 16 ), and fed into the subroutine presented below . this subroutine is compiled ( or assembled , if rewritten in assembly language ) and installed in the rom of the icd &# 39 ; s embedded microprocessor . this subroutine returns , contained within the same array yy , the 32 transformed haar wavelet amplitude values described above . ( the constant value \u201c recipsqrttwo \u201d is the reciprocal of the square root of two ). an illustrative version of the subroutine for computing all 32 of the haar transformed values in an efficient manner is presented here : void haar ( double yy []) { int i , j , l ; double zz [ 32 ]; for ( i = 5 ; i & gt ; 0 ; i \u2212\u2212) { l = 1 ; for ( j = 1 ; j & lt ;= i ; ++ j ) l = 2 * l ; for ( j = 0 ; j & lt ; l ; ++ j ) zz [ j ] = yy [ j ]; for ( j = 0 ; j & lt ; l \u2212 1 ; j = j + 2 ) ( yy [ j / 2 ] = recipsqrttwo * ( zz [ j ] + zz [ j + 1 ]); yy [( j + l )/ 2 ] = rcipsqrttwo * ( zz [ j ] \u2212 zz [ j + 1 ]); } } return ; } the above program computes the 32 haar transformed values 1 to 32 from the 32 atrial signal potential time - sequenced input values . it does so with only 31 additions , 31 subtractions , and 62 multiplications . it is carefully designed to compute each value in a systematic way , making multiple use of intermediate results . first , the program computes sixteen sums of and sixteen differences between eight adjacent pairs of the 16 time sequenced atrial signal potential values , and the sixteen difference values become the haar transformed values 17 to 32 , reflecting the strength of the highest frequency haar wavelets positioned at 16 different positions in time , corresponding to the wavelets w 17 to w 32 shown in fig3 . all the sum and difference values are scaled by multiplication by 2 \u2212 1 / 2 . next , taking the 16 sums of atrial signal potential values as an intermediate result , the above algorithm generates eight sums of and eight differences between adjacent pairs of these sixteen intermediate values that resulted from the first sixteen additions , and the eight newly - computed difference values become the haar transformed values 9 to 16 , reflecting the strength of the second to the highest frequency values at eight different points in time , corresponding to wavelets w 9 to w 16 in fig3 . all of these eight sum and eight difference values are again scaled by 2 \u2212 1 / 2 so that the wavelets w 9 to w 16 are scaled by \u00bd ( 2 \u2212 2 / 2 or 2 \u2212 1 / 2 times 2 \u2212 1 / 2 ). next , taking these eight sums of sums of adjacent atrial signal potential values as an intermediate result , the above algorithm generates four sum and four difference values , again scaling by 2 \u2212 1 / 2 , and the four difference values become haar transformed values 5 to 8 , reflecting the strength of the middle frequency wavelets at four different points in time , and corresponding to wavelets w 5 to w 8 in fig3 . next , taking these four sums of sums of sums of adjacent atrial signal potential values , the above algorithm generates two sum and two difference values , again scaling by 2 \u2212 1 / 2 , and the two difference values become haar transformed values 3 and 4 , reflecting the strength of the second to the lowest frequency wavelets only two of which encompass all the time domain data , corresponding to the wavelets w 3 and w 4 shown in fig3 . and finally , the algorithm generates the sum of and the difference between the final remaining two sums of sums of sums of sums of atrial signal potential values , again scaling by 2 \u2212 1 / 2 . the difference value is then the haar transformed value 2 , representing the strength of the lowest - frequency wavelet , the one corresponding to the wavelet w 2 in fig3 the wavelet that extends the full length of the time scale . the sum value is then the first , or d . c ., transformed value , representing the average signal potential level over the 32 sampled points in time , which corresponds to the wavelet w 1 in fig3 . another way of viewing this computation is illustrated in fig4 a and 4b . the 32 atrial signal potential values are shown at the top of fig4 a and are identified as c 0 5 through c 31 5 . the superscript \u201c 5 \u201d indicates that these values are processed when the index value \u201c n \u201d in the above computer program is equal to \u201c 5 \u201d\u2014 that is , during the first pass through the data generating intermediary sums and differences . during subsequent passes , the value of n is decremented to 4 , 3 , 2 , and finally to 1 . during each pass through the data , the computer generates sums c n - 1 and differences d n - 1 , as shown in fig4 b , between adjacent pairs of the values c 0 n and c 1 n ; c 2 n and c 3 n ; and so on , so that the number of newly - generated c n - 1 and d n - 1 terms is reduced by half with each computer pass through the intermediary results . at the end of all these computations , the value c 0 0 is the first , or d . c ., haar transformed value ; and the values d 0 0 ; d 0 1 and d 1 1 ; d 0 2 , d 1 2 , d 2 2 , and d 3 2 ; d 0 3 , d 1 3 , . . . and d 7 3 ; and d 0 4 , d 1 4 , and d 15 4 are , respectively , the remaining haar transform values 2 through 32 . fig4 thus illustrates quite succinctly how all the transform computations are carried out , and how the intermediary \u201c c \u201d sum values , such as c 0 4 , are used to compute multiple haar values , such as the values d 0 3 , d 0 2 , d 0 1 , d 0 0 , and c 0 0 all of which are computed from the intermediary value c 0 4 . saving and reusing these intermediary \u201c c \u201d values saves much computational time . fig4 b indicates the precise addition and subtraction operations that are carried out at each level to compute the values in the next lower level , proceeding down through the chart presented in fig4 a . but in any given application to heart waveform analysis , all of these computations may not be needed , and accordingly the number of computations may be reduced much further . since the present invention teaches that only a small number of these haar transformed values need actually be considered , a far less computationally intensive transform can be developed which only generates the intermediate and final transformed values that are actually needed to generate the specific haar transformed values which have proved to be significant in discriminating between sinus tachycardia ( or st ) and non - st conditions . all others need not be computed , and the above program may be reduced to a special algorithm that omits as many sums , differences , and multiplications as possible . for example , if only the transformed values 1 , 5 , 9 , and 24 are significant , then 31 additions are required to compute the first transformed value ( the d . c . value \u2014 the simple sum of all the time domain signal values ); six additions and one subtraction are required to compute the fifth transformed value ( in fig3 the difference between the sums of time domain values under each half of the wavelet w 5 at 314 in fig3 ); two additions and one subtraction are required to compute the ninth transformed value ( in fig3 the difference between the sums of the time domain values under each half of the wavelet w9 at 322 in fig3 ); and only one subtraction is required to compute the 24 th transformed value ( in fig3 the difference between the two time domain values under the respective halves of the wavelet w 24 ( not shown ) which is the same size as , but differently positioned in time than , the wavelet w 18 at 330 ( fig3 ). accordingly , if only the transformed values 1 , 5 , 9 , and 24 actually evaluated , then only 42 additions and subtractions are required . but even this number can be reduced further . if the computational algorithm set forth in the above illustrative computer program is followed as a guide , then the intermediary sums used in computing the haar first , or d . c ., transformed value can be re - used to compute the sums for the haar transformed values 5 and 9 , assuming these intermediary results are saved in the manner described in the above program example . then 31 additions are still required to compute the first transformed value , but only one subtraction is required to compute each of the transformed values 5 , 9 , and 24 , giving a total of additions and subtraction of only 34 operations . and if only transformed values 1 , 5 , and 9 are computed , the number of additions and subtractions is reduced to 33 . and if the first transformed value is omitted and if transformed values 5 , 9 , and 24 are selected , then the 31 additions needed to compute the first transformed value are not required . then the number of computations for transformed value 5 is 8 additions and one subtraction ; the number of computations for transformed value 9 is 2 additions and one subtraction ; and the number of computations for transformed value 24 is still just one subtraction . so the total number of additions and subtractions is just 13 . but even this number can be reduced to 11 when it is realized that the two additions done for the transformed value 9 are also done ( in the above computational algorithm ) when computing the transformed value 5 . so the number of additions and subtractions can be reduced to 11 . also , when computing such a small number of transformed values , the number of multiplications can be reduced as well by postponing the multiplications until a transformed value is actually computed . with reference to fig4 a , instead of dividing every sum and difference by the square root of two ( as shown in fig4 b ), several vertical sums of \u201c c \u201d values in different rows of the fig4 a table can be formed , and then a \u201c d \u201d value can be computed by subtraction , and then the resulting unscaled transformed value can be scaled with a single multiplication by 2 \u2212 j / 2 where j is the number of rows in the table of fig4 a traversed by the one or more sum and the one difference computations . thus , the number of scaling multiplications can sometimes be equal to the number of transformed values that are computed , typically 3 or 4 . accordingly , by using wavelet analysis and transformation , instead of fourier or discrete cosine or other sinusoid analysis and transformation , a highly useful and highly frequency - specific result can be achieved with a very small number of computations , thereby conserving power and battery life , and yet achieving a high degree of precision in recognizing changes in morphology . in our tests , we have selected certain transformed values as being much more significant than other values in distinguishing between st and non - st . we focused upon those transform values whose \u201c+ 1 \u201d and \u201c\u2212 1 \u201d scope included the middle 8 time points most significant to analysis of the peak of the atrial depolarization . working across test data samples obtained from patients who had dual chamber icds , we computed the variance of each transformed value for st and non - st events using the standard statistical formula for computing variance . we then calculated the ratio of the variance of non - st events to the variance of st events , and selected those terms with the highest variance ratios for further consideration . in this manner , we reduced the number of transformed values that were included in the cwa or correlation process while always maintaining or improving the performance achieved . ultimately we settled upon the three or four transformed values having the highest ratios . these were the transform values 1 , 5 , 9 , and 24 . we tested 1 , 5 , and 24 together ; 5 , 9 , and 24 together ; and 1 , 5 , 9 , and 24 together . these selected haar transform values , when used in these combinations , required minimal computations and thereby produced a savings in battery power , and that also gave better performance at distinguishing st from non - st than did all of the transformed values used together . so an increase in accuracy was achieved as well as a decrease in computational complexity by this approach to atrial waveform analysis . [ 0054 ] fig5 for example , illustrates actual plots of digital information illustrating the effect of using only the three coefficients 1 , 5 , and 24 . at 502 , a normal waveform is shown , with amplitude plotted against sample numbers from 1 to 32 . at 504 , an abnormal waveform is shown , again with amplitude plotted against sample numbers . after the haar transformation , at 506 and 508 , the wavelet coefficient amplitude values are indicated for the coefficients 1 to 32 . at 506 , the coefficients generated by the normal waveform 502 are shown , and at 508 , the coefficients generated by the abnormal waveform 504 are shown . one may directly compare these component values and verify that the coefficients 1 , 5 , and 24 at 506 and at 508 vary significantly in amplitude . comparison of this same information among different patients ( not shown in this figure ) also indicated that this variation is relatively constant from one patient to another . at 510 and 512 , using only the coefficients 1 , 5 , and 24 , the heartbeat waveforms are reconstructed by a reverse transformation , the normal reconstructed waveform shown at 510 and the abnormal reconstructed waveform shown at 512 . the marked differences between these two reconstructed waveforms highlights the way in which these three coefficients can signal an abnormal condition with less computation . in a practical system , a small number of transform values are selected , in the manner just described . the algorithm for computing these values is then refined , as explained above , to reduce the number of additions , subtractions , and multiplications to the minimum possible while preserving the accuracy of the computations . we first compute a value \u03c1 , which is the correlation between these values with the corresponding values in a template that represents the values for an average normal population . we then compare this value \u03c1 to a threshold correlation value \u03b2 that is chosen to give optimal results on experimental data . ( see the examples in the tables presented below .) for each waveform analyzed , a test is made to determine whether \u03c1 is greater than \u03b2 . if so , then this waveform is placed into the buffer marked \u201c st \u201d at step 210 . if not , then this waveform is placed into the buffer marked \u201c non - st \u201d at step 211 . finally , after ten waveforms have been analyzed , a count of st and non - st marked waveforms is made to see if the count of non - st waveforms is greater than some threshold value x ( at step 212 ), where x can be , for example , 7 . if more than seven waveforms are non - st , then therapy is delivered at step 214 . the buffers at steps 210 and 211 in fig2 are buffers that hold the last ten decision values . these buffers can be pictured as sliding windows revealing the most recent ten values to permit the decision at 212 to be continuously updated . the buffers thus function as a nonlinear filter preventing irregular values from delivering therapy improperly . in developing a working prototype of this system , we used a development data set consisting of 20 episodes of st taken from 4 patients and 18 episodes of af taken from 7 patients . patient number episodes of st episodes of af 1 2 6 2 12 0 13 0 1 4 0 2 5 5 1 6 0 3 7 0 3 8 0 2 9 1 0 total 20 18 we used this data to optimize our choice of \u03c1 and \u03b2 and also the particular coefficients that we chose to examine . next , we tested our prototype using a set of 16 episodes of st obtained from 4 patients together with 17 episodes of af obtained from 7 patients . patient number episodes of st episodes of af 1 8 0 2 3 1 3 3 0 4 2 0 5 0 4 6 0 4 7 0 3 8 0 3 9 0 1 10 0 1 total 16 17 x out threshold coefficients of 10 ( \u03b2 ) sensitivity specificity 1 , 5 , 9 , & amp ; 24 3 0 . 808 76 % 94 % 1 , 5 , 9 , & amp ; 24 5 0 . 975 88 % 94 % 1 , 5 , 9 , & amp ; 24 6 0 . 976 82 % 94 % 1 , 5 , 9 , & amp ; 24 7 0 . 983 88 % 94 % 1 , 5 , & amp ; 24 3 0 . 815 82 % 94 % 1 , 5 , & amp ; 24 4 0 . 943 88 % 94 % 1 , 5 , & amp ; 24 5 0 . 956 82 % 94 % 1 , 5 , & amp ; 24 7 0 . 991 94 % 94 % 5 , 9 , & amp ; 24 8 0 . 9993 82 % 94 % 5 , 9 , & amp ; 24 9 0 . 9997 76 % 81 % the quality of the selected subset of transformed values for characterizing changes in morphology may be demonstrated graphically , as indicated in fig5 by performing a reverse haar transform using only the three or four or so selected transformed values . as can be seen ( at 510 and 512 in fig5 ), the inverse transformations produce distorted representations of the original atrial waveforms which illustrate how the choice of transform values can emphasize the changes . this is another useful way to assist one in selecting which transformed values are most useful . one feature of the invention is its ability to use a template derived from an average normal population , rather than deriving a patient customized average normal template for each individual patient . prior systems have required each new patient to be monitored while in a normal state , and the captured data was then averaged to form a patient specific normal template . contrary to this , the present invention achieved the results shown above using the same template for all patients . accordingly , the invention may be used with a new patient without the necessity of such preliminary testing of each patient and without a new customized template necessarily having to be created for each patient . one reason why the present invention can function using a template that represents average values for a normal population is because the specific coefficients selected in the transform domain , in addition to having been selected to emphasize the difference between normal and abnormal values , are also preferably chosen to minimize the differences between different patients . in some cases , values that were good at distinguishing between normal and abnormal rhythms for one patient were not as good at doing so for some other patient . these values are preferably not selected for use in implementing the present invention . for example , the transform coefficients may be selected such that all ( or most of ) the normal values of a particular coefficient gathered from many patients had the same sign ( positive or negative ) and similar amplitudes ( or if they varied in sign , they were near to zero ). in addition , the abnormal coefficients are significantly different in value from the normal coefficients for each particular patient . [ 0068 ] fig6 at 600 , illustrates one way in which this may be done . first , at step 602 , one gathers a large number of st and non - st data samples . next , all of this data is transformed , generating coefficients for every waveform set of data ( step 604 ). the variance of each coefficient is then computed first for the st samples ( step 606 ) and then for the non - st samples . then , for each coefficient , the ratio of the variance of the non - st samples to that of the st samples is computed ( step 608 ). finally , those coefficients whose variance ratio was large may be selected a coefficients for use in creating a population template and in testing patients ( step 612 ). in addition , at step 614 , the number of coefficients retained may be further reduced to reduce the number of computations that need to be performed , as explained above . in the version of the invention that was used to generate the data shown above , the population template was computed as follows , using the 20 st episodes in the development data set ( also used in the first of the two tables presented above ). this is illustrated at 700 in fig7 . first , the coefficients were selected as was explained above and as shown in fig6 . next , for each patient episode of st ( step 702 ), the recorded heartbeats were broken up into groups of ten consecutive heartbeats ( step 704 ). for each group ( step 706 ), the selected coefficients of the haar transform were computed ( step 708 ) for each of the ten heartbeats and the median value was then selected ( step 710 ) for each coefficient to form a proto - template . if these median values correlated well with the coefficient values for each of the ten heartbeats , the proto - template was retained ( step 714 ). otherwise , it was discarded . in this manner , a whole bunch of proto - templates were obtained from each patient episode . next , median values of the coefficients from all of the proto - templates ( step 716 ) were computed ( step 718 ). these values were then used as the population template for distinguishing st from non - st events ( step 720 ). while the preferred embodiment of the invention has been described above , it is to be understood that numerous modifications and changes will occur to those who are skilled in the art to which the invention pertains . accordingly , the following claims annexed to and forming a part of this specification are intended to define the true spirit and scope of the invention \u2014 that is , what is new and what is desired to be secured by letters patent of the united states ."}
|
{"category": "Textiles; Paper", "patent": "the present invention is directed towards improving the quality and functionality of both implantable cardioverter defibrillators ( icds ) and implantable atrial defibrillators ( iads ). all such devices need to have a method and mechanism for accurately distinguishing sinus tachycardia ( st ) form non - st , such as atrial fibrillation and atrial flutter . this method and mechanism should be conservative , biased towards not initiating therapy unless non - st is clearly indicated . it should also require as few machine computational cycles as possible to reduce icd and iad battery drain . the principle underlying the present invention is that st and non - st can be distinguished by studying and measuring the atrial electrogram morphology to detect aberrant conduction in the atria . thus , by comparing normal st waveforms to a series of newly - occurring waveforms , st is indicated by abnormal atrial waveform morphology which signals aberrant conduction within the atria improper electrical patterns of depolarization . this will show up in the signal potentials captured by a bipolar atrial lead . with reference to fig1 the present invention can be implemented in an icd 100 ( which could be an iad ) that is implanted in the chest of a patient who suffers from unpleasant and possibly life - threatening irregularities in heart operation due to improper electrical stimulation of the heart muscle cells . during normal heart action , the heart &# 39 ; s electrical impulses originate in the sino - atrial node as an action potential that is transmitted smoothly to all portions of the atria , causing contraction of the atrial chambers . the electrical impulse continues in its path to a cluster of conduction fibrils known as the atrioventricular node . after a delay of about one - tenth of a second , an action potential flows over the ventricles and causes them to contract in synchronism with and following shortly after the atrial contraction . in this manner , the heart pumps blood to the lungs and from the lungs to the body . a bipolar lead 102 is implanted within the atria 104 of a heart 103 to measure the electrical potential between nearby cells of the atria . as the action potential actively passes from cell to cell past the two electrodes of the lead 102 , an oscillating signal potential is developed across the two electrodes and is conveyed by the bipolar lead 102 to the icd 100 where the signal is sampled about 1000 times per second and is digitized , with the sequential samples stored within a ram memory within the icd 100 for further analysis . another bipolar lead 105 may extend to the ventricle ( not shown ) of the heart 103 to measure the action potentials generated within the ventricles . these same leads , or other leads , may be used by the icd for administering various types of therapy , such as pacing or defibrillation shock therapy . the data samples collected from within the atria are now analyzed . first , with reference to data gathered from the ventricles , if some event in the ventricle , such as a mis - timed depolarization event that partially overlaps the atria &# 39 ; s p wave , could leak across to the atria and distort the measurement of the potentials for a given atrial depolarization , then the data for that particular heartbeat is discarded and is not used in further analyses . such distortion can also be found simply because a given heartbeat set of data gathered from the atria is itself badly distorted , and this approach must be taken in the case of an iad having no ventricular lead . the remaining data is retained for further processing . in preparation for further processing , the data for individual heartbeats is identified and gathered . the position within the gathered data of the negative spike to the atrial p depolarization waveform is determined and is selected as the center of the waveform for each beat . then , since data close to this p - wave is to be analyzed , and since data further away may be distorted to some degree by ventricular activity or by variations in the p - to - p interval , 32 sequential data samples are selected for each heartbeat such that the negative spike of the p - wave becomes signal potential sample number 16 of the 32 sequential samples selected for further analyses . these samples for a normal heartbeat appear at 302 in fig3 and at 502 in fig5 where the amplitude of the atrial signal potential is plotted against time over the 32 sampled values , with the negative p - wave spike positioned as signal sample number 16 . likewise , the samples for an abnormal heartbeat appear at 504 in fig5 . straightforward correlation ( cwa ) of the 32 samples against a template containing a typical or normal pattern could be performed at this juncture , as taught on page 563 ( equation ( 1 )) of the article by thorne , cited in the introductory portion of this specification . but the performance of such a cross correlation at repeated regular intervals would generate numerous computations and would drain the icd &# 39 ; s battery more rapidly than is desirable . in addition , provision would have to be made for a full template waveform that can be stored within the icd to be used as a comparison reference . in addition , such templates generally need to be patient specific , and they are more susceptible to normal changes in shape . accordingly , to reduce the mathematical complexity of the cross correlation computations , it is desirable to reduce the number of data points from 32 down to some much lower number through the use of some form of transformation into a different set of values . for example , the data could be transformed using the karhunen - loeve transformation described in the morris article cited at the beginning of this specification . but that transformation is fairly computationally intensive . computationally , a fast fourier transformation would be more efficient , but any such transformation into the frequency domain , where variable frequency sine and cosine wave amplitudes result from the transformation , does not match itself well to the morphology of atrial heart waveforms , which tend to be formed from large , ringing , spike - like representations of p - wave depolarization events travelling as moving action potential fronts past the pair of electrodes attached to the bipolar lead , and not as steady harmonics extending across time . accordingly , fourier - class transforms do not adequately reduce the number of significant data values that must be considered . and in addition , when the \u201c window \u201d size for a fourier analysis is selected , if it is wide , then the frequency values are not time - specific , but represent average signal harmonic content over the entire window time duration . and if the window is narrow , then the harmonics are too spread out in frequency , and the frequency - domain data generated by the transform does not resolve things finely enough in the frequency domain . for all of the above reasons , the present invention analyzes the waveforms using a discrete wavelet transform . this has the advantage that while low frequency wavelets are broad in time , high frequency wavelets have a much narrower timeslot focus . thus , for example , with 32 data samples taken sequentially over time , a wavelet transformation generates a d . c . wavelet that spans the entire set of 32 points ; a first reversing wavelet that also spans the entire set of points ; two second harmonic wavelets that each focus upon only one - half of the 32 data points ; four third harmonic wavelets that each focus upon only one - fourth of the 32 data points ; eight fourth harmonic wavelets that each focus upon four of the 32 data points ; and 16 fifth harmonic wavelets that each focuses upon only two of the 32 data points . accordingly , the higher - frequency wavelet amplitudes are quite time specific , unlike fourier sinusoidal amplitudes , while the lower - frequency and d . c . wavelet amplitudes give one broad information about the whole set of 32 points . and like the fourier transform , the discrete wavelet transform preserves all of the information of the original atrial signal , such that the transformation is fully reversible . 32 discrete wavelet transform values may be reverse transformed back into 32 values indicating the atrial signal potentials at 32 sequential points in time . in essence , the haar function is performing multiple digital filtering operations , at variable positions and utilizing varying - width windowing functions , to develop component values that represent varying types of heartbeat activity , at varying frequencies , spread over varying widths , and located at varying positions . these component values can be studied to see which subset of these component values are good at distinguishing st from non - st , or ( more generally ) which subset of these component values are good at distinguishing one type of heartbeat waveform from another . further study of such a selected subset can further determine which of the subset of components are relatively invariant from one individual &# 39 ; s heart to another . one preferably selects component values that are suitable from both of these two perspectives . the particular wavelet transformation to choose can be tailored to the nature of the data , with the wavelets chosen to resemble somewhat the values to be found within the data so that some transformed values are of much larger amplitude than others , and such that the low amplitude transformed values may be disregarded . of particular importance with an atrial p waveform of the type being analyzed here are wavelets having frequency values roughly comparable to and timed to coincide with the ringing of the atrial depolarization . this tailoring can minimize the number of transformed data values that are significant and that are candidates for full participation in the correlation analyses to determine if there has been a substantial change in waveform morphology . another factor in selecting a particular wavelet transformation is reducing the number of computations that must be performed to carry out the transformation . yet another factor is selecting a wavelet transformation where some of the transformed component values vary from normal to abnormal heart rhythms in much the same over a population of individuals so that a generic template of these component values may be derived that can be used with many individuals , rather than with just one individual . these factors suggest that the haar transformation would be a suitable candidate for use in analysis of atrial waveforms and possibly other waveforms as well . with reference to fig3 in the case of 32 voltage values sampled over time , the 32 applicable haar discrete transform wavelets appear as shown in this figure . other discrete wavelet transforms based upon wavelets having triangular or other shapes may also prove usable . the haar transform wavelets , in particular , have the shape of a square waveform , as will be described , and this can simplify the computations required . in fig3 time increases from left to right . a scale 304 indicates the 32 points in time at which the atrial waveform is sampled , and an analog atrial waveform ( before digitization ) is shown at 302 . the p wave negative notch 303 is shown centered at the 16 th timeslot so that the signal potential of the atria waveform sampled at this point in time becomes the 16 th data value in the set of 32 data values that are to be subjected to the haar transformation . a haar wavelet is simply one cycle of a square waveform that starts at zero , then swings positive ( to \u201c+ 1 \u201d), then swings negative ( to \u201c\u2212 1 \u201d), and then swings back to zero . the wavelet w 2 , shown at 308 in fig3 for example , is at \u201c+ 1 \u201d at points in time 1 to 16 and is at \u201c\u2212 1 \u201d at points in time 17 to 32 . the shorter waveform w 6 , shown at 316 in fig3 is at \u201c+ 1 \u201d at points in time 9 to 12 and is at \u201c\u2212 1 \u201d at points in time 13 to 16 , and is at \u201c 0 \u201d at all other points in time . the waveform w 1 , shown at 306 in fig3 is so slow to fluctuate in time that its \u201c\u2212 1 \u201d portion is off of the chart ( in fig3 ) to the right , and is ignored ; and accordingly , it has the value \u201c+ 1 \u201d for all 32 of the points in time shown in fig3 . it thus represents the average , or d . c ., component of the atrial signal potential . the lowest frequency wavelets w 1 and w 2 encompass the entire set of 32 points in time . the higher frequency wavelets w 3 and w 4 each encompasses only half of the points in time , and the two wavelets taken together encompass all the points in time . the four wavelets w 5 , w 6 , w 7 , and w 8 each encompasses only eight points in time , and the four wavelets taken together encompass all points in time . the eight wavelets w 9 . . . w 16 each encompasses only four points in time , and together all eight wavelets encompass all points in time . and finally , the wavelets w 17 . . . w 32 each encompass only two points in time , but the sixteen wavelets together encompass all points in time . thus , each wavelet has a characteristic time span and position as well as a characteristic frequency , with higher - frequency wavelets having a narrower and more specific time span and position than lower - frequency wavelets . this is advantageous with an impulse - type , ringing signal such as the atrial depolarization waveform considered here , since many higher - frequency transform values that correspond to haar wavelets positioned in time away from the p waveform or that do not correspond to its frequency of ringing may be low in amplitude such that they may safely be ignored during the analysis steps , thereby reducing the data that must be processed during the correlation steps . ( the above , while presented in the context of the haar wavelet , is also applicable to other wavelet shapes that may used to perform a discrete wavelet transform ( dwt )). each of the 32 haar transformed values is computed as follows : multiply each of the 32 sampled and digitized voltage values representing the analog atrial waveform 302 by the correspondingly - positioned -( in - time ) amplitude values of one of the 32 haar wavelets ( shown in fig3 ); then sum the resulting products ; and then multiply the resulting sums by a discrete wavelet transform scaling factor 2 \u2212 j / 2 , where j equals 1 for the wavelets w 17 to w 32 , 2 for w 9 to w 16 , 3 for w 5 to w 8 , 4 for w 3 to w 4 , and 5 for w 1 and w 2 . for example , and referring to fig3 : the first wavelet w 1 is always + 1 , so the atrial signal potential values are simply summed and then multiplied by 2 \u2212 5 / 2 ; the second wavelet is + 1 for time points 1 to 16 and \u2212 1 for time points 17 to 32 , so the first 16 values are summed , the second 16 values are summed , the difference between the first and second sums is computed , and the result is multiplied by 2 \u2212 5 / 2 ; and so on until all the transformed values have been computed , one for each haar wavelet . the 32 time - sequential atrial signal potential values are thus transformed into 32 haar wavelet amplitude values which may be called \u201c transformed values \u201d and which may be assigned the numbers 1 to 32 corresponding to the subscript numbers of the haar wavelets to which they each correspond and whose amplitudes they represent . the above description of how to compute the haar wavelet transformed values is accurate , but it is not the most efficient way to proceed in the case of haar wavelets . like the fourier transform , which has a corresponding fast fourier transform that saves intermediate results and thereby avoids re - computing them and thus reduces substantially the number of computations , the haar transformation also may be carried out in a manner that saves intermediate sums and re - uses them to reduce substantially the number of computations . this is described below at the point where fig4 is described , since fig4 illustrates graphically how this can be done . the illustrative program listing presented below is also an implementation of the fast haar wavelet transformation algorithm . the haar transformation can readily be carried out by any digital computer . as an example of how the haar transformation can be carried out on 32 values , the following program , written in the language c , is illustrative of many possible programs that may be written . in the exemplary program that follows , a 32 - element array yy contains 32 data values representing the 32 sampled atrial signal potential values representing the fluctuations over time of the signal supplied by the bipolar lead 102 . the analog atrial signal is sampled , digitized , broken into separate beat data sets , pre - processed ( to remove waveforms distorted by ventricular activity ), centered ( with the negative depolarization spike at data point 16 ), and fed into the subroutine presented below . this subroutine is compiled ( or assembled , if rewritten in assembly language ) and installed in the rom of the icd &# 39 ; s embedded microprocessor . this subroutine returns , contained within the same array yy , the 32 transformed haar wavelet amplitude values described above . ( the constant value \u201c recipsqrttwo \u201d is the reciprocal of the square root of two ). an illustrative version of the subroutine for computing all 32 of the haar transformed values in an efficient manner is presented here : void haar ( double yy []) { int i , j , l ; double zz [ 32 ]; for ( i = 5 ; i & gt ; 0 ; i \u2212\u2212) { l = 1 ; for ( j = 1 ; j & lt ;= i ; ++ j ) l = 2 * l ; for ( j = 0 ; j & lt ; l ; ++ j ) zz [ j ] = yy [ j ]; for ( j = 0 ; j & lt ; l \u2212 1 ; j = j + 2 ) ( yy [ j / 2 ] = recipsqrttwo * ( zz [ j ] + zz [ j + 1 ]); yy [( j + l )/ 2 ] = rcipsqrttwo * ( zz [ j ] \u2212 zz [ j + 1 ]); } } return ; } the above program computes the 32 haar transformed values 1 to 32 from the 32 atrial signal potential time - sequenced input values . it does so with only 31 additions , 31 subtractions , and 62 multiplications . it is carefully designed to compute each value in a systematic way , making multiple use of intermediate results . first , the program computes sixteen sums of and sixteen differences between eight adjacent pairs of the 16 time sequenced atrial signal potential values , and the sixteen difference values become the haar transformed values 17 to 32 , reflecting the strength of the highest frequency haar wavelets positioned at 16 different positions in time , corresponding to the wavelets w 17 to w 32 shown in fig3 . all the sum and difference values are scaled by multiplication by 2 \u2212 1 / 2 . next , taking the 16 sums of atrial signal potential values as an intermediate result , the above algorithm generates eight sums of and eight differences between adjacent pairs of these sixteen intermediate values that resulted from the first sixteen additions , and the eight newly - computed difference values become the haar transformed values 9 to 16 , reflecting the strength of the second to the highest frequency values at eight different points in time , corresponding to wavelets w 9 to w 16 in fig3 . all of these eight sum and eight difference values are again scaled by 2 \u2212 1 / 2 so that the wavelets w 9 to w 16 are scaled by \u00bd ( 2 \u2212 2 / 2 or 2 \u2212 1 / 2 times 2 \u2212 1 / 2 ). next , taking these eight sums of sums of adjacent atrial signal potential values as an intermediate result , the above algorithm generates four sum and four difference values , again scaling by 2 \u2212 1 / 2 , and the four difference values become haar transformed values 5 to 8 , reflecting the strength of the middle frequency wavelets at four different points in time , and corresponding to wavelets w 5 to w 8 in fig3 . next , taking these four sums of sums of sums of adjacent atrial signal potential values , the above algorithm generates two sum and two difference values , again scaling by 2 \u2212 1 / 2 , and the two difference values become haar transformed values 3 and 4 , reflecting the strength of the second to the lowest frequency wavelets only two of which encompass all the time domain data , corresponding to the wavelets w 3 and w 4 shown in fig3 . and finally , the algorithm generates the sum of and the difference between the final remaining two sums of sums of sums of sums of atrial signal potential values , again scaling by 2 \u2212 1 / 2 . the difference value is then the haar transformed value 2 , representing the strength of the lowest - frequency wavelet , the one corresponding to the wavelet w 2 in fig3 the wavelet that extends the full length of the time scale . the sum value is then the first , or d . c ., transformed value , representing the average signal potential level over the 32 sampled points in time , which corresponds to the wavelet w 1 in fig3 . another way of viewing this computation is illustrated in fig4 a and 4b . the 32 atrial signal potential values are shown at the top of fig4 a and are identified as c 0 5 through c 31 5 . the superscript \u201c 5 \u201d indicates that these values are processed when the index value \u201c n \u201d in the above computer program is equal to \u201c 5 \u201d\u2014 that is , during the first pass through the data generating intermediary sums and differences . during subsequent passes , the value of n is decremented to 4 , 3 , 2 , and finally to 1 . during each pass through the data , the computer generates sums c n - 1 and differences d n - 1 , as shown in fig4 b , between adjacent pairs of the values c 0 n and c 1 n ; c 2 n and c 3 n ; and so on , so that the number of newly - generated c n - 1 and d n - 1 terms is reduced by half with each computer pass through the intermediary results . at the end of all these computations , the value c 0 0 is the first , or d . c ., haar transformed value ; and the values d 0 0 ; d 0 1 and d 1 1 ; d 0 2 , d 1 2 , d 2 2 , and d 3 2 ; d 0 3 , d 1 3 , . . . and d 7 3 ; and d 0 4 , d 1 4 , and d 15 4 are , respectively , the remaining haar transform values 2 through 32 . fig4 thus illustrates quite succinctly how all the transform computations are carried out , and how the intermediary \u201c c \u201d sum values , such as c 0 4 , are used to compute multiple haar values , such as the values d 0 3 , d 0 2 , d 0 1 , d 0 0 , and c 0 0 all of which are computed from the intermediary value c 0 4 . saving and reusing these intermediary \u201c c \u201d values saves much computational time . fig4 b indicates the precise addition and subtraction operations that are carried out at each level to compute the values in the next lower level , proceeding down through the chart presented in fig4 a . but in any given application to heart waveform analysis , all of these computations may not be needed , and accordingly the number of computations may be reduced much further . since the present invention teaches that only a small number of these haar transformed values need actually be considered , a far less computationally intensive transform can be developed which only generates the intermediate and final transformed values that are actually needed to generate the specific haar transformed values which have proved to be significant in discriminating between sinus tachycardia ( or st ) and non - st conditions . all others need not be computed , and the above program may be reduced to a special algorithm that omits as many sums , differences , and multiplications as possible . for example , if only the transformed values 1 , 5 , 9 , and 24 are significant , then 31 additions are required to compute the first transformed value ( the d . c . value \u2014 the simple sum of all the time domain signal values ); six additions and one subtraction are required to compute the fifth transformed value ( in fig3 the difference between the sums of time domain values under each half of the wavelet w 5 at 314 in fig3 ); two additions and one subtraction are required to compute the ninth transformed value ( in fig3 the difference between the sums of the time domain values under each half of the wavelet w9 at 322 in fig3 ); and only one subtraction is required to compute the 24 th transformed value ( in fig3 the difference between the two time domain values under the respective halves of the wavelet w 24 ( not shown ) which is the same size as , but differently positioned in time than , the wavelet w 18 at 330 ( fig3 ). accordingly , if only the transformed values 1 , 5 , 9 , and 24 actually evaluated , then only 42 additions and subtractions are required . but even this number can be reduced further . if the computational algorithm set forth in the above illustrative computer program is followed as a guide , then the intermediary sums used in computing the haar first , or d . c ., transformed value can be re - used to compute the sums for the haar transformed values 5 and 9 , assuming these intermediary results are saved in the manner described in the above program example . then 31 additions are still required to compute the first transformed value , but only one subtraction is required to compute each of the transformed values 5 , 9 , and 24 , giving a total of additions and subtraction of only 34 operations . and if only transformed values 1 , 5 , and 9 are computed , the number of additions and subtractions is reduced to 33 . and if the first transformed value is omitted and if transformed values 5 , 9 , and 24 are selected , then the 31 additions needed to compute the first transformed value are not required . then the number of computations for transformed value 5 is 8 additions and one subtraction ; the number of computations for transformed value 9 is 2 additions and one subtraction ; and the number of computations for transformed value 24 is still just one subtraction . so the total number of additions and subtractions is just 13 . but even this number can be reduced to 11 when it is realized that the two additions done for the transformed value 9 are also done ( in the above computational algorithm ) when computing the transformed value 5 . so the number of additions and subtractions can be reduced to 11 . also , when computing such a small number of transformed values , the number of multiplications can be reduced as well by postponing the multiplications until a transformed value is actually computed . with reference to fig4 a , instead of dividing every sum and difference by the square root of two ( as shown in fig4 b ), several vertical sums of \u201c c \u201d values in different rows of the fig4 a table can be formed , and then a \u201c d \u201d value can be computed by subtraction , and then the resulting unscaled transformed value can be scaled with a single multiplication by 2 \u2212 j / 2 where j is the number of rows in the table of fig4 a traversed by the one or more sum and the one difference computations . thus , the number of scaling multiplications can sometimes be equal to the number of transformed values that are computed , typically 3 or 4 . accordingly , by using wavelet analysis and transformation , instead of fourier or discrete cosine or other sinusoid analysis and transformation , a highly useful and highly frequency - specific result can be achieved with a very small number of computations , thereby conserving power and battery life , and yet achieving a high degree of precision in recognizing changes in morphology . in our tests , we have selected certain transformed values as being much more significant than other values in distinguishing between st and non - st . we focused upon those transform values whose \u201c+ 1 \u201d and \u201c\u2212 1 \u201d scope included the middle 8 time points most significant to analysis of the peak of the atrial depolarization . working across test data samples obtained from patients who had dual chamber icds , we computed the variance of each transformed value for st and non - st events using the standard statistical formula for computing variance . we then calculated the ratio of the variance of non - st events to the variance of st events , and selected those terms with the highest variance ratios for further consideration . in this manner , we reduced the number of transformed values that were included in the cwa or correlation process while always maintaining or improving the performance achieved . ultimately we settled upon the three or four transformed values having the highest ratios . these were the transform values 1 , 5 , 9 , and 24 . we tested 1 , 5 , and 24 together ; 5 , 9 , and 24 together ; and 1 , 5 , 9 , and 24 together . these selected haar transform values , when used in these combinations , required minimal computations and thereby produced a savings in battery power , and that also gave better performance at distinguishing st from non - st than did all of the transformed values used together . so an increase in accuracy was achieved as well as a decrease in computational complexity by this approach to atrial waveform analysis . [ 0054 ] fig5 for example , illustrates actual plots of digital information illustrating the effect of using only the three coefficients 1 , 5 , and 24 . at 502 , a normal waveform is shown , with amplitude plotted against sample numbers from 1 to 32 . at 504 , an abnormal waveform is shown , again with amplitude plotted against sample numbers . after the haar transformation , at 506 and 508 , the wavelet coefficient amplitude values are indicated for the coefficients 1 to 32 . at 506 , the coefficients generated by the normal waveform 502 are shown , and at 508 , the coefficients generated by the abnormal waveform 504 are shown . one may directly compare these component values and verify that the coefficients 1 , 5 , and 24 at 506 and at 508 vary significantly in amplitude . comparison of this same information among different patients ( not shown in this figure ) also indicated that this variation is relatively constant from one patient to another . at 510 and 512 , using only the coefficients 1 , 5 , and 24 , the heartbeat waveforms are reconstructed by a reverse transformation , the normal reconstructed waveform shown at 510 and the abnormal reconstructed waveform shown at 512 . the marked differences between these two reconstructed waveforms highlights the way in which these three coefficients can signal an abnormal condition with less computation . in a practical system , a small number of transform values are selected , in the manner just described . the algorithm for computing these values is then refined , as explained above , to reduce the number of additions , subtractions , and multiplications to the minimum possible while preserving the accuracy of the computations . we first compute a value \u03c1 , which is the correlation between these values with the corresponding values in a template that represents the values for an average normal population . we then compare this value \u03c1 to a threshold correlation value \u03b2 that is chosen to give optimal results on experimental data . ( see the examples in the tables presented below .) for each waveform analyzed , a test is made to determine whether \u03c1 is greater than \u03b2 . if so , then this waveform is placed into the buffer marked \u201c st \u201d at step 210 . if not , then this waveform is placed into the buffer marked \u201c non - st \u201d at step 211 . finally , after ten waveforms have been analyzed , a count of st and non - st marked waveforms is made to see if the count of non - st waveforms is greater than some threshold value x ( at step 212 ), where x can be , for example , 7 . if more than seven waveforms are non - st , then therapy is delivered at step 214 . the buffers at steps 210 and 211 in fig2 are buffers that hold the last ten decision values . these buffers can be pictured as sliding windows revealing the most recent ten values to permit the decision at 212 to be continuously updated . the buffers thus function as a nonlinear filter preventing irregular values from delivering therapy improperly . in developing a working prototype of this system , we used a development data set consisting of 20 episodes of st taken from 4 patients and 18 episodes of af taken from 7 patients . patient number episodes of st episodes of af 1 2 6 2 12 0 13 0 1 4 0 2 5 5 1 6 0 3 7 0 3 8 0 2 9 1 0 total 20 18 we used this data to optimize our choice of \u03c1 and \u03b2 and also the particular coefficients that we chose to examine . next , we tested our prototype using a set of 16 episodes of st obtained from 4 patients together with 17 episodes of af obtained from 7 patients . patient number episodes of st episodes of af 1 8 0 2 3 1 3 3 0 4 2 0 5 0 4 6 0 4 7 0 3 8 0 3 9 0 1 10 0 1 total 16 17 x out threshold coefficients of 10 ( \u03b2 ) sensitivity specificity 1 , 5 , 9 , & amp ; 24 3 0 . 808 76 % 94 % 1 , 5 , 9 , & amp ; 24 5 0 . 975 88 % 94 % 1 , 5 , 9 , & amp ; 24 6 0 . 976 82 % 94 % 1 , 5 , 9 , & amp ; 24 7 0 . 983 88 % 94 % 1 , 5 , & amp ; 24 3 0 . 815 82 % 94 % 1 , 5 , & amp ; 24 4 0 . 943 88 % 94 % 1 , 5 , & amp ; 24 5 0 . 956 82 % 94 % 1 , 5 , & amp ; 24 7 0 . 991 94 % 94 % 5 , 9 , & amp ; 24 8 0 . 9993 82 % 94 % 5 , 9 , & amp ; 24 9 0 . 9997 76 % 81 % the quality of the selected subset of transformed values for characterizing changes in morphology may be demonstrated graphically , as indicated in fig5 by performing a reverse haar transform using only the three or four or so selected transformed values . as can be seen ( at 510 and 512 in fig5 ), the inverse transformations produce distorted representations of the original atrial waveforms which illustrate how the choice of transform values can emphasize the changes . this is another useful way to assist one in selecting which transformed values are most useful . one feature of the invention is its ability to use a template derived from an average normal population , rather than deriving a patient customized average normal template for each individual patient . prior systems have required each new patient to be monitored while in a normal state , and the captured data was then averaged to form a patient specific normal template . contrary to this , the present invention achieved the results shown above using the same template for all patients . accordingly , the invention may be used with a new patient without the necessity of such preliminary testing of each patient and without a new customized template necessarily having to be created for each patient . one reason why the present invention can function using a template that represents average values for a normal population is because the specific coefficients selected in the transform domain , in addition to having been selected to emphasize the difference between normal and abnormal values , are also preferably chosen to minimize the differences between different patients . in some cases , values that were good at distinguishing between normal and abnormal rhythms for one patient were not as good at doing so for some other patient . these values are preferably not selected for use in implementing the present invention . for example , the transform coefficients may be selected such that all ( or most of ) the normal values of a particular coefficient gathered from many patients had the same sign ( positive or negative ) and similar amplitudes ( or if they varied in sign , they were near to zero ). in addition , the abnormal coefficients are significantly different in value from the normal coefficients for each particular patient . [ 0068 ] fig6 at 600 , illustrates one way in which this may be done . first , at step 602 , one gathers a large number of st and non - st data samples . next , all of this data is transformed , generating coefficients for every waveform set of data ( step 604 ). the variance of each coefficient is then computed first for the st samples ( step 606 ) and then for the non - st samples . then , for each coefficient , the ratio of the variance of the non - st samples to that of the st samples is computed ( step 608 ). finally , those coefficients whose variance ratio was large may be selected a coefficients for use in creating a population template and in testing patients ( step 612 ). in addition , at step 614 , the number of coefficients retained may be further reduced to reduce the number of computations that need to be performed , as explained above . in the version of the invention that was used to generate the data shown above , the population template was computed as follows , using the 20 st episodes in the development data set ( also used in the first of the two tables presented above ). this is illustrated at 700 in fig7 . first , the coefficients were selected as was explained above and as shown in fig6 . next , for each patient episode of st ( step 702 ), the recorded heartbeats were broken up into groups of ten consecutive heartbeats ( step 704 ). for each group ( step 706 ), the selected coefficients of the haar transform were computed ( step 708 ) for each of the ten heartbeats and the median value was then selected ( step 710 ) for each coefficient to form a proto - template . if these median values correlated well with the coefficient values for each of the ten heartbeats , the proto - template was retained ( step 714 ). otherwise , it was discarded . in this manner , a whole bunch of proto - templates were obtained from each patient episode . next , median values of the coefficients from all of the proto - templates ( step 716 ) were computed ( step 718 ). these values were then used as the population template for distinguishing st from non - st events ( step 720 ). while the preferred embodiment of the invention has been described above , it is to be understood that numerous modifications and changes will occur to those who are skilled in the art to which the invention pertains . accordingly , the following claims annexed to and forming a part of this specification are intended to define the true spirit and scope of the invention \u2014 that is , what is new and what is desired to be secured by letters patent of the united states ."}
|
Does the patent belong in this category?
| 0.25 |
1cd5305994f1692ec7ec09368a2a725ff3c3dd15c21bf56c12e04e558cdb5fbc
| 0.145508 | 0.027222 | 0.172852 | 0.000553 | 0.314453 | 0.044678 |
null |
{"category": "Human Necessities", "patent": "the present invention is directed towards improving the quality and functionality of both implantable cardioverter defibrillators ( icds ) and implantable atrial defibrillators ( iads ). all such devices need to have a method and mechanism for accurately distinguishing sinus tachycardia ( st ) form non - st , such as atrial fibrillation and atrial flutter . this method and mechanism should be conservative , biased towards not initiating therapy unless non - st is clearly indicated . it should also require as few machine computational cycles as possible to reduce icd and iad battery drain . the principle underlying the present invention is that st and non - st can be distinguished by studying and measuring the atrial electrogram morphology to detect aberrant conduction in the atria . thus , by comparing normal st waveforms to a series of newly - occurring waveforms , st is indicated by abnormal atrial waveform morphology which signals aberrant conduction within the atria improper electrical patterns of depolarization . this will show up in the signal potentials captured by a bipolar atrial lead . with reference to fig1 the present invention can be implemented in an icd 100 ( which could be an iad ) that is implanted in the chest of a patient who suffers from unpleasant and possibly life - threatening irregularities in heart operation due to improper electrical stimulation of the heart muscle cells . during normal heart action , the heart &# 39 ; s electrical impulses originate in the sino - atrial node as an action potential that is transmitted smoothly to all portions of the atria , causing contraction of the atrial chambers . the electrical impulse continues in its path to a cluster of conduction fibrils known as the atrioventricular node . after a delay of about one - tenth of a second , an action potential flows over the ventricles and causes them to contract in synchronism with and following shortly after the atrial contraction . in this manner , the heart pumps blood to the lungs and from the lungs to the body . a bipolar lead 102 is implanted within the atria 104 of a heart 103 to measure the electrical potential between nearby cells of the atria . as the action potential actively passes from cell to cell past the two electrodes of the lead 102 , an oscillating signal potential is developed across the two electrodes and is conveyed by the bipolar lead 102 to the icd 100 where the signal is sampled about 1000 times per second and is digitized , with the sequential samples stored within a ram memory within the icd 100 for further analysis . another bipolar lead 105 may extend to the ventricle ( not shown ) of the heart 103 to measure the action potentials generated within the ventricles . these same leads , or other leads , may be used by the icd for administering various types of therapy , such as pacing or defibrillation shock therapy . the data samples collected from within the atria are now analyzed . first , with reference to data gathered from the ventricles , if some event in the ventricle , such as a mis - timed depolarization event that partially overlaps the atria &# 39 ; s p wave , could leak across to the atria and distort the measurement of the potentials for a given atrial depolarization , then the data for that particular heartbeat is discarded and is not used in further analyses . such distortion can also be found simply because a given heartbeat set of data gathered from the atria is itself badly distorted , and this approach must be taken in the case of an iad having no ventricular lead . the remaining data is retained for further processing . in preparation for further processing , the data for individual heartbeats is identified and gathered . the position within the gathered data of the negative spike to the atrial p depolarization waveform is determined and is selected as the center of the waveform for each beat . then , since data close to this p - wave is to be analyzed , and since data further away may be distorted to some degree by ventricular activity or by variations in the p - to - p interval , 32 sequential data samples are selected for each heartbeat such that the negative spike of the p - wave becomes signal potential sample number 16 of the 32 sequential samples selected for further analyses . these samples for a normal heartbeat appear at 302 in fig3 and at 502 in fig5 where the amplitude of the atrial signal potential is plotted against time over the 32 sampled values , with the negative p - wave spike positioned as signal sample number 16 . likewise , the samples for an abnormal heartbeat appear at 504 in fig5 . straightforward correlation ( cwa ) of the 32 samples against a template containing a typical or normal pattern could be performed at this juncture , as taught on page 563 ( equation ( 1 )) of the article by thorne , cited in the introductory portion of this specification . but the performance of such a cross correlation at repeated regular intervals would generate numerous computations and would drain the icd &# 39 ; s battery more rapidly than is desirable . in addition , provision would have to be made for a full template waveform that can be stored within the icd to be used as a comparison reference . in addition , such templates generally need to be patient specific , and they are more susceptible to normal changes in shape . accordingly , to reduce the mathematical complexity of the cross correlation computations , it is desirable to reduce the number of data points from 32 down to some much lower number through the use of some form of transformation into a different set of values . for example , the data could be transformed using the karhunen - loeve transformation described in the morris article cited at the beginning of this specification . but that transformation is fairly computationally intensive . computationally , a fast fourier transformation would be more efficient , but any such transformation into the frequency domain , where variable frequency sine and cosine wave amplitudes result from the transformation , does not match itself well to the morphology of atrial heart waveforms , which tend to be formed from large , ringing , spike - like representations of p - wave depolarization events travelling as moving action potential fronts past the pair of electrodes attached to the bipolar lead , and not as steady harmonics extending across time . accordingly , fourier - class transforms do not adequately reduce the number of significant data values that must be considered . and in addition , when the \u201c window \u201d size for a fourier analysis is selected , if it is wide , then the frequency values are not time - specific , but represent average signal harmonic content over the entire window time duration . and if the window is narrow , then the harmonics are too spread out in frequency , and the frequency - domain data generated by the transform does not resolve things finely enough in the frequency domain . for all of the above reasons , the present invention analyzes the waveforms using a discrete wavelet transform . this has the advantage that while low frequency wavelets are broad in time , high frequency wavelets have a much narrower timeslot focus . thus , for example , with 32 data samples taken sequentially over time , a wavelet transformation generates a d . c . wavelet that spans the entire set of 32 points ; a first reversing wavelet that also spans the entire set of points ; two second harmonic wavelets that each focus upon only one - half of the 32 data points ; four third harmonic wavelets that each focus upon only one - fourth of the 32 data points ; eight fourth harmonic wavelets that each focus upon four of the 32 data points ; and 16 fifth harmonic wavelets that each focuses upon only two of the 32 data points . accordingly , the higher - frequency wavelet amplitudes are quite time specific , unlike fourier sinusoidal amplitudes , while the lower - frequency and d . c . wavelet amplitudes give one broad information about the whole set of 32 points . and like the fourier transform , the discrete wavelet transform preserves all of the information of the original atrial signal , such that the transformation is fully reversible . 32 discrete wavelet transform values may be reverse transformed back into 32 values indicating the atrial signal potentials at 32 sequential points in time . in essence , the haar function is performing multiple digital filtering operations , at variable positions and utilizing varying - width windowing functions , to develop component values that represent varying types of heartbeat activity , at varying frequencies , spread over varying widths , and located at varying positions . these component values can be studied to see which subset of these component values are good at distinguishing st from non - st , or ( more generally ) which subset of these component values are good at distinguishing one type of heartbeat waveform from another . further study of such a selected subset can further determine which of the subset of components are relatively invariant from one individual &# 39 ; s heart to another . one preferably selects component values that are suitable from both of these two perspectives . the particular wavelet transformation to choose can be tailored to the nature of the data , with the wavelets chosen to resemble somewhat the values to be found within the data so that some transformed values are of much larger amplitude than others , and such that the low amplitude transformed values may be disregarded . of particular importance with an atrial p waveform of the type being analyzed here are wavelets having frequency values roughly comparable to and timed to coincide with the ringing of the atrial depolarization . this tailoring can minimize the number of transformed data values that are significant and that are candidates for full participation in the correlation analyses to determine if there has been a substantial change in waveform morphology . another factor in selecting a particular wavelet transformation is reducing the number of computations that must be performed to carry out the transformation . yet another factor is selecting a wavelet transformation where some of the transformed component values vary from normal to abnormal heart rhythms in much the same over a population of individuals so that a generic template of these component values may be derived that can be used with many individuals , rather than with just one individual . these factors suggest that the haar transformation would be a suitable candidate for use in analysis of atrial waveforms and possibly other waveforms as well . with reference to fig3 in the case of 32 voltage values sampled over time , the 32 applicable haar discrete transform wavelets appear as shown in this figure . other discrete wavelet transforms based upon wavelets having triangular or other shapes may also prove usable . the haar transform wavelets , in particular , have the shape of a square waveform , as will be described , and this can simplify the computations required . in fig3 time increases from left to right . a scale 304 indicates the 32 points in time at which the atrial waveform is sampled , and an analog atrial waveform ( before digitization ) is shown at 302 . the p wave negative notch 303 is shown centered at the 16 th timeslot so that the signal potential of the atria waveform sampled at this point in time becomes the 16 th data value in the set of 32 data values that are to be subjected to the haar transformation . a haar wavelet is simply one cycle of a square waveform that starts at zero , then swings positive ( to \u201c+ 1 \u201d), then swings negative ( to \u201c\u2212 1 \u201d), and then swings back to zero . the wavelet w 2 , shown at 308 in fig3 for example , is at \u201c+ 1 \u201d at points in time 1 to 16 and is at \u201c\u2212 1 \u201d at points in time 17 to 32 . the shorter waveform w 6 , shown at 316 in fig3 is at \u201c+ 1 \u201d at points in time 9 to 12 and is at \u201c\u2212 1 \u201d at points in time 13 to 16 , and is at \u201c 0 \u201d at all other points in time . the waveform w 1 , shown at 306 in fig3 is so slow to fluctuate in time that its \u201c\u2212 1 \u201d portion is off of the chart ( in fig3 ) to the right , and is ignored ; and accordingly , it has the value \u201c+ 1 \u201d for all 32 of the points in time shown in fig3 . it thus represents the average , or d . c ., component of the atrial signal potential . the lowest frequency wavelets w 1 and w 2 encompass the entire set of 32 points in time . the higher frequency wavelets w 3 and w 4 each encompasses only half of the points in time , and the two wavelets taken together encompass all the points in time . the four wavelets w 5 , w 6 , w 7 , and w 8 each encompasses only eight points in time , and the four wavelets taken together encompass all points in time . the eight wavelets w 9 . . . w 16 each encompasses only four points in time , and together all eight wavelets encompass all points in time . and finally , the wavelets w 17 . . . w 32 each encompass only two points in time , but the sixteen wavelets together encompass all points in time . thus , each wavelet has a characteristic time span and position as well as a characteristic frequency , with higher - frequency wavelets having a narrower and more specific time span and position than lower - frequency wavelets . this is advantageous with an impulse - type , ringing signal such as the atrial depolarization waveform considered here , since many higher - frequency transform values that correspond to haar wavelets positioned in time away from the p waveform or that do not correspond to its frequency of ringing may be low in amplitude such that they may safely be ignored during the analysis steps , thereby reducing the data that must be processed during the correlation steps . ( the above , while presented in the context of the haar wavelet , is also applicable to other wavelet shapes that may used to perform a discrete wavelet transform ( dwt )). each of the 32 haar transformed values is computed as follows : multiply each of the 32 sampled and digitized voltage values representing the analog atrial waveform 302 by the correspondingly - positioned -( in - time ) amplitude values of one of the 32 haar wavelets ( shown in fig3 ); then sum the resulting products ; and then multiply the resulting sums by a discrete wavelet transform scaling factor 2 \u2212 j / 2 , where j equals 1 for the wavelets w 17 to w 32 , 2 for w 9 to w 16 , 3 for w 5 to w 8 , 4 for w 3 to w 4 , and 5 for w 1 and w 2 . for example , and referring to fig3 : the first wavelet w 1 is always + 1 , so the atrial signal potential values are simply summed and then multiplied by 2 \u2212 5 / 2 ; the second wavelet is + 1 for time points 1 to 16 and \u2212 1 for time points 17 to 32 , so the first 16 values are summed , the second 16 values are summed , the difference between the first and second sums is computed , and the result is multiplied by 2 \u2212 5 / 2 ; and so on until all the transformed values have been computed , one for each haar wavelet . the 32 time - sequential atrial signal potential values are thus transformed into 32 haar wavelet amplitude values which may be called \u201c transformed values \u201d and which may be assigned the numbers 1 to 32 corresponding to the subscript numbers of the haar wavelets to which they each correspond and whose amplitudes they represent . the above description of how to compute the haar wavelet transformed values is accurate , but it is not the most efficient way to proceed in the case of haar wavelets . like the fourier transform , which has a corresponding fast fourier transform that saves intermediate results and thereby avoids re - computing them and thus reduces substantially the number of computations , the haar transformation also may be carried out in a manner that saves intermediate sums and re - uses them to reduce substantially the number of computations . this is described below at the point where fig4 is described , since fig4 illustrates graphically how this can be done . the illustrative program listing presented below is also an implementation of the fast haar wavelet transformation algorithm . the haar transformation can readily be carried out by any digital computer . as an example of how the haar transformation can be carried out on 32 values , the following program , written in the language c , is illustrative of many possible programs that may be written . in the exemplary program that follows , a 32 - element array yy contains 32 data values representing the 32 sampled atrial signal potential values representing the fluctuations over time of the signal supplied by the bipolar lead 102 . the analog atrial signal is sampled , digitized , broken into separate beat data sets , pre - processed ( to remove waveforms distorted by ventricular activity ), centered ( with the negative depolarization spike at data point 16 ), and fed into the subroutine presented below . this subroutine is compiled ( or assembled , if rewritten in assembly language ) and installed in the rom of the icd &# 39 ; s embedded microprocessor . this subroutine returns , contained within the same array yy , the 32 transformed haar wavelet amplitude values described above . ( the constant value \u201c recipsqrttwo \u201d is the reciprocal of the square root of two ). an illustrative version of the subroutine for computing all 32 of the haar transformed values in an efficient manner is presented here : void haar ( double yy []) { int i , j , l ; double zz [ 32 ]; for ( i = 5 ; i & gt ; 0 ; i \u2212\u2212) { l = 1 ; for ( j = 1 ; j & lt ;= i ; ++ j ) l = 2 * l ; for ( j = 0 ; j & lt ; l ; ++ j ) zz [ j ] = yy [ j ]; for ( j = 0 ; j & lt ; l \u2212 1 ; j = j + 2 ) ( yy [ j / 2 ] = recipsqrttwo * ( zz [ j ] + zz [ j + 1 ]); yy [( j + l )/ 2 ] = rcipsqrttwo * ( zz [ j ] \u2212 zz [ j + 1 ]); } } return ; } the above program computes the 32 haar transformed values 1 to 32 from the 32 atrial signal potential time - sequenced input values . it does so with only 31 additions , 31 subtractions , and 62 multiplications . it is carefully designed to compute each value in a systematic way , making multiple use of intermediate results . first , the program computes sixteen sums of and sixteen differences between eight adjacent pairs of the 16 time sequenced atrial signal potential values , and the sixteen difference values become the haar transformed values 17 to 32 , reflecting the strength of the highest frequency haar wavelets positioned at 16 different positions in time , corresponding to the wavelets w 17 to w 32 shown in fig3 . all the sum and difference values are scaled by multiplication by 2 \u2212 1 / 2 . next , taking the 16 sums of atrial signal potential values as an intermediate result , the above algorithm generates eight sums of and eight differences between adjacent pairs of these sixteen intermediate values that resulted from the first sixteen additions , and the eight newly - computed difference values become the haar transformed values 9 to 16 , reflecting the strength of the second to the highest frequency values at eight different points in time , corresponding to wavelets w 9 to w 16 in fig3 . all of these eight sum and eight difference values are again scaled by 2 \u2212 1 / 2 so that the wavelets w 9 to w 16 are scaled by \u00bd ( 2 \u2212 2 / 2 or 2 \u2212 1 / 2 times 2 \u2212 1 / 2 ). next , taking these eight sums of sums of adjacent atrial signal potential values as an intermediate result , the above algorithm generates four sum and four difference values , again scaling by 2 \u2212 1 / 2 , and the four difference values become haar transformed values 5 to 8 , reflecting the strength of the middle frequency wavelets at four different points in time , and corresponding to wavelets w 5 to w 8 in fig3 . next , taking these four sums of sums of sums of adjacent atrial signal potential values , the above algorithm generates two sum and two difference values , again scaling by 2 \u2212 1 / 2 , and the two difference values become haar transformed values 3 and 4 , reflecting the strength of the second to the lowest frequency wavelets only two of which encompass all the time domain data , corresponding to the wavelets w 3 and w 4 shown in fig3 . and finally , the algorithm generates the sum of and the difference between the final remaining two sums of sums of sums of sums of atrial signal potential values , again scaling by 2 \u2212 1 / 2 . the difference value is then the haar transformed value 2 , representing the strength of the lowest - frequency wavelet , the one corresponding to the wavelet w 2 in fig3 the wavelet that extends the full length of the time scale . the sum value is then the first , or d . c ., transformed value , representing the average signal potential level over the 32 sampled points in time , which corresponds to the wavelet w 1 in fig3 . another way of viewing this computation is illustrated in fig4 a and 4b . the 32 atrial signal potential values are shown at the top of fig4 a and are identified as c 0 5 through c 31 5 . the superscript \u201c 5 \u201d indicates that these values are processed when the index value \u201c n \u201d in the above computer program is equal to \u201c 5 \u201d\u2014 that is , during the first pass through the data generating intermediary sums and differences . during subsequent passes , the value of n is decremented to 4 , 3 , 2 , and finally to 1 . during each pass through the data , the computer generates sums c n - 1 and differences d n - 1 , as shown in fig4 b , between adjacent pairs of the values c 0 n and c 1 n ; c 2 n and c 3 n ; and so on , so that the number of newly - generated c n - 1 and d n - 1 terms is reduced by half with each computer pass through the intermediary results . at the end of all these computations , the value c 0 0 is the first , or d . c ., haar transformed value ; and the values d 0 0 ; d 0 1 and d 1 1 ; d 0 2 , d 1 2 , d 2 2 , and d 3 2 ; d 0 3 , d 1 3 , . . . and d 7 3 ; and d 0 4 , d 1 4 , and d 15 4 are , respectively , the remaining haar transform values 2 through 32 . fig4 thus illustrates quite succinctly how all the transform computations are carried out , and how the intermediary \u201c c \u201d sum values , such as c 0 4 , are used to compute multiple haar values , such as the values d 0 3 , d 0 2 , d 0 1 , d 0 0 , and c 0 0 all of which are computed from the intermediary value c 0 4 . saving and reusing these intermediary \u201c c \u201d values saves much computational time . fig4 b indicates the precise addition and subtraction operations that are carried out at each level to compute the values in the next lower level , proceeding down through the chart presented in fig4 a . but in any given application to heart waveform analysis , all of these computations may not be needed , and accordingly the number of computations may be reduced much further . since the present invention teaches that only a small number of these haar transformed values need actually be considered , a far less computationally intensive transform can be developed which only generates the intermediate and final transformed values that are actually needed to generate the specific haar transformed values which have proved to be significant in discriminating between sinus tachycardia ( or st ) and non - st conditions . all others need not be computed , and the above program may be reduced to a special algorithm that omits as many sums , differences , and multiplications as possible . for example , if only the transformed values 1 , 5 , 9 , and 24 are significant , then 31 additions are required to compute the first transformed value ( the d . c . value \u2014 the simple sum of all the time domain signal values ); six additions and one subtraction are required to compute the fifth transformed value ( in fig3 the difference between the sums of time domain values under each half of the wavelet w 5 at 314 in fig3 ); two additions and one subtraction are required to compute the ninth transformed value ( in fig3 the difference between the sums of the time domain values under each half of the wavelet w9 at 322 in fig3 ); and only one subtraction is required to compute the 24 th transformed value ( in fig3 the difference between the two time domain values under the respective halves of the wavelet w 24 ( not shown ) which is the same size as , but differently positioned in time than , the wavelet w 18 at 330 ( fig3 ). accordingly , if only the transformed values 1 , 5 , 9 , and 24 actually evaluated , then only 42 additions and subtractions are required . but even this number can be reduced further . if the computational algorithm set forth in the above illustrative computer program is followed as a guide , then the intermediary sums used in computing the haar first , or d . c ., transformed value can be re - used to compute the sums for the haar transformed values 5 and 9 , assuming these intermediary results are saved in the manner described in the above program example . then 31 additions are still required to compute the first transformed value , but only one subtraction is required to compute each of the transformed values 5 , 9 , and 24 , giving a total of additions and subtraction of only 34 operations . and if only transformed values 1 , 5 , and 9 are computed , the number of additions and subtractions is reduced to 33 . and if the first transformed value is omitted and if transformed values 5 , 9 , and 24 are selected , then the 31 additions needed to compute the first transformed value are not required . then the number of computations for transformed value 5 is 8 additions and one subtraction ; the number of computations for transformed value 9 is 2 additions and one subtraction ; and the number of computations for transformed value 24 is still just one subtraction . so the total number of additions and subtractions is just 13 . but even this number can be reduced to 11 when it is realized that the two additions done for the transformed value 9 are also done ( in the above computational algorithm ) when computing the transformed value 5 . so the number of additions and subtractions can be reduced to 11 . also , when computing such a small number of transformed values , the number of multiplications can be reduced as well by postponing the multiplications until a transformed value is actually computed . with reference to fig4 a , instead of dividing every sum and difference by the square root of two ( as shown in fig4 b ), several vertical sums of \u201c c \u201d values in different rows of the fig4 a table can be formed , and then a \u201c d \u201d value can be computed by subtraction , and then the resulting unscaled transformed value can be scaled with a single multiplication by 2 \u2212 j / 2 where j is the number of rows in the table of fig4 a traversed by the one or more sum and the one difference computations . thus , the number of scaling multiplications can sometimes be equal to the number of transformed values that are computed , typically 3 or 4 . accordingly , by using wavelet analysis and transformation , instead of fourier or discrete cosine or other sinusoid analysis and transformation , a highly useful and highly frequency - specific result can be achieved with a very small number of computations , thereby conserving power and battery life , and yet achieving a high degree of precision in recognizing changes in morphology . in our tests , we have selected certain transformed values as being much more significant than other values in distinguishing between st and non - st . we focused upon those transform values whose \u201c+ 1 \u201d and \u201c\u2212 1 \u201d scope included the middle 8 time points most significant to analysis of the peak of the atrial depolarization . working across test data samples obtained from patients who had dual chamber icds , we computed the variance of each transformed value for st and non - st events using the standard statistical formula for computing variance . we then calculated the ratio of the variance of non - st events to the variance of st events , and selected those terms with the highest variance ratios for further consideration . in this manner , we reduced the number of transformed values that were included in the cwa or correlation process while always maintaining or improving the performance achieved . ultimately we settled upon the three or four transformed values having the highest ratios . these were the transform values 1 , 5 , 9 , and 24 . we tested 1 , 5 , and 24 together ; 5 , 9 , and 24 together ; and 1 , 5 , 9 , and 24 together . these selected haar transform values , when used in these combinations , required minimal computations and thereby produced a savings in battery power , and that also gave better performance at distinguishing st from non - st than did all of the transformed values used together . so an increase in accuracy was achieved as well as a decrease in computational complexity by this approach to atrial waveform analysis . [ 0054 ] fig5 for example , illustrates actual plots of digital information illustrating the effect of using only the three coefficients 1 , 5 , and 24 . at 502 , a normal waveform is shown , with amplitude plotted against sample numbers from 1 to 32 . at 504 , an abnormal waveform is shown , again with amplitude plotted against sample numbers . after the haar transformation , at 506 and 508 , the wavelet coefficient amplitude values are indicated for the coefficients 1 to 32 . at 506 , the coefficients generated by the normal waveform 502 are shown , and at 508 , the coefficients generated by the abnormal waveform 504 are shown . one may directly compare these component values and verify that the coefficients 1 , 5 , and 24 at 506 and at 508 vary significantly in amplitude . comparison of this same information among different patients ( not shown in this figure ) also indicated that this variation is relatively constant from one patient to another . at 510 and 512 , using only the coefficients 1 , 5 , and 24 , the heartbeat waveforms are reconstructed by a reverse transformation , the normal reconstructed waveform shown at 510 and the abnormal reconstructed waveform shown at 512 . the marked differences between these two reconstructed waveforms highlights the way in which these three coefficients can signal an abnormal condition with less computation . in a practical system , a small number of transform values are selected , in the manner just described . the algorithm for computing these values is then refined , as explained above , to reduce the number of additions , subtractions , and multiplications to the minimum possible while preserving the accuracy of the computations . we first compute a value \u03c1 , which is the correlation between these values with the corresponding values in a template that represents the values for an average normal population . we then compare this value \u03c1 to a threshold correlation value \u03b2 that is chosen to give optimal results on experimental data . ( see the examples in the tables presented below .) for each waveform analyzed , a test is made to determine whether \u03c1 is greater than \u03b2 . if so , then this waveform is placed into the buffer marked \u201c st \u201d at step 210 . if not , then this waveform is placed into the buffer marked \u201c non - st \u201d at step 211 . finally , after ten waveforms have been analyzed , a count of st and non - st marked waveforms is made to see if the count of non - st waveforms is greater than some threshold value x ( at step 212 ), where x can be , for example , 7 . if more than seven waveforms are non - st , then therapy is delivered at step 214 . the buffers at steps 210 and 211 in fig2 are buffers that hold the last ten decision values . these buffers can be pictured as sliding windows revealing the most recent ten values to permit the decision at 212 to be continuously updated . the buffers thus function as a nonlinear filter preventing irregular values from delivering therapy improperly . in developing a working prototype of this system , we used a development data set consisting of 20 episodes of st taken from 4 patients and 18 episodes of af taken from 7 patients . patient number episodes of st episodes of af 1 2 6 2 12 0 13 0 1 4 0 2 5 5 1 6 0 3 7 0 3 8 0 2 9 1 0 total 20 18 we used this data to optimize our choice of \u03c1 and \u03b2 and also the particular coefficients that we chose to examine . next , we tested our prototype using a set of 16 episodes of st obtained from 4 patients together with 17 episodes of af obtained from 7 patients . patient number episodes of st episodes of af 1 8 0 2 3 1 3 3 0 4 2 0 5 0 4 6 0 4 7 0 3 8 0 3 9 0 1 10 0 1 total 16 17 x out threshold coefficients of 10 ( \u03b2 ) sensitivity specificity 1 , 5 , 9 , & amp ; 24 3 0 . 808 76 % 94 % 1 , 5 , 9 , & amp ; 24 5 0 . 975 88 % 94 % 1 , 5 , 9 , & amp ; 24 6 0 . 976 82 % 94 % 1 , 5 , 9 , & amp ; 24 7 0 . 983 88 % 94 % 1 , 5 , & amp ; 24 3 0 . 815 82 % 94 % 1 , 5 , & amp ; 24 4 0 . 943 88 % 94 % 1 , 5 , & amp ; 24 5 0 . 956 82 % 94 % 1 , 5 , & amp ; 24 7 0 . 991 94 % 94 % 5 , 9 , & amp ; 24 8 0 . 9993 82 % 94 % 5 , 9 , & amp ; 24 9 0 . 9997 76 % 81 % the quality of the selected subset of transformed values for characterizing changes in morphology may be demonstrated graphically , as indicated in fig5 by performing a reverse haar transform using only the three or four or so selected transformed values . as can be seen ( at 510 and 512 in fig5 ), the inverse transformations produce distorted representations of the original atrial waveforms which illustrate how the choice of transform values can emphasize the changes . this is another useful way to assist one in selecting which transformed values are most useful . one feature of the invention is its ability to use a template derived from an average normal population , rather than deriving a patient customized average normal template for each individual patient . prior systems have required each new patient to be monitored while in a normal state , and the captured data was then averaged to form a patient specific normal template . contrary to this , the present invention achieved the results shown above using the same template for all patients . accordingly , the invention may be used with a new patient without the necessity of such preliminary testing of each patient and without a new customized template necessarily having to be created for each patient . one reason why the present invention can function using a template that represents average values for a normal population is because the specific coefficients selected in the transform domain , in addition to having been selected to emphasize the difference between normal and abnormal values , are also preferably chosen to minimize the differences between different patients . in some cases , values that were good at distinguishing between normal and abnormal rhythms for one patient were not as good at doing so for some other patient . these values are preferably not selected for use in implementing the present invention . for example , the transform coefficients may be selected such that all ( or most of ) the normal values of a particular coefficient gathered from many patients had the same sign ( positive or negative ) and similar amplitudes ( or if they varied in sign , they were near to zero ). in addition , the abnormal coefficients are significantly different in value from the normal coefficients for each particular patient . [ 0068 ] fig6 at 600 , illustrates one way in which this may be done . first , at step 602 , one gathers a large number of st and non - st data samples . next , all of this data is transformed , generating coefficients for every waveform set of data ( step 604 ). the variance of each coefficient is then computed first for the st samples ( step 606 ) and then for the non - st samples . then , for each coefficient , the ratio of the variance of the non - st samples to that of the st samples is computed ( step 608 ). finally , those coefficients whose variance ratio was large may be selected a coefficients for use in creating a population template and in testing patients ( step 612 ). in addition , at step 614 , the number of coefficients retained may be further reduced to reduce the number of computations that need to be performed , as explained above . in the version of the invention that was used to generate the data shown above , the population template was computed as follows , using the 20 st episodes in the development data set ( also used in the first of the two tables presented above ). this is illustrated at 700 in fig7 . first , the coefficients were selected as was explained above and as shown in fig6 . next , for each patient episode of st ( step 702 ), the recorded heartbeats were broken up into groups of ten consecutive heartbeats ( step 704 ). for each group ( step 706 ), the selected coefficients of the haar transform were computed ( step 708 ) for each of the ten heartbeats and the median value was then selected ( step 710 ) for each coefficient to form a proto - template . if these median values correlated well with the coefficient values for each of the ten heartbeats , the proto - template was retained ( step 714 ). otherwise , it was discarded . in this manner , a whole bunch of proto - templates were obtained from each patient episode . next , median values of the coefficients from all of the proto - templates ( step 716 ) were computed ( step 718 ). these values were then used as the population template for distinguishing st from non - st events ( step 720 ). while the preferred embodiment of the invention has been described above , it is to be understood that numerous modifications and changes will occur to those who are skilled in the art to which the invention pertains . accordingly , the following claims annexed to and forming a part of this specification are intended to define the true spirit and scope of the invention \u2014 that is , what is new and what is desired to be secured by letters patent of the united states ."}
|
{"category": "Fixed Constructions", "patent": "the present invention is directed towards improving the quality and functionality of both implantable cardioverter defibrillators ( icds ) and implantable atrial defibrillators ( iads ). all such devices need to have a method and mechanism for accurately distinguishing sinus tachycardia ( st ) form non - st , such as atrial fibrillation and atrial flutter . this method and mechanism should be conservative , biased towards not initiating therapy unless non - st is clearly indicated . it should also require as few machine computational cycles as possible to reduce icd and iad battery drain . the principle underlying the present invention is that st and non - st can be distinguished by studying and measuring the atrial electrogram morphology to detect aberrant conduction in the atria . thus , by comparing normal st waveforms to a series of newly - occurring waveforms , st is indicated by abnormal atrial waveform morphology which signals aberrant conduction within the atria improper electrical patterns of depolarization . this will show up in the signal potentials captured by a bipolar atrial lead . with reference to fig1 the present invention can be implemented in an icd 100 ( which could be an iad ) that is implanted in the chest of a patient who suffers from unpleasant and possibly life - threatening irregularities in heart operation due to improper electrical stimulation of the heart muscle cells . during normal heart action , the heart &# 39 ; s electrical impulses originate in the sino - atrial node as an action potential that is transmitted smoothly to all portions of the atria , causing contraction of the atrial chambers . the electrical impulse continues in its path to a cluster of conduction fibrils known as the atrioventricular node . after a delay of about one - tenth of a second , an action potential flows over the ventricles and causes them to contract in synchronism with and following shortly after the atrial contraction . in this manner , the heart pumps blood to the lungs and from the lungs to the body . a bipolar lead 102 is implanted within the atria 104 of a heart 103 to measure the electrical potential between nearby cells of the atria . as the action potential actively passes from cell to cell past the two electrodes of the lead 102 , an oscillating signal potential is developed across the two electrodes and is conveyed by the bipolar lead 102 to the icd 100 where the signal is sampled about 1000 times per second and is digitized , with the sequential samples stored within a ram memory within the icd 100 for further analysis . another bipolar lead 105 may extend to the ventricle ( not shown ) of the heart 103 to measure the action potentials generated within the ventricles . these same leads , or other leads , may be used by the icd for administering various types of therapy , such as pacing or defibrillation shock therapy . the data samples collected from within the atria are now analyzed . first , with reference to data gathered from the ventricles , if some event in the ventricle , such as a mis - timed depolarization event that partially overlaps the atria &# 39 ; s p wave , could leak across to the atria and distort the measurement of the potentials for a given atrial depolarization , then the data for that particular heartbeat is discarded and is not used in further analyses . such distortion can also be found simply because a given heartbeat set of data gathered from the atria is itself badly distorted , and this approach must be taken in the case of an iad having no ventricular lead . the remaining data is retained for further processing . in preparation for further processing , the data for individual heartbeats is identified and gathered . the position within the gathered data of the negative spike to the atrial p depolarization waveform is determined and is selected as the center of the waveform for each beat . then , since data close to this p - wave is to be analyzed , and since data further away may be distorted to some degree by ventricular activity or by variations in the p - to - p interval , 32 sequential data samples are selected for each heartbeat such that the negative spike of the p - wave becomes signal potential sample number 16 of the 32 sequential samples selected for further analyses . these samples for a normal heartbeat appear at 302 in fig3 and at 502 in fig5 where the amplitude of the atrial signal potential is plotted against time over the 32 sampled values , with the negative p - wave spike positioned as signal sample number 16 . likewise , the samples for an abnormal heartbeat appear at 504 in fig5 . straightforward correlation ( cwa ) of the 32 samples against a template containing a typical or normal pattern could be performed at this juncture , as taught on page 563 ( equation ( 1 )) of the article by thorne , cited in the introductory portion of this specification . but the performance of such a cross correlation at repeated regular intervals would generate numerous computations and would drain the icd &# 39 ; s battery more rapidly than is desirable . in addition , provision would have to be made for a full template waveform that can be stored within the icd to be used as a comparison reference . in addition , such templates generally need to be patient specific , and they are more susceptible to normal changes in shape . accordingly , to reduce the mathematical complexity of the cross correlation computations , it is desirable to reduce the number of data points from 32 down to some much lower number through the use of some form of transformation into a different set of values . for example , the data could be transformed using the karhunen - loeve transformation described in the morris article cited at the beginning of this specification . but that transformation is fairly computationally intensive . computationally , a fast fourier transformation would be more efficient , but any such transformation into the frequency domain , where variable frequency sine and cosine wave amplitudes result from the transformation , does not match itself well to the morphology of atrial heart waveforms , which tend to be formed from large , ringing , spike - like representations of p - wave depolarization events travelling as moving action potential fronts past the pair of electrodes attached to the bipolar lead , and not as steady harmonics extending across time . accordingly , fourier - class transforms do not adequately reduce the number of significant data values that must be considered . and in addition , when the \u201c window \u201d size for a fourier analysis is selected , if it is wide , then the frequency values are not time - specific , but represent average signal harmonic content over the entire window time duration . and if the window is narrow , then the harmonics are too spread out in frequency , and the frequency - domain data generated by the transform does not resolve things finely enough in the frequency domain . for all of the above reasons , the present invention analyzes the waveforms using a discrete wavelet transform . this has the advantage that while low frequency wavelets are broad in time , high frequency wavelets have a much narrower timeslot focus . thus , for example , with 32 data samples taken sequentially over time , a wavelet transformation generates a d . c . wavelet that spans the entire set of 32 points ; a first reversing wavelet that also spans the entire set of points ; two second harmonic wavelets that each focus upon only one - half of the 32 data points ; four third harmonic wavelets that each focus upon only one - fourth of the 32 data points ; eight fourth harmonic wavelets that each focus upon four of the 32 data points ; and 16 fifth harmonic wavelets that each focuses upon only two of the 32 data points . accordingly , the higher - frequency wavelet amplitudes are quite time specific , unlike fourier sinusoidal amplitudes , while the lower - frequency and d . c . wavelet amplitudes give one broad information about the whole set of 32 points . and like the fourier transform , the discrete wavelet transform preserves all of the information of the original atrial signal , such that the transformation is fully reversible . 32 discrete wavelet transform values may be reverse transformed back into 32 values indicating the atrial signal potentials at 32 sequential points in time . in essence , the haar function is performing multiple digital filtering operations , at variable positions and utilizing varying - width windowing functions , to develop component values that represent varying types of heartbeat activity , at varying frequencies , spread over varying widths , and located at varying positions . these component values can be studied to see which subset of these component values are good at distinguishing st from non - st , or ( more generally ) which subset of these component values are good at distinguishing one type of heartbeat waveform from another . further study of such a selected subset can further determine which of the subset of components are relatively invariant from one individual &# 39 ; s heart to another . one preferably selects component values that are suitable from both of these two perspectives . the particular wavelet transformation to choose can be tailored to the nature of the data , with the wavelets chosen to resemble somewhat the values to be found within the data so that some transformed values are of much larger amplitude than others , and such that the low amplitude transformed values may be disregarded . of particular importance with an atrial p waveform of the type being analyzed here are wavelets having frequency values roughly comparable to and timed to coincide with the ringing of the atrial depolarization . this tailoring can minimize the number of transformed data values that are significant and that are candidates for full participation in the correlation analyses to determine if there has been a substantial change in waveform morphology . another factor in selecting a particular wavelet transformation is reducing the number of computations that must be performed to carry out the transformation . yet another factor is selecting a wavelet transformation where some of the transformed component values vary from normal to abnormal heart rhythms in much the same over a population of individuals so that a generic template of these component values may be derived that can be used with many individuals , rather than with just one individual . these factors suggest that the haar transformation would be a suitable candidate for use in analysis of atrial waveforms and possibly other waveforms as well . with reference to fig3 in the case of 32 voltage values sampled over time , the 32 applicable haar discrete transform wavelets appear as shown in this figure . other discrete wavelet transforms based upon wavelets having triangular or other shapes may also prove usable . the haar transform wavelets , in particular , have the shape of a square waveform , as will be described , and this can simplify the computations required . in fig3 time increases from left to right . a scale 304 indicates the 32 points in time at which the atrial waveform is sampled , and an analog atrial waveform ( before digitization ) is shown at 302 . the p wave negative notch 303 is shown centered at the 16 th timeslot so that the signal potential of the atria waveform sampled at this point in time becomes the 16 th data value in the set of 32 data values that are to be subjected to the haar transformation . a haar wavelet is simply one cycle of a square waveform that starts at zero , then swings positive ( to \u201c+ 1 \u201d), then swings negative ( to \u201c\u2212 1 \u201d), and then swings back to zero . the wavelet w 2 , shown at 308 in fig3 for example , is at \u201c+ 1 \u201d at points in time 1 to 16 and is at \u201c\u2212 1 \u201d at points in time 17 to 32 . the shorter waveform w 6 , shown at 316 in fig3 is at \u201c+ 1 \u201d at points in time 9 to 12 and is at \u201c\u2212 1 \u201d at points in time 13 to 16 , and is at \u201c 0 \u201d at all other points in time . the waveform w 1 , shown at 306 in fig3 is so slow to fluctuate in time that its \u201c\u2212 1 \u201d portion is off of the chart ( in fig3 ) to the right , and is ignored ; and accordingly , it has the value \u201c+ 1 \u201d for all 32 of the points in time shown in fig3 . it thus represents the average , or d . c ., component of the atrial signal potential . the lowest frequency wavelets w 1 and w 2 encompass the entire set of 32 points in time . the higher frequency wavelets w 3 and w 4 each encompasses only half of the points in time , and the two wavelets taken together encompass all the points in time . the four wavelets w 5 , w 6 , w 7 , and w 8 each encompasses only eight points in time , and the four wavelets taken together encompass all points in time . the eight wavelets w 9 . . . w 16 each encompasses only four points in time , and together all eight wavelets encompass all points in time . and finally , the wavelets w 17 . . . w 32 each encompass only two points in time , but the sixteen wavelets together encompass all points in time . thus , each wavelet has a characteristic time span and position as well as a characteristic frequency , with higher - frequency wavelets having a narrower and more specific time span and position than lower - frequency wavelets . this is advantageous with an impulse - type , ringing signal such as the atrial depolarization waveform considered here , since many higher - frequency transform values that correspond to haar wavelets positioned in time away from the p waveform or that do not correspond to its frequency of ringing may be low in amplitude such that they may safely be ignored during the analysis steps , thereby reducing the data that must be processed during the correlation steps . ( the above , while presented in the context of the haar wavelet , is also applicable to other wavelet shapes that may used to perform a discrete wavelet transform ( dwt )). each of the 32 haar transformed values is computed as follows : multiply each of the 32 sampled and digitized voltage values representing the analog atrial waveform 302 by the correspondingly - positioned -( in - time ) amplitude values of one of the 32 haar wavelets ( shown in fig3 ); then sum the resulting products ; and then multiply the resulting sums by a discrete wavelet transform scaling factor 2 \u2212 j / 2 , where j equals 1 for the wavelets w 17 to w 32 , 2 for w 9 to w 16 , 3 for w 5 to w 8 , 4 for w 3 to w 4 , and 5 for w 1 and w 2 . for example , and referring to fig3 : the first wavelet w 1 is always + 1 , so the atrial signal potential values are simply summed and then multiplied by 2 \u2212 5 / 2 ; the second wavelet is + 1 for time points 1 to 16 and \u2212 1 for time points 17 to 32 , so the first 16 values are summed , the second 16 values are summed , the difference between the first and second sums is computed , and the result is multiplied by 2 \u2212 5 / 2 ; and so on until all the transformed values have been computed , one for each haar wavelet . the 32 time - sequential atrial signal potential values are thus transformed into 32 haar wavelet amplitude values which may be called \u201c transformed values \u201d and which may be assigned the numbers 1 to 32 corresponding to the subscript numbers of the haar wavelets to which they each correspond and whose amplitudes they represent . the above description of how to compute the haar wavelet transformed values is accurate , but it is not the most efficient way to proceed in the case of haar wavelets . like the fourier transform , which has a corresponding fast fourier transform that saves intermediate results and thereby avoids re - computing them and thus reduces substantially the number of computations , the haar transformation also may be carried out in a manner that saves intermediate sums and re - uses them to reduce substantially the number of computations . this is described below at the point where fig4 is described , since fig4 illustrates graphically how this can be done . the illustrative program listing presented below is also an implementation of the fast haar wavelet transformation algorithm . the haar transformation can readily be carried out by any digital computer . as an example of how the haar transformation can be carried out on 32 values , the following program , written in the language c , is illustrative of many possible programs that may be written . in the exemplary program that follows , a 32 - element array yy contains 32 data values representing the 32 sampled atrial signal potential values representing the fluctuations over time of the signal supplied by the bipolar lead 102 . the analog atrial signal is sampled , digitized , broken into separate beat data sets , pre - processed ( to remove waveforms distorted by ventricular activity ), centered ( with the negative depolarization spike at data point 16 ), and fed into the subroutine presented below . this subroutine is compiled ( or assembled , if rewritten in assembly language ) and installed in the rom of the icd &# 39 ; s embedded microprocessor . this subroutine returns , contained within the same array yy , the 32 transformed haar wavelet amplitude values described above . ( the constant value \u201c recipsqrttwo \u201d is the reciprocal of the square root of two ). an illustrative version of the subroutine for computing all 32 of the haar transformed values in an efficient manner is presented here : void haar ( double yy []) { int i , j , l ; double zz [ 32 ]; for ( i = 5 ; i & gt ; 0 ; i \u2212\u2212) { l = 1 ; for ( j = 1 ; j & lt ;= i ; ++ j ) l = 2 * l ; for ( j = 0 ; j & lt ; l ; ++ j ) zz [ j ] = yy [ j ]; for ( j = 0 ; j & lt ; l \u2212 1 ; j = j + 2 ) ( yy [ j / 2 ] = recipsqrttwo * ( zz [ j ] + zz [ j + 1 ]); yy [( j + l )/ 2 ] = rcipsqrttwo * ( zz [ j ] \u2212 zz [ j + 1 ]); } } return ; } the above program computes the 32 haar transformed values 1 to 32 from the 32 atrial signal potential time - sequenced input values . it does so with only 31 additions , 31 subtractions , and 62 multiplications . it is carefully designed to compute each value in a systematic way , making multiple use of intermediate results . first , the program computes sixteen sums of and sixteen differences between eight adjacent pairs of the 16 time sequenced atrial signal potential values , and the sixteen difference values become the haar transformed values 17 to 32 , reflecting the strength of the highest frequency haar wavelets positioned at 16 different positions in time , corresponding to the wavelets w 17 to w 32 shown in fig3 . all the sum and difference values are scaled by multiplication by 2 \u2212 1 / 2 . next , taking the 16 sums of atrial signal potential values as an intermediate result , the above algorithm generates eight sums of and eight differences between adjacent pairs of these sixteen intermediate values that resulted from the first sixteen additions , and the eight newly - computed difference values become the haar transformed values 9 to 16 , reflecting the strength of the second to the highest frequency values at eight different points in time , corresponding to wavelets w 9 to w 16 in fig3 . all of these eight sum and eight difference values are again scaled by 2 \u2212 1 / 2 so that the wavelets w 9 to w 16 are scaled by \u00bd ( 2 \u2212 2 / 2 or 2 \u2212 1 / 2 times 2 \u2212 1 / 2 ). next , taking these eight sums of sums of adjacent atrial signal potential values as an intermediate result , the above algorithm generates four sum and four difference values , again scaling by 2 \u2212 1 / 2 , and the four difference values become haar transformed values 5 to 8 , reflecting the strength of the middle frequency wavelets at four different points in time , and corresponding to wavelets w 5 to w 8 in fig3 . next , taking these four sums of sums of sums of adjacent atrial signal potential values , the above algorithm generates two sum and two difference values , again scaling by 2 \u2212 1 / 2 , and the two difference values become haar transformed values 3 and 4 , reflecting the strength of the second to the lowest frequency wavelets only two of which encompass all the time domain data , corresponding to the wavelets w 3 and w 4 shown in fig3 . and finally , the algorithm generates the sum of and the difference between the final remaining two sums of sums of sums of sums of atrial signal potential values , again scaling by 2 \u2212 1 / 2 . the difference value is then the haar transformed value 2 , representing the strength of the lowest - frequency wavelet , the one corresponding to the wavelet w 2 in fig3 the wavelet that extends the full length of the time scale . the sum value is then the first , or d . c ., transformed value , representing the average signal potential level over the 32 sampled points in time , which corresponds to the wavelet w 1 in fig3 . another way of viewing this computation is illustrated in fig4 a and 4b . the 32 atrial signal potential values are shown at the top of fig4 a and are identified as c 0 5 through c 31 5 . the superscript \u201c 5 \u201d indicates that these values are processed when the index value \u201c n \u201d in the above computer program is equal to \u201c 5 \u201d\u2014 that is , during the first pass through the data generating intermediary sums and differences . during subsequent passes , the value of n is decremented to 4 , 3 , 2 , and finally to 1 . during each pass through the data , the computer generates sums c n - 1 and differences d n - 1 , as shown in fig4 b , between adjacent pairs of the values c 0 n and c 1 n ; c 2 n and c 3 n ; and so on , so that the number of newly - generated c n - 1 and d n - 1 terms is reduced by half with each computer pass through the intermediary results . at the end of all these computations , the value c 0 0 is the first , or d . c ., haar transformed value ; and the values d 0 0 ; d 0 1 and d 1 1 ; d 0 2 , d 1 2 , d 2 2 , and d 3 2 ; d 0 3 , d 1 3 , . . . and d 7 3 ; and d 0 4 , d 1 4 , and d 15 4 are , respectively , the remaining haar transform values 2 through 32 . fig4 thus illustrates quite succinctly how all the transform computations are carried out , and how the intermediary \u201c c \u201d sum values , such as c 0 4 , are used to compute multiple haar values , such as the values d 0 3 , d 0 2 , d 0 1 , d 0 0 , and c 0 0 all of which are computed from the intermediary value c 0 4 . saving and reusing these intermediary \u201c c \u201d values saves much computational time . fig4 b indicates the precise addition and subtraction operations that are carried out at each level to compute the values in the next lower level , proceeding down through the chart presented in fig4 a . but in any given application to heart waveform analysis , all of these computations may not be needed , and accordingly the number of computations may be reduced much further . since the present invention teaches that only a small number of these haar transformed values need actually be considered , a far less computationally intensive transform can be developed which only generates the intermediate and final transformed values that are actually needed to generate the specific haar transformed values which have proved to be significant in discriminating between sinus tachycardia ( or st ) and non - st conditions . all others need not be computed , and the above program may be reduced to a special algorithm that omits as many sums , differences , and multiplications as possible . for example , if only the transformed values 1 , 5 , 9 , and 24 are significant , then 31 additions are required to compute the first transformed value ( the d . c . value \u2014 the simple sum of all the time domain signal values ); six additions and one subtraction are required to compute the fifth transformed value ( in fig3 the difference between the sums of time domain values under each half of the wavelet w 5 at 314 in fig3 ); two additions and one subtraction are required to compute the ninth transformed value ( in fig3 the difference between the sums of the time domain values under each half of the wavelet w9 at 322 in fig3 ); and only one subtraction is required to compute the 24 th transformed value ( in fig3 the difference between the two time domain values under the respective halves of the wavelet w 24 ( not shown ) which is the same size as , but differently positioned in time than , the wavelet w 18 at 330 ( fig3 ). accordingly , if only the transformed values 1 , 5 , 9 , and 24 actually evaluated , then only 42 additions and subtractions are required . but even this number can be reduced further . if the computational algorithm set forth in the above illustrative computer program is followed as a guide , then the intermediary sums used in computing the haar first , or d . c ., transformed value can be re - used to compute the sums for the haar transformed values 5 and 9 , assuming these intermediary results are saved in the manner described in the above program example . then 31 additions are still required to compute the first transformed value , but only one subtraction is required to compute each of the transformed values 5 , 9 , and 24 , giving a total of additions and subtraction of only 34 operations . and if only transformed values 1 , 5 , and 9 are computed , the number of additions and subtractions is reduced to 33 . and if the first transformed value is omitted and if transformed values 5 , 9 , and 24 are selected , then the 31 additions needed to compute the first transformed value are not required . then the number of computations for transformed value 5 is 8 additions and one subtraction ; the number of computations for transformed value 9 is 2 additions and one subtraction ; and the number of computations for transformed value 24 is still just one subtraction . so the total number of additions and subtractions is just 13 . but even this number can be reduced to 11 when it is realized that the two additions done for the transformed value 9 are also done ( in the above computational algorithm ) when computing the transformed value 5 . so the number of additions and subtractions can be reduced to 11 . also , when computing such a small number of transformed values , the number of multiplications can be reduced as well by postponing the multiplications until a transformed value is actually computed . with reference to fig4 a , instead of dividing every sum and difference by the square root of two ( as shown in fig4 b ), several vertical sums of \u201c c \u201d values in different rows of the fig4 a table can be formed , and then a \u201c d \u201d value can be computed by subtraction , and then the resulting unscaled transformed value can be scaled with a single multiplication by 2 \u2212 j / 2 where j is the number of rows in the table of fig4 a traversed by the one or more sum and the one difference computations . thus , the number of scaling multiplications can sometimes be equal to the number of transformed values that are computed , typically 3 or 4 . accordingly , by using wavelet analysis and transformation , instead of fourier or discrete cosine or other sinusoid analysis and transformation , a highly useful and highly frequency - specific result can be achieved with a very small number of computations , thereby conserving power and battery life , and yet achieving a high degree of precision in recognizing changes in morphology . in our tests , we have selected certain transformed values as being much more significant than other values in distinguishing between st and non - st . we focused upon those transform values whose \u201c+ 1 \u201d and \u201c\u2212 1 \u201d scope included the middle 8 time points most significant to analysis of the peak of the atrial depolarization . working across test data samples obtained from patients who had dual chamber icds , we computed the variance of each transformed value for st and non - st events using the standard statistical formula for computing variance . we then calculated the ratio of the variance of non - st events to the variance of st events , and selected those terms with the highest variance ratios for further consideration . in this manner , we reduced the number of transformed values that were included in the cwa or correlation process while always maintaining or improving the performance achieved . ultimately we settled upon the three or four transformed values having the highest ratios . these were the transform values 1 , 5 , 9 , and 24 . we tested 1 , 5 , and 24 together ; 5 , 9 , and 24 together ; and 1 , 5 , 9 , and 24 together . these selected haar transform values , when used in these combinations , required minimal computations and thereby produced a savings in battery power , and that also gave better performance at distinguishing st from non - st than did all of the transformed values used together . so an increase in accuracy was achieved as well as a decrease in computational complexity by this approach to atrial waveform analysis . [ 0054 ] fig5 for example , illustrates actual plots of digital information illustrating the effect of using only the three coefficients 1 , 5 , and 24 . at 502 , a normal waveform is shown , with amplitude plotted against sample numbers from 1 to 32 . at 504 , an abnormal waveform is shown , again with amplitude plotted against sample numbers . after the haar transformation , at 506 and 508 , the wavelet coefficient amplitude values are indicated for the coefficients 1 to 32 . at 506 , the coefficients generated by the normal waveform 502 are shown , and at 508 , the coefficients generated by the abnormal waveform 504 are shown . one may directly compare these component values and verify that the coefficients 1 , 5 , and 24 at 506 and at 508 vary significantly in amplitude . comparison of this same information among different patients ( not shown in this figure ) also indicated that this variation is relatively constant from one patient to another . at 510 and 512 , using only the coefficients 1 , 5 , and 24 , the heartbeat waveforms are reconstructed by a reverse transformation , the normal reconstructed waveform shown at 510 and the abnormal reconstructed waveform shown at 512 . the marked differences between these two reconstructed waveforms highlights the way in which these three coefficients can signal an abnormal condition with less computation . in a practical system , a small number of transform values are selected , in the manner just described . the algorithm for computing these values is then refined , as explained above , to reduce the number of additions , subtractions , and multiplications to the minimum possible while preserving the accuracy of the computations . we first compute a value \u03c1 , which is the correlation between these values with the corresponding values in a template that represents the values for an average normal population . we then compare this value \u03c1 to a threshold correlation value \u03b2 that is chosen to give optimal results on experimental data . ( see the examples in the tables presented below .) for each waveform analyzed , a test is made to determine whether \u03c1 is greater than \u03b2 . if so , then this waveform is placed into the buffer marked \u201c st \u201d at step 210 . if not , then this waveform is placed into the buffer marked \u201c non - st \u201d at step 211 . finally , after ten waveforms have been analyzed , a count of st and non - st marked waveforms is made to see if the count of non - st waveforms is greater than some threshold value x ( at step 212 ), where x can be , for example , 7 . if more than seven waveforms are non - st , then therapy is delivered at step 214 . the buffers at steps 210 and 211 in fig2 are buffers that hold the last ten decision values . these buffers can be pictured as sliding windows revealing the most recent ten values to permit the decision at 212 to be continuously updated . the buffers thus function as a nonlinear filter preventing irregular values from delivering therapy improperly . in developing a working prototype of this system , we used a development data set consisting of 20 episodes of st taken from 4 patients and 18 episodes of af taken from 7 patients . patient number episodes of st episodes of af 1 2 6 2 12 0 13 0 1 4 0 2 5 5 1 6 0 3 7 0 3 8 0 2 9 1 0 total 20 18 we used this data to optimize our choice of \u03c1 and \u03b2 and also the particular coefficients that we chose to examine . next , we tested our prototype using a set of 16 episodes of st obtained from 4 patients together with 17 episodes of af obtained from 7 patients . patient number episodes of st episodes of af 1 8 0 2 3 1 3 3 0 4 2 0 5 0 4 6 0 4 7 0 3 8 0 3 9 0 1 10 0 1 total 16 17 x out threshold coefficients of 10 ( \u03b2 ) sensitivity specificity 1 , 5 , 9 , & amp ; 24 3 0 . 808 76 % 94 % 1 , 5 , 9 , & amp ; 24 5 0 . 975 88 % 94 % 1 , 5 , 9 , & amp ; 24 6 0 . 976 82 % 94 % 1 , 5 , 9 , & amp ; 24 7 0 . 983 88 % 94 % 1 , 5 , & amp ; 24 3 0 . 815 82 % 94 % 1 , 5 , & amp ; 24 4 0 . 943 88 % 94 % 1 , 5 , & amp ; 24 5 0 . 956 82 % 94 % 1 , 5 , & amp ; 24 7 0 . 991 94 % 94 % 5 , 9 , & amp ; 24 8 0 . 9993 82 % 94 % 5 , 9 , & amp ; 24 9 0 . 9997 76 % 81 % the quality of the selected subset of transformed values for characterizing changes in morphology may be demonstrated graphically , as indicated in fig5 by performing a reverse haar transform using only the three or four or so selected transformed values . as can be seen ( at 510 and 512 in fig5 ), the inverse transformations produce distorted representations of the original atrial waveforms which illustrate how the choice of transform values can emphasize the changes . this is another useful way to assist one in selecting which transformed values are most useful . one feature of the invention is its ability to use a template derived from an average normal population , rather than deriving a patient customized average normal template for each individual patient . prior systems have required each new patient to be monitored while in a normal state , and the captured data was then averaged to form a patient specific normal template . contrary to this , the present invention achieved the results shown above using the same template for all patients . accordingly , the invention may be used with a new patient without the necessity of such preliminary testing of each patient and without a new customized template necessarily having to be created for each patient . one reason why the present invention can function using a template that represents average values for a normal population is because the specific coefficients selected in the transform domain , in addition to having been selected to emphasize the difference between normal and abnormal values , are also preferably chosen to minimize the differences between different patients . in some cases , values that were good at distinguishing between normal and abnormal rhythms for one patient were not as good at doing so for some other patient . these values are preferably not selected for use in implementing the present invention . for example , the transform coefficients may be selected such that all ( or most of ) the normal values of a particular coefficient gathered from many patients had the same sign ( positive or negative ) and similar amplitudes ( or if they varied in sign , they were near to zero ). in addition , the abnormal coefficients are significantly different in value from the normal coefficients for each particular patient . [ 0068 ] fig6 at 600 , illustrates one way in which this may be done . first , at step 602 , one gathers a large number of st and non - st data samples . next , all of this data is transformed , generating coefficients for every waveform set of data ( step 604 ). the variance of each coefficient is then computed first for the st samples ( step 606 ) and then for the non - st samples . then , for each coefficient , the ratio of the variance of the non - st samples to that of the st samples is computed ( step 608 ). finally , those coefficients whose variance ratio was large may be selected a coefficients for use in creating a population template and in testing patients ( step 612 ). in addition , at step 614 , the number of coefficients retained may be further reduced to reduce the number of computations that need to be performed , as explained above . in the version of the invention that was used to generate the data shown above , the population template was computed as follows , using the 20 st episodes in the development data set ( also used in the first of the two tables presented above ). this is illustrated at 700 in fig7 . first , the coefficients were selected as was explained above and as shown in fig6 . next , for each patient episode of st ( step 702 ), the recorded heartbeats were broken up into groups of ten consecutive heartbeats ( step 704 ). for each group ( step 706 ), the selected coefficients of the haar transform were computed ( step 708 ) for each of the ten heartbeats and the median value was then selected ( step 710 ) for each coefficient to form a proto - template . if these median values correlated well with the coefficient values for each of the ten heartbeats , the proto - template was retained ( step 714 ). otherwise , it was discarded . in this manner , a whole bunch of proto - templates were obtained from each patient episode . next , median values of the coefficients from all of the proto - templates ( step 716 ) were computed ( step 718 ). these values were then used as the population template for distinguishing st from non - st events ( step 720 ). while the preferred embodiment of the invention has been described above , it is to be understood that numerous modifications and changes will occur to those who are skilled in the art to which the invention pertains . accordingly , the following claims annexed to and forming a part of this specification are intended to define the true spirit and scope of the invention \u2014 that is , what is new and what is desired to be secured by letters patent of the united states ."}
|
Is the category the most suitable category for the given patent?
| 0.25 |
1cd5305994f1692ec7ec09368a2a725ff3c3dd15c21bf56c12e04e558cdb5fbc
| 0.046631 | 0.035156 | 0.013245 | 0.047363 | 0.150391 | 0.114258 |
null |
{"category": "Human Necessities", "patent": "the present invention is directed towards improving the quality and functionality of both implantable cardioverter defibrillators ( icds ) and implantable atrial defibrillators ( iads ). all such devices need to have a method and mechanism for accurately distinguishing sinus tachycardia ( st ) form non - st , such as atrial fibrillation and atrial flutter . this method and mechanism should be conservative , biased towards not initiating therapy unless non - st is clearly indicated . it should also require as few machine computational cycles as possible to reduce icd and iad battery drain . the principle underlying the present invention is that st and non - st can be distinguished by studying and measuring the atrial electrogram morphology to detect aberrant conduction in the atria . thus , by comparing normal st waveforms to a series of newly - occurring waveforms , st is indicated by abnormal atrial waveform morphology which signals aberrant conduction within the atria improper electrical patterns of depolarization . this will show up in the signal potentials captured by a bipolar atrial lead . with reference to fig1 the present invention can be implemented in an icd 100 ( which could be an iad ) that is implanted in the chest of a patient who suffers from unpleasant and possibly life - threatening irregularities in heart operation due to improper electrical stimulation of the heart muscle cells . during normal heart action , the heart &# 39 ; s electrical impulses originate in the sino - atrial node as an action potential that is transmitted smoothly to all portions of the atria , causing contraction of the atrial chambers . the electrical impulse continues in its path to a cluster of conduction fibrils known as the atrioventricular node . after a delay of about one - tenth of a second , an action potential flows over the ventricles and causes them to contract in synchronism with and following shortly after the atrial contraction . in this manner , the heart pumps blood to the lungs and from the lungs to the body . a bipolar lead 102 is implanted within the atria 104 of a heart 103 to measure the electrical potential between nearby cells of the atria . as the action potential actively passes from cell to cell past the two electrodes of the lead 102 , an oscillating signal potential is developed across the two electrodes and is conveyed by the bipolar lead 102 to the icd 100 where the signal is sampled about 1000 times per second and is digitized , with the sequential samples stored within a ram memory within the icd 100 for further analysis . another bipolar lead 105 may extend to the ventricle ( not shown ) of the heart 103 to measure the action potentials generated within the ventricles . these same leads , or other leads , may be used by the icd for administering various types of therapy , such as pacing or defibrillation shock therapy . the data samples collected from within the atria are now analyzed . first , with reference to data gathered from the ventricles , if some event in the ventricle , such as a mis - timed depolarization event that partially overlaps the atria &# 39 ; s p wave , could leak across to the atria and distort the measurement of the potentials for a given atrial depolarization , then the data for that particular heartbeat is discarded and is not used in further analyses . such distortion can also be found simply because a given heartbeat set of data gathered from the atria is itself badly distorted , and this approach must be taken in the case of an iad having no ventricular lead . the remaining data is retained for further processing . in preparation for further processing , the data for individual heartbeats is identified and gathered . the position within the gathered data of the negative spike to the atrial p depolarization waveform is determined and is selected as the center of the waveform for each beat . then , since data close to this p - wave is to be analyzed , and since data further away may be distorted to some degree by ventricular activity or by variations in the p - to - p interval , 32 sequential data samples are selected for each heartbeat such that the negative spike of the p - wave becomes signal potential sample number 16 of the 32 sequential samples selected for further analyses . these samples for a normal heartbeat appear at 302 in fig3 and at 502 in fig5 where the amplitude of the atrial signal potential is plotted against time over the 32 sampled values , with the negative p - wave spike positioned as signal sample number 16 . likewise , the samples for an abnormal heartbeat appear at 504 in fig5 . straightforward correlation ( cwa ) of the 32 samples against a template containing a typical or normal pattern could be performed at this juncture , as taught on page 563 ( equation ( 1 )) of the article by thorne , cited in the introductory portion of this specification . but the performance of such a cross correlation at repeated regular intervals would generate numerous computations and would drain the icd &# 39 ; s battery more rapidly than is desirable . in addition , provision would have to be made for a full template waveform that can be stored within the icd to be used as a comparison reference . in addition , such templates generally need to be patient specific , and they are more susceptible to normal changes in shape . accordingly , to reduce the mathematical complexity of the cross correlation computations , it is desirable to reduce the number of data points from 32 down to some much lower number through the use of some form of transformation into a different set of values . for example , the data could be transformed using the karhunen - loeve transformation described in the morris article cited at the beginning of this specification . but that transformation is fairly computationally intensive . computationally , a fast fourier transformation would be more efficient , but any such transformation into the frequency domain , where variable frequency sine and cosine wave amplitudes result from the transformation , does not match itself well to the morphology of atrial heart waveforms , which tend to be formed from large , ringing , spike - like representations of p - wave depolarization events travelling as moving action potential fronts past the pair of electrodes attached to the bipolar lead , and not as steady harmonics extending across time . accordingly , fourier - class transforms do not adequately reduce the number of significant data values that must be considered . and in addition , when the \u201c window \u201d size for a fourier analysis is selected , if it is wide , then the frequency values are not time - specific , but represent average signal harmonic content over the entire window time duration . and if the window is narrow , then the harmonics are too spread out in frequency , and the frequency - domain data generated by the transform does not resolve things finely enough in the frequency domain . for all of the above reasons , the present invention analyzes the waveforms using a discrete wavelet transform . this has the advantage that while low frequency wavelets are broad in time , high frequency wavelets have a much narrower timeslot focus . thus , for example , with 32 data samples taken sequentially over time , a wavelet transformation generates a d . c . wavelet that spans the entire set of 32 points ; a first reversing wavelet that also spans the entire set of points ; two second harmonic wavelets that each focus upon only one - half of the 32 data points ; four third harmonic wavelets that each focus upon only one - fourth of the 32 data points ; eight fourth harmonic wavelets that each focus upon four of the 32 data points ; and 16 fifth harmonic wavelets that each focuses upon only two of the 32 data points . accordingly , the higher - frequency wavelet amplitudes are quite time specific , unlike fourier sinusoidal amplitudes , while the lower - frequency and d . c . wavelet amplitudes give one broad information about the whole set of 32 points . and like the fourier transform , the discrete wavelet transform preserves all of the information of the original atrial signal , such that the transformation is fully reversible . 32 discrete wavelet transform values may be reverse transformed back into 32 values indicating the atrial signal potentials at 32 sequential points in time . in essence , the haar function is performing multiple digital filtering operations , at variable positions and utilizing varying - width windowing functions , to develop component values that represent varying types of heartbeat activity , at varying frequencies , spread over varying widths , and located at varying positions . these component values can be studied to see which subset of these component values are good at distinguishing st from non - st , or ( more generally ) which subset of these component values are good at distinguishing one type of heartbeat waveform from another . further study of such a selected subset can further determine which of the subset of components are relatively invariant from one individual &# 39 ; s heart to another . one preferably selects component values that are suitable from both of these two perspectives . the particular wavelet transformation to choose can be tailored to the nature of the data , with the wavelets chosen to resemble somewhat the values to be found within the data so that some transformed values are of much larger amplitude than others , and such that the low amplitude transformed values may be disregarded . of particular importance with an atrial p waveform of the type being analyzed here are wavelets having frequency values roughly comparable to and timed to coincide with the ringing of the atrial depolarization . this tailoring can minimize the number of transformed data values that are significant and that are candidates for full participation in the correlation analyses to determine if there has been a substantial change in waveform morphology . another factor in selecting a particular wavelet transformation is reducing the number of computations that must be performed to carry out the transformation . yet another factor is selecting a wavelet transformation where some of the transformed component values vary from normal to abnormal heart rhythms in much the same over a population of individuals so that a generic template of these component values may be derived that can be used with many individuals , rather than with just one individual . these factors suggest that the haar transformation would be a suitable candidate for use in analysis of atrial waveforms and possibly other waveforms as well . with reference to fig3 in the case of 32 voltage values sampled over time , the 32 applicable haar discrete transform wavelets appear as shown in this figure . other discrete wavelet transforms based upon wavelets having triangular or other shapes may also prove usable . the haar transform wavelets , in particular , have the shape of a square waveform , as will be described , and this can simplify the computations required . in fig3 time increases from left to right . a scale 304 indicates the 32 points in time at which the atrial waveform is sampled , and an analog atrial waveform ( before digitization ) is shown at 302 . the p wave negative notch 303 is shown centered at the 16 th timeslot so that the signal potential of the atria waveform sampled at this point in time becomes the 16 th data value in the set of 32 data values that are to be subjected to the haar transformation . a haar wavelet is simply one cycle of a square waveform that starts at zero , then swings positive ( to \u201c+ 1 \u201d), then swings negative ( to \u201c\u2212 1 \u201d), and then swings back to zero . the wavelet w 2 , shown at 308 in fig3 for example , is at \u201c+ 1 \u201d at points in time 1 to 16 and is at \u201c\u2212 1 \u201d at points in time 17 to 32 . the shorter waveform w 6 , shown at 316 in fig3 is at \u201c+ 1 \u201d at points in time 9 to 12 and is at \u201c\u2212 1 \u201d at points in time 13 to 16 , and is at \u201c 0 \u201d at all other points in time . the waveform w 1 , shown at 306 in fig3 is so slow to fluctuate in time that its \u201c\u2212 1 \u201d portion is off of the chart ( in fig3 ) to the right , and is ignored ; and accordingly , it has the value \u201c+ 1 \u201d for all 32 of the points in time shown in fig3 . it thus represents the average , or d . c ., component of the atrial signal potential . the lowest frequency wavelets w 1 and w 2 encompass the entire set of 32 points in time . the higher frequency wavelets w 3 and w 4 each encompasses only half of the points in time , and the two wavelets taken together encompass all the points in time . the four wavelets w 5 , w 6 , w 7 , and w 8 each encompasses only eight points in time , and the four wavelets taken together encompass all points in time . the eight wavelets w 9 . . . w 16 each encompasses only four points in time , and together all eight wavelets encompass all points in time . and finally , the wavelets w 17 . . . w 32 each encompass only two points in time , but the sixteen wavelets together encompass all points in time . thus , each wavelet has a characteristic time span and position as well as a characteristic frequency , with higher - frequency wavelets having a narrower and more specific time span and position than lower - frequency wavelets . this is advantageous with an impulse - type , ringing signal such as the atrial depolarization waveform considered here , since many higher - frequency transform values that correspond to haar wavelets positioned in time away from the p waveform or that do not correspond to its frequency of ringing may be low in amplitude such that they may safely be ignored during the analysis steps , thereby reducing the data that must be processed during the correlation steps . ( the above , while presented in the context of the haar wavelet , is also applicable to other wavelet shapes that may used to perform a discrete wavelet transform ( dwt )). each of the 32 haar transformed values is computed as follows : multiply each of the 32 sampled and digitized voltage values representing the analog atrial waveform 302 by the correspondingly - positioned -( in - time ) amplitude values of one of the 32 haar wavelets ( shown in fig3 ); then sum the resulting products ; and then multiply the resulting sums by a discrete wavelet transform scaling factor 2 \u2212 j / 2 , where j equals 1 for the wavelets w 17 to w 32 , 2 for w 9 to w 16 , 3 for w 5 to w 8 , 4 for w 3 to w 4 , and 5 for w 1 and w 2 . for example , and referring to fig3 : the first wavelet w 1 is always + 1 , so the atrial signal potential values are simply summed and then multiplied by 2 \u2212 5 / 2 ; the second wavelet is + 1 for time points 1 to 16 and \u2212 1 for time points 17 to 32 , so the first 16 values are summed , the second 16 values are summed , the difference between the first and second sums is computed , and the result is multiplied by 2 \u2212 5 / 2 ; and so on until all the transformed values have been computed , one for each haar wavelet . the 32 time - sequential atrial signal potential values are thus transformed into 32 haar wavelet amplitude values which may be called \u201c transformed values \u201d and which may be assigned the numbers 1 to 32 corresponding to the subscript numbers of the haar wavelets to which they each correspond and whose amplitudes they represent . the above description of how to compute the haar wavelet transformed values is accurate , but it is not the most efficient way to proceed in the case of haar wavelets . like the fourier transform , which has a corresponding fast fourier transform that saves intermediate results and thereby avoids re - computing them and thus reduces substantially the number of computations , the haar transformation also may be carried out in a manner that saves intermediate sums and re - uses them to reduce substantially the number of computations . this is described below at the point where fig4 is described , since fig4 illustrates graphically how this can be done . the illustrative program listing presented below is also an implementation of the fast haar wavelet transformation algorithm . the haar transformation can readily be carried out by any digital computer . as an example of how the haar transformation can be carried out on 32 values , the following program , written in the language c , is illustrative of many possible programs that may be written . in the exemplary program that follows , a 32 - element array yy contains 32 data values representing the 32 sampled atrial signal potential values representing the fluctuations over time of the signal supplied by the bipolar lead 102 . the analog atrial signal is sampled , digitized , broken into separate beat data sets , pre - processed ( to remove waveforms distorted by ventricular activity ), centered ( with the negative depolarization spike at data point 16 ), and fed into the subroutine presented below . this subroutine is compiled ( or assembled , if rewritten in assembly language ) and installed in the rom of the icd &# 39 ; s embedded microprocessor . this subroutine returns , contained within the same array yy , the 32 transformed haar wavelet amplitude values described above . ( the constant value \u201c recipsqrttwo \u201d is the reciprocal of the square root of two ). an illustrative version of the subroutine for computing all 32 of the haar transformed values in an efficient manner is presented here : void haar ( double yy []) { int i , j , l ; double zz [ 32 ]; for ( i = 5 ; i & gt ; 0 ; i \u2212\u2212) { l = 1 ; for ( j = 1 ; j & lt ;= i ; ++ j ) l = 2 * l ; for ( j = 0 ; j & lt ; l ; ++ j ) zz [ j ] = yy [ j ]; for ( j = 0 ; j & lt ; l \u2212 1 ; j = j + 2 ) ( yy [ j / 2 ] = recipsqrttwo * ( zz [ j ] + zz [ j + 1 ]); yy [( j + l )/ 2 ] = rcipsqrttwo * ( zz [ j ] \u2212 zz [ j + 1 ]); } } return ; } the above program computes the 32 haar transformed values 1 to 32 from the 32 atrial signal potential time - sequenced input values . it does so with only 31 additions , 31 subtractions , and 62 multiplications . it is carefully designed to compute each value in a systematic way , making multiple use of intermediate results . first , the program computes sixteen sums of and sixteen differences between eight adjacent pairs of the 16 time sequenced atrial signal potential values , and the sixteen difference values become the haar transformed values 17 to 32 , reflecting the strength of the highest frequency haar wavelets positioned at 16 different positions in time , corresponding to the wavelets w 17 to w 32 shown in fig3 . all the sum and difference values are scaled by multiplication by 2 \u2212 1 / 2 . next , taking the 16 sums of atrial signal potential values as an intermediate result , the above algorithm generates eight sums of and eight differences between adjacent pairs of these sixteen intermediate values that resulted from the first sixteen additions , and the eight newly - computed difference values become the haar transformed values 9 to 16 , reflecting the strength of the second to the highest frequency values at eight different points in time , corresponding to wavelets w 9 to w 16 in fig3 . all of these eight sum and eight difference values are again scaled by 2 \u2212 1 / 2 so that the wavelets w 9 to w 16 are scaled by \u00bd ( 2 \u2212 2 / 2 or 2 \u2212 1 / 2 times 2 \u2212 1 / 2 ). next , taking these eight sums of sums of adjacent atrial signal potential values as an intermediate result , the above algorithm generates four sum and four difference values , again scaling by 2 \u2212 1 / 2 , and the four difference values become haar transformed values 5 to 8 , reflecting the strength of the middle frequency wavelets at four different points in time , and corresponding to wavelets w 5 to w 8 in fig3 . next , taking these four sums of sums of sums of adjacent atrial signal potential values , the above algorithm generates two sum and two difference values , again scaling by 2 \u2212 1 / 2 , and the two difference values become haar transformed values 3 and 4 , reflecting the strength of the second to the lowest frequency wavelets only two of which encompass all the time domain data , corresponding to the wavelets w 3 and w 4 shown in fig3 . and finally , the algorithm generates the sum of and the difference between the final remaining two sums of sums of sums of sums of atrial signal potential values , again scaling by 2 \u2212 1 / 2 . the difference value is then the haar transformed value 2 , representing the strength of the lowest - frequency wavelet , the one corresponding to the wavelet w 2 in fig3 the wavelet that extends the full length of the time scale . the sum value is then the first , or d . c ., transformed value , representing the average signal potential level over the 32 sampled points in time , which corresponds to the wavelet w 1 in fig3 . another way of viewing this computation is illustrated in fig4 a and 4b . the 32 atrial signal potential values are shown at the top of fig4 a and are identified as c 0 5 through c 31 5 . the superscript \u201c 5 \u201d indicates that these values are processed when the index value \u201c n \u201d in the above computer program is equal to \u201c 5 \u201d\u2014 that is , during the first pass through the data generating intermediary sums and differences . during subsequent passes , the value of n is decremented to 4 , 3 , 2 , and finally to 1 . during each pass through the data , the computer generates sums c n - 1 and differences d n - 1 , as shown in fig4 b , between adjacent pairs of the values c 0 n and c 1 n ; c 2 n and c 3 n ; and so on , so that the number of newly - generated c n - 1 and d n - 1 terms is reduced by half with each computer pass through the intermediary results . at the end of all these computations , the value c 0 0 is the first , or d . c ., haar transformed value ; and the values d 0 0 ; d 0 1 and d 1 1 ; d 0 2 , d 1 2 , d 2 2 , and d 3 2 ; d 0 3 , d 1 3 , . . . and d 7 3 ; and d 0 4 , d 1 4 , and d 15 4 are , respectively , the remaining haar transform values 2 through 32 . fig4 thus illustrates quite succinctly how all the transform computations are carried out , and how the intermediary \u201c c \u201d sum values , such as c 0 4 , are used to compute multiple haar values , such as the values d 0 3 , d 0 2 , d 0 1 , d 0 0 , and c 0 0 all of which are computed from the intermediary value c 0 4 . saving and reusing these intermediary \u201c c \u201d values saves much computational time . fig4 b indicates the precise addition and subtraction operations that are carried out at each level to compute the values in the next lower level , proceeding down through the chart presented in fig4 a . but in any given application to heart waveform analysis , all of these computations may not be needed , and accordingly the number of computations may be reduced much further . since the present invention teaches that only a small number of these haar transformed values need actually be considered , a far less computationally intensive transform can be developed which only generates the intermediate and final transformed values that are actually needed to generate the specific haar transformed values which have proved to be significant in discriminating between sinus tachycardia ( or st ) and non - st conditions . all others need not be computed , and the above program may be reduced to a special algorithm that omits as many sums , differences , and multiplications as possible . for example , if only the transformed values 1 , 5 , 9 , and 24 are significant , then 31 additions are required to compute the first transformed value ( the d . c . value \u2014 the simple sum of all the time domain signal values ); six additions and one subtraction are required to compute the fifth transformed value ( in fig3 the difference between the sums of time domain values under each half of the wavelet w 5 at 314 in fig3 ); two additions and one subtraction are required to compute the ninth transformed value ( in fig3 the difference between the sums of the time domain values under each half of the wavelet w9 at 322 in fig3 ); and only one subtraction is required to compute the 24 th transformed value ( in fig3 the difference between the two time domain values under the respective halves of the wavelet w 24 ( not shown ) which is the same size as , but differently positioned in time than , the wavelet w 18 at 330 ( fig3 ). accordingly , if only the transformed values 1 , 5 , 9 , and 24 actually evaluated , then only 42 additions and subtractions are required . but even this number can be reduced further . if the computational algorithm set forth in the above illustrative computer program is followed as a guide , then the intermediary sums used in computing the haar first , or d . c ., transformed value can be re - used to compute the sums for the haar transformed values 5 and 9 , assuming these intermediary results are saved in the manner described in the above program example . then 31 additions are still required to compute the first transformed value , but only one subtraction is required to compute each of the transformed values 5 , 9 , and 24 , giving a total of additions and subtraction of only 34 operations . and if only transformed values 1 , 5 , and 9 are computed , the number of additions and subtractions is reduced to 33 . and if the first transformed value is omitted and if transformed values 5 , 9 , and 24 are selected , then the 31 additions needed to compute the first transformed value are not required . then the number of computations for transformed value 5 is 8 additions and one subtraction ; the number of computations for transformed value 9 is 2 additions and one subtraction ; and the number of computations for transformed value 24 is still just one subtraction . so the total number of additions and subtractions is just 13 . but even this number can be reduced to 11 when it is realized that the two additions done for the transformed value 9 are also done ( in the above computational algorithm ) when computing the transformed value 5 . so the number of additions and subtractions can be reduced to 11 . also , when computing such a small number of transformed values , the number of multiplications can be reduced as well by postponing the multiplications until a transformed value is actually computed . with reference to fig4 a , instead of dividing every sum and difference by the square root of two ( as shown in fig4 b ), several vertical sums of \u201c c \u201d values in different rows of the fig4 a table can be formed , and then a \u201c d \u201d value can be computed by subtraction , and then the resulting unscaled transformed value can be scaled with a single multiplication by 2 \u2212 j / 2 where j is the number of rows in the table of fig4 a traversed by the one or more sum and the one difference computations . thus , the number of scaling multiplications can sometimes be equal to the number of transformed values that are computed , typically 3 or 4 . accordingly , by using wavelet analysis and transformation , instead of fourier or discrete cosine or other sinusoid analysis and transformation , a highly useful and highly frequency - specific result can be achieved with a very small number of computations , thereby conserving power and battery life , and yet achieving a high degree of precision in recognizing changes in morphology . in our tests , we have selected certain transformed values as being much more significant than other values in distinguishing between st and non - st . we focused upon those transform values whose \u201c+ 1 \u201d and \u201c\u2212 1 \u201d scope included the middle 8 time points most significant to analysis of the peak of the atrial depolarization . working across test data samples obtained from patients who had dual chamber icds , we computed the variance of each transformed value for st and non - st events using the standard statistical formula for computing variance . we then calculated the ratio of the variance of non - st events to the variance of st events , and selected those terms with the highest variance ratios for further consideration . in this manner , we reduced the number of transformed values that were included in the cwa or correlation process while always maintaining or improving the performance achieved . ultimately we settled upon the three or four transformed values having the highest ratios . these were the transform values 1 , 5 , 9 , and 24 . we tested 1 , 5 , and 24 together ; 5 , 9 , and 24 together ; and 1 , 5 , 9 , and 24 together . these selected haar transform values , when used in these combinations , required minimal computations and thereby produced a savings in battery power , and that also gave better performance at distinguishing st from non - st than did all of the transformed values used together . so an increase in accuracy was achieved as well as a decrease in computational complexity by this approach to atrial waveform analysis . [ 0054 ] fig5 for example , illustrates actual plots of digital information illustrating the effect of using only the three coefficients 1 , 5 , and 24 . at 502 , a normal waveform is shown , with amplitude plotted against sample numbers from 1 to 32 . at 504 , an abnormal waveform is shown , again with amplitude plotted against sample numbers . after the haar transformation , at 506 and 508 , the wavelet coefficient amplitude values are indicated for the coefficients 1 to 32 . at 506 , the coefficients generated by the normal waveform 502 are shown , and at 508 , the coefficients generated by the abnormal waveform 504 are shown . one may directly compare these component values and verify that the coefficients 1 , 5 , and 24 at 506 and at 508 vary significantly in amplitude . comparison of this same information among different patients ( not shown in this figure ) also indicated that this variation is relatively constant from one patient to another . at 510 and 512 , using only the coefficients 1 , 5 , and 24 , the heartbeat waveforms are reconstructed by a reverse transformation , the normal reconstructed waveform shown at 510 and the abnormal reconstructed waveform shown at 512 . the marked differences between these two reconstructed waveforms highlights the way in which these three coefficients can signal an abnormal condition with less computation . in a practical system , a small number of transform values are selected , in the manner just described . the algorithm for computing these values is then refined , as explained above , to reduce the number of additions , subtractions , and multiplications to the minimum possible while preserving the accuracy of the computations . we first compute a value \u03c1 , which is the correlation between these values with the corresponding values in a template that represents the values for an average normal population . we then compare this value \u03c1 to a threshold correlation value \u03b2 that is chosen to give optimal results on experimental data . ( see the examples in the tables presented below .) for each waveform analyzed , a test is made to determine whether \u03c1 is greater than \u03b2 . if so , then this waveform is placed into the buffer marked \u201c st \u201d at step 210 . if not , then this waveform is placed into the buffer marked \u201c non - st \u201d at step 211 . finally , after ten waveforms have been analyzed , a count of st and non - st marked waveforms is made to see if the count of non - st waveforms is greater than some threshold value x ( at step 212 ), where x can be , for example , 7 . if more than seven waveforms are non - st , then therapy is delivered at step 214 . the buffers at steps 210 and 211 in fig2 are buffers that hold the last ten decision values . these buffers can be pictured as sliding windows revealing the most recent ten values to permit the decision at 212 to be continuously updated . the buffers thus function as a nonlinear filter preventing irregular values from delivering therapy improperly . in developing a working prototype of this system , we used a development data set consisting of 20 episodes of st taken from 4 patients and 18 episodes of af taken from 7 patients . patient number episodes of st episodes of af 1 2 6 2 12 0 13 0 1 4 0 2 5 5 1 6 0 3 7 0 3 8 0 2 9 1 0 total 20 18 we used this data to optimize our choice of \u03c1 and \u03b2 and also the particular coefficients that we chose to examine . next , we tested our prototype using a set of 16 episodes of st obtained from 4 patients together with 17 episodes of af obtained from 7 patients . patient number episodes of st episodes of af 1 8 0 2 3 1 3 3 0 4 2 0 5 0 4 6 0 4 7 0 3 8 0 3 9 0 1 10 0 1 total 16 17 x out threshold coefficients of 10 ( \u03b2 ) sensitivity specificity 1 , 5 , 9 , & amp ; 24 3 0 . 808 76 % 94 % 1 , 5 , 9 , & amp ; 24 5 0 . 975 88 % 94 % 1 , 5 , 9 , & amp ; 24 6 0 . 976 82 % 94 % 1 , 5 , 9 , & amp ; 24 7 0 . 983 88 % 94 % 1 , 5 , & amp ; 24 3 0 . 815 82 % 94 % 1 , 5 , & amp ; 24 4 0 . 943 88 % 94 % 1 , 5 , & amp ; 24 5 0 . 956 82 % 94 % 1 , 5 , & amp ; 24 7 0 . 991 94 % 94 % 5 , 9 , & amp ; 24 8 0 . 9993 82 % 94 % 5 , 9 , & amp ; 24 9 0 . 9997 76 % 81 % the quality of the selected subset of transformed values for characterizing changes in morphology may be demonstrated graphically , as indicated in fig5 by performing a reverse haar transform using only the three or four or so selected transformed values . as can be seen ( at 510 and 512 in fig5 ), the inverse transformations produce distorted representations of the original atrial waveforms which illustrate how the choice of transform values can emphasize the changes . this is another useful way to assist one in selecting which transformed values are most useful . one feature of the invention is its ability to use a template derived from an average normal population , rather than deriving a patient customized average normal template for each individual patient . prior systems have required each new patient to be monitored while in a normal state , and the captured data was then averaged to form a patient specific normal template . contrary to this , the present invention achieved the results shown above using the same template for all patients . accordingly , the invention may be used with a new patient without the necessity of such preliminary testing of each patient and without a new customized template necessarily having to be created for each patient . one reason why the present invention can function using a template that represents average values for a normal population is because the specific coefficients selected in the transform domain , in addition to having been selected to emphasize the difference between normal and abnormal values , are also preferably chosen to minimize the differences between different patients . in some cases , values that were good at distinguishing between normal and abnormal rhythms for one patient were not as good at doing so for some other patient . these values are preferably not selected for use in implementing the present invention . for example , the transform coefficients may be selected such that all ( or most of ) the normal values of a particular coefficient gathered from many patients had the same sign ( positive or negative ) and similar amplitudes ( or if they varied in sign , they were near to zero ). in addition , the abnormal coefficients are significantly different in value from the normal coefficients for each particular patient . [ 0068 ] fig6 at 600 , illustrates one way in which this may be done . first , at step 602 , one gathers a large number of st and non - st data samples . next , all of this data is transformed , generating coefficients for every waveform set of data ( step 604 ). the variance of each coefficient is then computed first for the st samples ( step 606 ) and then for the non - st samples . then , for each coefficient , the ratio of the variance of the non - st samples to that of the st samples is computed ( step 608 ). finally , those coefficients whose variance ratio was large may be selected a coefficients for use in creating a population template and in testing patients ( step 612 ). in addition , at step 614 , the number of coefficients retained may be further reduced to reduce the number of computations that need to be performed , as explained above . in the version of the invention that was used to generate the data shown above , the population template was computed as follows , using the 20 st episodes in the development data set ( also used in the first of the two tables presented above ). this is illustrated at 700 in fig7 . first , the coefficients were selected as was explained above and as shown in fig6 . next , for each patient episode of st ( step 702 ), the recorded heartbeats were broken up into groups of ten consecutive heartbeats ( step 704 ). for each group ( step 706 ), the selected coefficients of the haar transform were computed ( step 708 ) for each of the ten heartbeats and the median value was then selected ( step 710 ) for each coefficient to form a proto - template . if these median values correlated well with the coefficient values for each of the ten heartbeats , the proto - template was retained ( step 714 ). otherwise , it was discarded . in this manner , a whole bunch of proto - templates were obtained from each patient episode . next , median values of the coefficients from all of the proto - templates ( step 716 ) were computed ( step 718 ). these values were then used as the population template for distinguishing st from non - st events ( step 720 ). while the preferred embodiment of the invention has been described above , it is to be understood that numerous modifications and changes will occur to those who are skilled in the art to which the invention pertains . accordingly , the following claims annexed to and forming a part of this specification are intended to define the true spirit and scope of the invention \u2014 that is , what is new and what is desired to be secured by letters patent of the united states ."}
|
{"category": "Mechanical Engineering; Lightning; Heating; Weapons; Blasting", "patent": "the present invention is directed towards improving the quality and functionality of both implantable cardioverter defibrillators ( icds ) and implantable atrial defibrillators ( iads ). all such devices need to have a method and mechanism for accurately distinguishing sinus tachycardia ( st ) form non - st , such as atrial fibrillation and atrial flutter . this method and mechanism should be conservative , biased towards not initiating therapy unless non - st is clearly indicated . it should also require as few machine computational cycles as possible to reduce icd and iad battery drain . the principle underlying the present invention is that st and non - st can be distinguished by studying and measuring the atrial electrogram morphology to detect aberrant conduction in the atria . thus , by comparing normal st waveforms to a series of newly - occurring waveforms , st is indicated by abnormal atrial waveform morphology which signals aberrant conduction within the atria improper electrical patterns of depolarization . this will show up in the signal potentials captured by a bipolar atrial lead . with reference to fig1 the present invention can be implemented in an icd 100 ( which could be an iad ) that is implanted in the chest of a patient who suffers from unpleasant and possibly life - threatening irregularities in heart operation due to improper electrical stimulation of the heart muscle cells . during normal heart action , the heart &# 39 ; s electrical impulses originate in the sino - atrial node as an action potential that is transmitted smoothly to all portions of the atria , causing contraction of the atrial chambers . the electrical impulse continues in its path to a cluster of conduction fibrils known as the atrioventricular node . after a delay of about one - tenth of a second , an action potential flows over the ventricles and causes them to contract in synchronism with and following shortly after the atrial contraction . in this manner , the heart pumps blood to the lungs and from the lungs to the body . a bipolar lead 102 is implanted within the atria 104 of a heart 103 to measure the electrical potential between nearby cells of the atria . as the action potential actively passes from cell to cell past the two electrodes of the lead 102 , an oscillating signal potential is developed across the two electrodes and is conveyed by the bipolar lead 102 to the icd 100 where the signal is sampled about 1000 times per second and is digitized , with the sequential samples stored within a ram memory within the icd 100 for further analysis . another bipolar lead 105 may extend to the ventricle ( not shown ) of the heart 103 to measure the action potentials generated within the ventricles . these same leads , or other leads , may be used by the icd for administering various types of therapy , such as pacing or defibrillation shock therapy . the data samples collected from within the atria are now analyzed . first , with reference to data gathered from the ventricles , if some event in the ventricle , such as a mis - timed depolarization event that partially overlaps the atria &# 39 ; s p wave , could leak across to the atria and distort the measurement of the potentials for a given atrial depolarization , then the data for that particular heartbeat is discarded and is not used in further analyses . such distortion can also be found simply because a given heartbeat set of data gathered from the atria is itself badly distorted , and this approach must be taken in the case of an iad having no ventricular lead . the remaining data is retained for further processing . in preparation for further processing , the data for individual heartbeats is identified and gathered . the position within the gathered data of the negative spike to the atrial p depolarization waveform is determined and is selected as the center of the waveform for each beat . then , since data close to this p - wave is to be analyzed , and since data further away may be distorted to some degree by ventricular activity or by variations in the p - to - p interval , 32 sequential data samples are selected for each heartbeat such that the negative spike of the p - wave becomes signal potential sample number 16 of the 32 sequential samples selected for further analyses . these samples for a normal heartbeat appear at 302 in fig3 and at 502 in fig5 where the amplitude of the atrial signal potential is plotted against time over the 32 sampled values , with the negative p - wave spike positioned as signal sample number 16 . likewise , the samples for an abnormal heartbeat appear at 504 in fig5 . straightforward correlation ( cwa ) of the 32 samples against a template containing a typical or normal pattern could be performed at this juncture , as taught on page 563 ( equation ( 1 )) of the article by thorne , cited in the introductory portion of this specification . but the performance of such a cross correlation at repeated regular intervals would generate numerous computations and would drain the icd &# 39 ; s battery more rapidly than is desirable . in addition , provision would have to be made for a full template waveform that can be stored within the icd to be used as a comparison reference . in addition , such templates generally need to be patient specific , and they are more susceptible to normal changes in shape . accordingly , to reduce the mathematical complexity of the cross correlation computations , it is desirable to reduce the number of data points from 32 down to some much lower number through the use of some form of transformation into a different set of values . for example , the data could be transformed using the karhunen - loeve transformation described in the morris article cited at the beginning of this specification . but that transformation is fairly computationally intensive . computationally , a fast fourier transformation would be more efficient , but any such transformation into the frequency domain , where variable frequency sine and cosine wave amplitudes result from the transformation , does not match itself well to the morphology of atrial heart waveforms , which tend to be formed from large , ringing , spike - like representations of p - wave depolarization events travelling as moving action potential fronts past the pair of electrodes attached to the bipolar lead , and not as steady harmonics extending across time . accordingly , fourier - class transforms do not adequately reduce the number of significant data values that must be considered . and in addition , when the \u201c window \u201d size for a fourier analysis is selected , if it is wide , then the frequency values are not time - specific , but represent average signal harmonic content over the entire window time duration . and if the window is narrow , then the harmonics are too spread out in frequency , and the frequency - domain data generated by the transform does not resolve things finely enough in the frequency domain . for all of the above reasons , the present invention analyzes the waveforms using a discrete wavelet transform . this has the advantage that while low frequency wavelets are broad in time , high frequency wavelets have a much narrower timeslot focus . thus , for example , with 32 data samples taken sequentially over time , a wavelet transformation generates a d . c . wavelet that spans the entire set of 32 points ; a first reversing wavelet that also spans the entire set of points ; two second harmonic wavelets that each focus upon only one - half of the 32 data points ; four third harmonic wavelets that each focus upon only one - fourth of the 32 data points ; eight fourth harmonic wavelets that each focus upon four of the 32 data points ; and 16 fifth harmonic wavelets that each focuses upon only two of the 32 data points . accordingly , the higher - frequency wavelet amplitudes are quite time specific , unlike fourier sinusoidal amplitudes , while the lower - frequency and d . c . wavelet amplitudes give one broad information about the whole set of 32 points . and like the fourier transform , the discrete wavelet transform preserves all of the information of the original atrial signal , such that the transformation is fully reversible . 32 discrete wavelet transform values may be reverse transformed back into 32 values indicating the atrial signal potentials at 32 sequential points in time . in essence , the haar function is performing multiple digital filtering operations , at variable positions and utilizing varying - width windowing functions , to develop component values that represent varying types of heartbeat activity , at varying frequencies , spread over varying widths , and located at varying positions . these component values can be studied to see which subset of these component values are good at distinguishing st from non - st , or ( more generally ) which subset of these component values are good at distinguishing one type of heartbeat waveform from another . further study of such a selected subset can further determine which of the subset of components are relatively invariant from one individual &# 39 ; s heart to another . one preferably selects component values that are suitable from both of these two perspectives . the particular wavelet transformation to choose can be tailored to the nature of the data , with the wavelets chosen to resemble somewhat the values to be found within the data so that some transformed values are of much larger amplitude than others , and such that the low amplitude transformed values may be disregarded . of particular importance with an atrial p waveform of the type being analyzed here are wavelets having frequency values roughly comparable to and timed to coincide with the ringing of the atrial depolarization . this tailoring can minimize the number of transformed data values that are significant and that are candidates for full participation in the correlation analyses to determine if there has been a substantial change in waveform morphology . another factor in selecting a particular wavelet transformation is reducing the number of computations that must be performed to carry out the transformation . yet another factor is selecting a wavelet transformation where some of the transformed component values vary from normal to abnormal heart rhythms in much the same over a population of individuals so that a generic template of these component values may be derived that can be used with many individuals , rather than with just one individual . these factors suggest that the haar transformation would be a suitable candidate for use in analysis of atrial waveforms and possibly other waveforms as well . with reference to fig3 in the case of 32 voltage values sampled over time , the 32 applicable haar discrete transform wavelets appear as shown in this figure . other discrete wavelet transforms based upon wavelets having triangular or other shapes may also prove usable . the haar transform wavelets , in particular , have the shape of a square waveform , as will be described , and this can simplify the computations required . in fig3 time increases from left to right . a scale 304 indicates the 32 points in time at which the atrial waveform is sampled , and an analog atrial waveform ( before digitization ) is shown at 302 . the p wave negative notch 303 is shown centered at the 16 th timeslot so that the signal potential of the atria waveform sampled at this point in time becomes the 16 th data value in the set of 32 data values that are to be subjected to the haar transformation . a haar wavelet is simply one cycle of a square waveform that starts at zero , then swings positive ( to \u201c+ 1 \u201d), then swings negative ( to \u201c\u2212 1 \u201d), and then swings back to zero . the wavelet w 2 , shown at 308 in fig3 for example , is at \u201c+ 1 \u201d at points in time 1 to 16 and is at \u201c\u2212 1 \u201d at points in time 17 to 32 . the shorter waveform w 6 , shown at 316 in fig3 is at \u201c+ 1 \u201d at points in time 9 to 12 and is at \u201c\u2212 1 \u201d at points in time 13 to 16 , and is at \u201c 0 \u201d at all other points in time . the waveform w 1 , shown at 306 in fig3 is so slow to fluctuate in time that its \u201c\u2212 1 \u201d portion is off of the chart ( in fig3 ) to the right , and is ignored ; and accordingly , it has the value \u201c+ 1 \u201d for all 32 of the points in time shown in fig3 . it thus represents the average , or d . c ., component of the atrial signal potential . the lowest frequency wavelets w 1 and w 2 encompass the entire set of 32 points in time . the higher frequency wavelets w 3 and w 4 each encompasses only half of the points in time , and the two wavelets taken together encompass all the points in time . the four wavelets w 5 , w 6 , w 7 , and w 8 each encompasses only eight points in time , and the four wavelets taken together encompass all points in time . the eight wavelets w 9 . . . w 16 each encompasses only four points in time , and together all eight wavelets encompass all points in time . and finally , the wavelets w 17 . . . w 32 each encompass only two points in time , but the sixteen wavelets together encompass all points in time . thus , each wavelet has a characteristic time span and position as well as a characteristic frequency , with higher - frequency wavelets having a narrower and more specific time span and position than lower - frequency wavelets . this is advantageous with an impulse - type , ringing signal such as the atrial depolarization waveform considered here , since many higher - frequency transform values that correspond to haar wavelets positioned in time away from the p waveform or that do not correspond to its frequency of ringing may be low in amplitude such that they may safely be ignored during the analysis steps , thereby reducing the data that must be processed during the correlation steps . ( the above , while presented in the context of the haar wavelet , is also applicable to other wavelet shapes that may used to perform a discrete wavelet transform ( dwt )). each of the 32 haar transformed values is computed as follows : multiply each of the 32 sampled and digitized voltage values representing the analog atrial waveform 302 by the correspondingly - positioned -( in - time ) amplitude values of one of the 32 haar wavelets ( shown in fig3 ); then sum the resulting products ; and then multiply the resulting sums by a discrete wavelet transform scaling factor 2 \u2212 j / 2 , where j equals 1 for the wavelets w 17 to w 32 , 2 for w 9 to w 16 , 3 for w 5 to w 8 , 4 for w 3 to w 4 , and 5 for w 1 and w 2 . for example , and referring to fig3 : the first wavelet w 1 is always + 1 , so the atrial signal potential values are simply summed and then multiplied by 2 \u2212 5 / 2 ; the second wavelet is + 1 for time points 1 to 16 and \u2212 1 for time points 17 to 32 , so the first 16 values are summed , the second 16 values are summed , the difference between the first and second sums is computed , and the result is multiplied by 2 \u2212 5 / 2 ; and so on until all the transformed values have been computed , one for each haar wavelet . the 32 time - sequential atrial signal potential values are thus transformed into 32 haar wavelet amplitude values which may be called \u201c transformed values \u201d and which may be assigned the numbers 1 to 32 corresponding to the subscript numbers of the haar wavelets to which they each correspond and whose amplitudes they represent . the above description of how to compute the haar wavelet transformed values is accurate , but it is not the most efficient way to proceed in the case of haar wavelets . like the fourier transform , which has a corresponding fast fourier transform that saves intermediate results and thereby avoids re - computing them and thus reduces substantially the number of computations , the haar transformation also may be carried out in a manner that saves intermediate sums and re - uses them to reduce substantially the number of computations . this is described below at the point where fig4 is described , since fig4 illustrates graphically how this can be done . the illustrative program listing presented below is also an implementation of the fast haar wavelet transformation algorithm . the haar transformation can readily be carried out by any digital computer . as an example of how the haar transformation can be carried out on 32 values , the following program , written in the language c , is illustrative of many possible programs that may be written . in the exemplary program that follows , a 32 - element array yy contains 32 data values representing the 32 sampled atrial signal potential values representing the fluctuations over time of the signal supplied by the bipolar lead 102 . the analog atrial signal is sampled , digitized , broken into separate beat data sets , pre - processed ( to remove waveforms distorted by ventricular activity ), centered ( with the negative depolarization spike at data point 16 ), and fed into the subroutine presented below . this subroutine is compiled ( or assembled , if rewritten in assembly language ) and installed in the rom of the icd &# 39 ; s embedded microprocessor . this subroutine returns , contained within the same array yy , the 32 transformed haar wavelet amplitude values described above . ( the constant value \u201c recipsqrttwo \u201d is the reciprocal of the square root of two ). an illustrative version of the subroutine for computing all 32 of the haar transformed values in an efficient manner is presented here : void haar ( double yy []) { int i , j , l ; double zz [ 32 ]; for ( i = 5 ; i & gt ; 0 ; i \u2212\u2212) { l = 1 ; for ( j = 1 ; j & lt ;= i ; ++ j ) l = 2 * l ; for ( j = 0 ; j & lt ; l ; ++ j ) zz [ j ] = yy [ j ]; for ( j = 0 ; j & lt ; l \u2212 1 ; j = j + 2 ) ( yy [ j / 2 ] = recipsqrttwo * ( zz [ j ] + zz [ j + 1 ]); yy [( j + l )/ 2 ] = rcipsqrttwo * ( zz [ j ] \u2212 zz [ j + 1 ]); } } return ; } the above program computes the 32 haar transformed values 1 to 32 from the 32 atrial signal potential time - sequenced input values . it does so with only 31 additions , 31 subtractions , and 62 multiplications . it is carefully designed to compute each value in a systematic way , making multiple use of intermediate results . first , the program computes sixteen sums of and sixteen differences between eight adjacent pairs of the 16 time sequenced atrial signal potential values , and the sixteen difference values become the haar transformed values 17 to 32 , reflecting the strength of the highest frequency haar wavelets positioned at 16 different positions in time , corresponding to the wavelets w 17 to w 32 shown in fig3 . all the sum and difference values are scaled by multiplication by 2 \u2212 1 / 2 . next , taking the 16 sums of atrial signal potential values as an intermediate result , the above algorithm generates eight sums of and eight differences between adjacent pairs of these sixteen intermediate values that resulted from the first sixteen additions , and the eight newly - computed difference values become the haar transformed values 9 to 16 , reflecting the strength of the second to the highest frequency values at eight different points in time , corresponding to wavelets w 9 to w 16 in fig3 . all of these eight sum and eight difference values are again scaled by 2 \u2212 1 / 2 so that the wavelets w 9 to w 16 are scaled by \u00bd ( 2 \u2212 2 / 2 or 2 \u2212 1 / 2 times 2 \u2212 1 / 2 ). next , taking these eight sums of sums of adjacent atrial signal potential values as an intermediate result , the above algorithm generates four sum and four difference values , again scaling by 2 \u2212 1 / 2 , and the four difference values become haar transformed values 5 to 8 , reflecting the strength of the middle frequency wavelets at four different points in time , and corresponding to wavelets w 5 to w 8 in fig3 . next , taking these four sums of sums of sums of adjacent atrial signal potential values , the above algorithm generates two sum and two difference values , again scaling by 2 \u2212 1 / 2 , and the two difference values become haar transformed values 3 and 4 , reflecting the strength of the second to the lowest frequency wavelets only two of which encompass all the time domain data , corresponding to the wavelets w 3 and w 4 shown in fig3 . and finally , the algorithm generates the sum of and the difference between the final remaining two sums of sums of sums of sums of atrial signal potential values , again scaling by 2 \u2212 1 / 2 . the difference value is then the haar transformed value 2 , representing the strength of the lowest - frequency wavelet , the one corresponding to the wavelet w 2 in fig3 the wavelet that extends the full length of the time scale . the sum value is then the first , or d . c ., transformed value , representing the average signal potential level over the 32 sampled points in time , which corresponds to the wavelet w 1 in fig3 . another way of viewing this computation is illustrated in fig4 a and 4b . the 32 atrial signal potential values are shown at the top of fig4 a and are identified as c 0 5 through c 31 5 . the superscript \u201c 5 \u201d indicates that these values are processed when the index value \u201c n \u201d in the above computer program is equal to \u201c 5 \u201d\u2014 that is , during the first pass through the data generating intermediary sums and differences . during subsequent passes , the value of n is decremented to 4 , 3 , 2 , and finally to 1 . during each pass through the data , the computer generates sums c n - 1 and differences d n - 1 , as shown in fig4 b , between adjacent pairs of the values c 0 n and c 1 n ; c 2 n and c 3 n ; and so on , so that the number of newly - generated c n - 1 and d n - 1 terms is reduced by half with each computer pass through the intermediary results . at the end of all these computations , the value c 0 0 is the first , or d . c ., haar transformed value ; and the values d 0 0 ; d 0 1 and d 1 1 ; d 0 2 , d 1 2 , d 2 2 , and d 3 2 ; d 0 3 , d 1 3 , . . . and d 7 3 ; and d 0 4 , d 1 4 , and d 15 4 are , respectively , the remaining haar transform values 2 through 32 . fig4 thus illustrates quite succinctly how all the transform computations are carried out , and how the intermediary \u201c c \u201d sum values , such as c 0 4 , are used to compute multiple haar values , such as the values d 0 3 , d 0 2 , d 0 1 , d 0 0 , and c 0 0 all of which are computed from the intermediary value c 0 4 . saving and reusing these intermediary \u201c c \u201d values saves much computational time . fig4 b indicates the precise addition and subtraction operations that are carried out at each level to compute the values in the next lower level , proceeding down through the chart presented in fig4 a . but in any given application to heart waveform analysis , all of these computations may not be needed , and accordingly the number of computations may be reduced much further . since the present invention teaches that only a small number of these haar transformed values need actually be considered , a far less computationally intensive transform can be developed which only generates the intermediate and final transformed values that are actually needed to generate the specific haar transformed values which have proved to be significant in discriminating between sinus tachycardia ( or st ) and non - st conditions . all others need not be computed , and the above program may be reduced to a special algorithm that omits as many sums , differences , and multiplications as possible . for example , if only the transformed values 1 , 5 , 9 , and 24 are significant , then 31 additions are required to compute the first transformed value ( the d . c . value \u2014 the simple sum of all the time domain signal values ); six additions and one subtraction are required to compute the fifth transformed value ( in fig3 the difference between the sums of time domain values under each half of the wavelet w 5 at 314 in fig3 ); two additions and one subtraction are required to compute the ninth transformed value ( in fig3 the difference between the sums of the time domain values under each half of the wavelet w9 at 322 in fig3 ); and only one subtraction is required to compute the 24 th transformed value ( in fig3 the difference between the two time domain values under the respective halves of the wavelet w 24 ( not shown ) which is the same size as , but differently positioned in time than , the wavelet w 18 at 330 ( fig3 ). accordingly , if only the transformed values 1 , 5 , 9 , and 24 actually evaluated , then only 42 additions and subtractions are required . but even this number can be reduced further . if the computational algorithm set forth in the above illustrative computer program is followed as a guide , then the intermediary sums used in computing the haar first , or d . c ., transformed value can be re - used to compute the sums for the haar transformed values 5 and 9 , assuming these intermediary results are saved in the manner described in the above program example . then 31 additions are still required to compute the first transformed value , but only one subtraction is required to compute each of the transformed values 5 , 9 , and 24 , giving a total of additions and subtraction of only 34 operations . and if only transformed values 1 , 5 , and 9 are computed , the number of additions and subtractions is reduced to 33 . and if the first transformed value is omitted and if transformed values 5 , 9 , and 24 are selected , then the 31 additions needed to compute the first transformed value are not required . then the number of computations for transformed value 5 is 8 additions and one subtraction ; the number of computations for transformed value 9 is 2 additions and one subtraction ; and the number of computations for transformed value 24 is still just one subtraction . so the total number of additions and subtractions is just 13 . but even this number can be reduced to 11 when it is realized that the two additions done for the transformed value 9 are also done ( in the above computational algorithm ) when computing the transformed value 5 . so the number of additions and subtractions can be reduced to 11 . also , when computing such a small number of transformed values , the number of multiplications can be reduced as well by postponing the multiplications until a transformed value is actually computed . with reference to fig4 a , instead of dividing every sum and difference by the square root of two ( as shown in fig4 b ), several vertical sums of \u201c c \u201d values in different rows of the fig4 a table can be formed , and then a \u201c d \u201d value can be computed by subtraction , and then the resulting unscaled transformed value can be scaled with a single multiplication by 2 \u2212 j / 2 where j is the number of rows in the table of fig4 a traversed by the one or more sum and the one difference computations . thus , the number of scaling multiplications can sometimes be equal to the number of transformed values that are computed , typically 3 or 4 . accordingly , by using wavelet analysis and transformation , instead of fourier or discrete cosine or other sinusoid analysis and transformation , a highly useful and highly frequency - specific result can be achieved with a very small number of computations , thereby conserving power and battery life , and yet achieving a high degree of precision in recognizing changes in morphology . in our tests , we have selected certain transformed values as being much more significant than other values in distinguishing between st and non - st . we focused upon those transform values whose \u201c+ 1 \u201d and \u201c\u2212 1 \u201d scope included the middle 8 time points most significant to analysis of the peak of the atrial depolarization . working across test data samples obtained from patients who had dual chamber icds , we computed the variance of each transformed value for st and non - st events using the standard statistical formula for computing variance . we then calculated the ratio of the variance of non - st events to the variance of st events , and selected those terms with the highest variance ratios for further consideration . in this manner , we reduced the number of transformed values that were included in the cwa or correlation process while always maintaining or improving the performance achieved . ultimately we settled upon the three or four transformed values having the highest ratios . these were the transform values 1 , 5 , 9 , and 24 . we tested 1 , 5 , and 24 together ; 5 , 9 , and 24 together ; and 1 , 5 , 9 , and 24 together . these selected haar transform values , when used in these combinations , required minimal computations and thereby produced a savings in battery power , and that also gave better performance at distinguishing st from non - st than did all of the transformed values used together . so an increase in accuracy was achieved as well as a decrease in computational complexity by this approach to atrial waveform analysis . [ 0054 ] fig5 for example , illustrates actual plots of digital information illustrating the effect of using only the three coefficients 1 , 5 , and 24 . at 502 , a normal waveform is shown , with amplitude plotted against sample numbers from 1 to 32 . at 504 , an abnormal waveform is shown , again with amplitude plotted against sample numbers . after the haar transformation , at 506 and 508 , the wavelet coefficient amplitude values are indicated for the coefficients 1 to 32 . at 506 , the coefficients generated by the normal waveform 502 are shown , and at 508 , the coefficients generated by the abnormal waveform 504 are shown . one may directly compare these component values and verify that the coefficients 1 , 5 , and 24 at 506 and at 508 vary significantly in amplitude . comparison of this same information among different patients ( not shown in this figure ) also indicated that this variation is relatively constant from one patient to another . at 510 and 512 , using only the coefficients 1 , 5 , and 24 , the heartbeat waveforms are reconstructed by a reverse transformation , the normal reconstructed waveform shown at 510 and the abnormal reconstructed waveform shown at 512 . the marked differences between these two reconstructed waveforms highlights the way in which these three coefficients can signal an abnormal condition with less computation . in a practical system , a small number of transform values are selected , in the manner just described . the algorithm for computing these values is then refined , as explained above , to reduce the number of additions , subtractions , and multiplications to the minimum possible while preserving the accuracy of the computations . we first compute a value \u03c1 , which is the correlation between these values with the corresponding values in a template that represents the values for an average normal population . we then compare this value \u03c1 to a threshold correlation value \u03b2 that is chosen to give optimal results on experimental data . ( see the examples in the tables presented below .) for each waveform analyzed , a test is made to determine whether \u03c1 is greater than \u03b2 . if so , then this waveform is placed into the buffer marked \u201c st \u201d at step 210 . if not , then this waveform is placed into the buffer marked \u201c non - st \u201d at step 211 . finally , after ten waveforms have been analyzed , a count of st and non - st marked waveforms is made to see if the count of non - st waveforms is greater than some threshold value x ( at step 212 ), where x can be , for example , 7 . if more than seven waveforms are non - st , then therapy is delivered at step 214 . the buffers at steps 210 and 211 in fig2 are buffers that hold the last ten decision values . these buffers can be pictured as sliding windows revealing the most recent ten values to permit the decision at 212 to be continuously updated . the buffers thus function as a nonlinear filter preventing irregular values from delivering therapy improperly . in developing a working prototype of this system , we used a development data set consisting of 20 episodes of st taken from 4 patients and 18 episodes of af taken from 7 patients . patient number episodes of st episodes of af 1 2 6 2 12 0 13 0 1 4 0 2 5 5 1 6 0 3 7 0 3 8 0 2 9 1 0 total 20 18 we used this data to optimize our choice of \u03c1 and \u03b2 and also the particular coefficients that we chose to examine . next , we tested our prototype using a set of 16 episodes of st obtained from 4 patients together with 17 episodes of af obtained from 7 patients . patient number episodes of st episodes of af 1 8 0 2 3 1 3 3 0 4 2 0 5 0 4 6 0 4 7 0 3 8 0 3 9 0 1 10 0 1 total 16 17 x out threshold coefficients of 10 ( \u03b2 ) sensitivity specificity 1 , 5 , 9 , & amp ; 24 3 0 . 808 76 % 94 % 1 , 5 , 9 , & amp ; 24 5 0 . 975 88 % 94 % 1 , 5 , 9 , & amp ; 24 6 0 . 976 82 % 94 % 1 , 5 , 9 , & amp ; 24 7 0 . 983 88 % 94 % 1 , 5 , & amp ; 24 3 0 . 815 82 % 94 % 1 , 5 , & amp ; 24 4 0 . 943 88 % 94 % 1 , 5 , & amp ; 24 5 0 . 956 82 % 94 % 1 , 5 , & amp ; 24 7 0 . 991 94 % 94 % 5 , 9 , & amp ; 24 8 0 . 9993 82 % 94 % 5 , 9 , & amp ; 24 9 0 . 9997 76 % 81 % the quality of the selected subset of transformed values for characterizing changes in morphology may be demonstrated graphically , as indicated in fig5 by performing a reverse haar transform using only the three or four or so selected transformed values . as can be seen ( at 510 and 512 in fig5 ), the inverse transformations produce distorted representations of the original atrial waveforms which illustrate how the choice of transform values can emphasize the changes . this is another useful way to assist one in selecting which transformed values are most useful . one feature of the invention is its ability to use a template derived from an average normal population , rather than deriving a patient customized average normal template for each individual patient . prior systems have required each new patient to be monitored while in a normal state , and the captured data was then averaged to form a patient specific normal template . contrary to this , the present invention achieved the results shown above using the same template for all patients . accordingly , the invention may be used with a new patient without the necessity of such preliminary testing of each patient and without a new customized template necessarily having to be created for each patient . one reason why the present invention can function using a template that represents average values for a normal population is because the specific coefficients selected in the transform domain , in addition to having been selected to emphasize the difference between normal and abnormal values , are also preferably chosen to minimize the differences between different patients . in some cases , values that were good at distinguishing between normal and abnormal rhythms for one patient were not as good at doing so for some other patient . these values are preferably not selected for use in implementing the present invention . for example , the transform coefficients may be selected such that all ( or most of ) the normal values of a particular coefficient gathered from many patients had the same sign ( positive or negative ) and similar amplitudes ( or if they varied in sign , they were near to zero ). in addition , the abnormal coefficients are significantly different in value from the normal coefficients for each particular patient . [ 0068 ] fig6 at 600 , illustrates one way in which this may be done . first , at step 602 , one gathers a large number of st and non - st data samples . next , all of this data is transformed , generating coefficients for every waveform set of data ( step 604 ). the variance of each coefficient is then computed first for the st samples ( step 606 ) and then for the non - st samples . then , for each coefficient , the ratio of the variance of the non - st samples to that of the st samples is computed ( step 608 ). finally , those coefficients whose variance ratio was large may be selected a coefficients for use in creating a population template and in testing patients ( step 612 ). in addition , at step 614 , the number of coefficients retained may be further reduced to reduce the number of computations that need to be performed , as explained above . in the version of the invention that was used to generate the data shown above , the population template was computed as follows , using the 20 st episodes in the development data set ( also used in the first of the two tables presented above ). this is illustrated at 700 in fig7 . first , the coefficients were selected as was explained above and as shown in fig6 . next , for each patient episode of st ( step 702 ), the recorded heartbeats were broken up into groups of ten consecutive heartbeats ( step 704 ). for each group ( step 706 ), the selected coefficients of the haar transform were computed ( step 708 ) for each of the ten heartbeats and the median value was then selected ( step 710 ) for each coefficient to form a proto - template . if these median values correlated well with the coefficient values for each of the ten heartbeats , the proto - template was retained ( step 714 ). otherwise , it was discarded . in this manner , a whole bunch of proto - templates were obtained from each patient episode . next , median values of the coefficients from all of the proto - templates ( step 716 ) were computed ( step 718 ). these values were then used as the population template for distinguishing st from non - st events ( step 720 ). while the preferred embodiment of the invention has been described above , it is to be understood that numerous modifications and changes will occur to those who are skilled in the art to which the invention pertains . accordingly , the following claims annexed to and forming a part of this specification are intended to define the true spirit and scope of the invention \u2014 that is , what is new and what is desired to be secured by letters patent of the united states ."}
|
Is the categorization of this patent accurate?
| 0.25 |
1cd5305994f1692ec7ec09368a2a725ff3c3dd15c21bf56c12e04e558cdb5fbc
| 0.123535 | 0.015869 | 0.066406 | 0.004913 | 0.138672 | 0.108398 |
null |
{"patent": "the present invention is directed towards improving the quality and functionality of both implantable cardioverter defibrillators ( icds ) and implantable atrial defibrillators ( iads ). all such devices need to have a method and mechanism for accurately distinguishing sinus tachycardia ( st ) form non - st , such as atrial fibrillation and atrial flutter . this method and mechanism should be conservative , biased towards not initiating therapy unless non - st is clearly indicated . it should also require as few machine computational cycles as possible to reduce icd and iad battery drain . the principle underlying the present invention is that st and non - st can be distinguished by studying and measuring the atrial electrogram morphology to detect aberrant conduction in the atria . thus , by comparing normal st waveforms to a series of newly - occurring waveforms , st is indicated by abnormal atrial waveform morphology which signals aberrant conduction within the atria improper electrical patterns of depolarization . this will show up in the signal potentials captured by a bipolar atrial lead . with reference to fig1 the present invention can be implemented in an icd 100 ( which could be an iad ) that is implanted in the chest of a patient who suffers from unpleasant and possibly life - threatening irregularities in heart operation due to improper electrical stimulation of the heart muscle cells . during normal heart action , the heart &# 39 ; s electrical impulses originate in the sino - atrial node as an action potential that is transmitted smoothly to all portions of the atria , causing contraction of the atrial chambers . the electrical impulse continues in its path to a cluster of conduction fibrils known as the atrioventricular node . after a delay of about one - tenth of a second , an action potential flows over the ventricles and causes them to contract in synchronism with and following shortly after the atrial contraction . in this manner , the heart pumps blood to the lungs and from the lungs to the body . a bipolar lead 102 is implanted within the atria 104 of a heart 103 to measure the electrical potential between nearby cells of the atria . as the action potential actively passes from cell to cell past the two electrodes of the lead 102 , an oscillating signal potential is developed across the two electrodes and is conveyed by the bipolar lead 102 to the icd 100 where the signal is sampled about 1000 times per second and is digitized , with the sequential samples stored within a ram memory within the icd 100 for further analysis . another bipolar lead 105 may extend to the ventricle ( not shown ) of the heart 103 to measure the action potentials generated within the ventricles . these same leads , or other leads , may be used by the icd for administering various types of therapy , such as pacing or defibrillation shock therapy . the data samples collected from within the atria are now analyzed . first , with reference to data gathered from the ventricles , if some event in the ventricle , such as a mis - timed depolarization event that partially overlaps the atria &# 39 ; s p wave , could leak across to the atria and distort the measurement of the potentials for a given atrial depolarization , then the data for that particular heartbeat is discarded and is not used in further analyses . such distortion can also be found simply because a given heartbeat set of data gathered from the atria is itself badly distorted , and this approach must be taken in the case of an iad having no ventricular lead . the remaining data is retained for further processing . in preparation for further processing , the data for individual heartbeats is identified and gathered . the position within the gathered data of the negative spike to the atrial p depolarization waveform is determined and is selected as the center of the waveform for each beat . then , since data close to this p - wave is to be analyzed , and since data further away may be distorted to some degree by ventricular activity or by variations in the p - to - p interval , 32 sequential data samples are selected for each heartbeat such that the negative spike of the p - wave becomes signal potential sample number 16 of the 32 sequential samples selected for further analyses . these samples for a normal heartbeat appear at 302 in fig3 and at 502 in fig5 where the amplitude of the atrial signal potential is plotted against time over the 32 sampled values , with the negative p - wave spike positioned as signal sample number 16 . likewise , the samples for an abnormal heartbeat appear at 504 in fig5 . straightforward correlation ( cwa ) of the 32 samples against a template containing a typical or normal pattern could be performed at this juncture , as taught on page 563 ( equation ( 1 )) of the article by thorne , cited in the introductory portion of this specification . but the performance of such a cross correlation at repeated regular intervals would generate numerous computations and would drain the icd &# 39 ; s battery more rapidly than is desirable . in addition , provision would have to be made for a full template waveform that can be stored within the icd to be used as a comparison reference . in addition , such templates generally need to be patient specific , and they are more susceptible to normal changes in shape . accordingly , to reduce the mathematical complexity of the cross correlation computations , it is desirable to reduce the number of data points from 32 down to some much lower number through the use of some form of transformation into a different set of values . for example , the data could be transformed using the karhunen - loeve transformation described in the morris article cited at the beginning of this specification . but that transformation is fairly computationally intensive . computationally , a fast fourier transformation would be more efficient , but any such transformation into the frequency domain , where variable frequency sine and cosine wave amplitudes result from the transformation , does not match itself well to the morphology of atrial heart waveforms , which tend to be formed from large , ringing , spike - like representations of p - wave depolarization events travelling as moving action potential fronts past the pair of electrodes attached to the bipolar lead , and not as steady harmonics extending across time . accordingly , fourier - class transforms do not adequately reduce the number of significant data values that must be considered . and in addition , when the \u201c window \u201d size for a fourier analysis is selected , if it is wide , then the frequency values are not time - specific , but represent average signal harmonic content over the entire window time duration . and if the window is narrow , then the harmonics are too spread out in frequency , and the frequency - domain data generated by the transform does not resolve things finely enough in the frequency domain . for all of the above reasons , the present invention analyzes the waveforms using a discrete wavelet transform . this has the advantage that while low frequency wavelets are broad in time , high frequency wavelets have a much narrower timeslot focus . thus , for example , with 32 data samples taken sequentially over time , a wavelet transformation generates a d . c . wavelet that spans the entire set of 32 points ; a first reversing wavelet that also spans the entire set of points ; two second harmonic wavelets that each focus upon only one - half of the 32 data points ; four third harmonic wavelets that each focus upon only one - fourth of the 32 data points ; eight fourth harmonic wavelets that each focus upon four of the 32 data points ; and 16 fifth harmonic wavelets that each focuses upon only two of the 32 data points . accordingly , the higher - frequency wavelet amplitudes are quite time specific , unlike fourier sinusoidal amplitudes , while the lower - frequency and d . c . wavelet amplitudes give one broad information about the whole set of 32 points . and like the fourier transform , the discrete wavelet transform preserves all of the information of the original atrial signal , such that the transformation is fully reversible . 32 discrete wavelet transform values may be reverse transformed back into 32 values indicating the atrial signal potentials at 32 sequential points in time . in essence , the haar function is performing multiple digital filtering operations , at variable positions and utilizing varying - width windowing functions , to develop component values that represent varying types of heartbeat activity , at varying frequencies , spread over varying widths , and located at varying positions . these component values can be studied to see which subset of these component values are good at distinguishing st from non - st , or ( more generally ) which subset of these component values are good at distinguishing one type of heartbeat waveform from another . further study of such a selected subset can further determine which of the subset of components are relatively invariant from one individual &# 39 ; s heart to another . one preferably selects component values that are suitable from both of these two perspectives . the particular wavelet transformation to choose can be tailored to the nature of the data , with the wavelets chosen to resemble somewhat the values to be found within the data so that some transformed values are of much larger amplitude than others , and such that the low amplitude transformed values may be disregarded . of particular importance with an atrial p waveform of the type being analyzed here are wavelets having frequency values roughly comparable to and timed to coincide with the ringing of the atrial depolarization . this tailoring can minimize the number of transformed data values that are significant and that are candidates for full participation in the correlation analyses to determine if there has been a substantial change in waveform morphology . another factor in selecting a particular wavelet transformation is reducing the number of computations that must be performed to carry out the transformation . yet another factor is selecting a wavelet transformation where some of the transformed component values vary from normal to abnormal heart rhythms in much the same over a population of individuals so that a generic template of these component values may be derived that can be used with many individuals , rather than with just one individual . these factors suggest that the haar transformation would be a suitable candidate for use in analysis of atrial waveforms and possibly other waveforms as well . with reference to fig3 in the case of 32 voltage values sampled over time , the 32 applicable haar discrete transform wavelets appear as shown in this figure . other discrete wavelet transforms based upon wavelets having triangular or other shapes may also prove usable . the haar transform wavelets , in particular , have the shape of a square waveform , as will be described , and this can simplify the computations required . in fig3 time increases from left to right . a scale 304 indicates the 32 points in time at which the atrial waveform is sampled , and an analog atrial waveform ( before digitization ) is shown at 302 . the p wave negative notch 303 is shown centered at the 16 th timeslot so that the signal potential of the atria waveform sampled at this point in time becomes the 16 th data value in the set of 32 data values that are to be subjected to the haar transformation . a haar wavelet is simply one cycle of a square waveform that starts at zero , then swings positive ( to \u201c+ 1 \u201d), then swings negative ( to \u201c\u2212 1 \u201d), and then swings back to zero . the wavelet w 2 , shown at 308 in fig3 for example , is at \u201c+ 1 \u201d at points in time 1 to 16 and is at \u201c\u2212 1 \u201d at points in time 17 to 32 . the shorter waveform w 6 , shown at 316 in fig3 is at \u201c+ 1 \u201d at points in time 9 to 12 and is at \u201c\u2212 1 \u201d at points in time 13 to 16 , and is at \u201c 0 \u201d at all other points in time . the waveform w 1 , shown at 306 in fig3 is so slow to fluctuate in time that its \u201c\u2212 1 \u201d portion is off of the chart ( in fig3 ) to the right , and is ignored ; and accordingly , it has the value \u201c+ 1 \u201d for all 32 of the points in time shown in fig3 . it thus represents the average , or d . c ., component of the atrial signal potential . the lowest frequency wavelets w 1 and w 2 encompass the entire set of 32 points in time . the higher frequency wavelets w 3 and w 4 each encompasses only half of the points in time , and the two wavelets taken together encompass all the points in time . the four wavelets w 5 , w 6 , w 7 , and w 8 each encompasses only eight points in time , and the four wavelets taken together encompass all points in time . the eight wavelets w 9 . . . w 16 each encompasses only four points in time , and together all eight wavelets encompass all points in time . and finally , the wavelets w 17 . . . w 32 each encompass only two points in time , but the sixteen wavelets together encompass all points in time . thus , each wavelet has a characteristic time span and position as well as a characteristic frequency , with higher - frequency wavelets having a narrower and more specific time span and position than lower - frequency wavelets . this is advantageous with an impulse - type , ringing signal such as the atrial depolarization waveform considered here , since many higher - frequency transform values that correspond to haar wavelets positioned in time away from the p waveform or that do not correspond to its frequency of ringing may be low in amplitude such that they may safely be ignored during the analysis steps , thereby reducing the data that must be processed during the correlation steps . ( the above , while presented in the context of the haar wavelet , is also applicable to other wavelet shapes that may used to perform a discrete wavelet transform ( dwt )). each of the 32 haar transformed values is computed as follows : multiply each of the 32 sampled and digitized voltage values representing the analog atrial waveform 302 by the correspondingly - positioned -( in - time ) amplitude values of one of the 32 haar wavelets ( shown in fig3 ); then sum the resulting products ; and then multiply the resulting sums by a discrete wavelet transform scaling factor 2 \u2212 j / 2 , where j equals 1 for the wavelets w 17 to w 32 , 2 for w 9 to w 16 , 3 for w 5 to w 8 , 4 for w 3 to w 4 , and 5 for w 1 and w 2 . for example , and referring to fig3 : the first wavelet w 1 is always + 1 , so the atrial signal potential values are simply summed and then multiplied by 2 \u2212 5 / 2 ; the second wavelet is + 1 for time points 1 to 16 and \u2212 1 for time points 17 to 32 , so the first 16 values are summed , the second 16 values are summed , the difference between the first and second sums is computed , and the result is multiplied by 2 \u2212 5 / 2 ; and so on until all the transformed values have been computed , one for each haar wavelet . the 32 time - sequential atrial signal potential values are thus transformed into 32 haar wavelet amplitude values which may be called \u201c transformed values \u201d and which may be assigned the numbers 1 to 32 corresponding to the subscript numbers of the haar wavelets to which they each correspond and whose amplitudes they represent . the above description of how to compute the haar wavelet transformed values is accurate , but it is not the most efficient way to proceed in the case of haar wavelets . like the fourier transform , which has a corresponding fast fourier transform that saves intermediate results and thereby avoids re - computing them and thus reduces substantially the number of computations , the haar transformation also may be carried out in a manner that saves intermediate sums and re - uses them to reduce substantially the number of computations . this is described below at the point where fig4 is described , since fig4 illustrates graphically how this can be done . the illustrative program listing presented below is also an implementation of the fast haar wavelet transformation algorithm . the haar transformation can readily be carried out by any digital computer . as an example of how the haar transformation can be carried out on 32 values , the following program , written in the language c , is illustrative of many possible programs that may be written . in the exemplary program that follows , a 32 - element array yy contains 32 data values representing the 32 sampled atrial signal potential values representing the fluctuations over time of the signal supplied by the bipolar lead 102 . the analog atrial signal is sampled , digitized , broken into separate beat data sets , pre - processed ( to remove waveforms distorted by ventricular activity ), centered ( with the negative depolarization spike at data point 16 ), and fed into the subroutine presented below . this subroutine is compiled ( or assembled , if rewritten in assembly language ) and installed in the rom of the icd &# 39 ; s embedded microprocessor . this subroutine returns , contained within the same array yy , the 32 transformed haar wavelet amplitude values described above . ( the constant value \u201c recipsqrttwo \u201d is the reciprocal of the square root of two ). an illustrative version of the subroutine for computing all 32 of the haar transformed values in an efficient manner is presented here : void haar ( double yy []) { int i , j , l ; double zz [ 32 ]; for ( i = 5 ; i & gt ; 0 ; i \u2212\u2212) { l = 1 ; for ( j = 1 ; j & lt ;= i ; ++ j ) l = 2 * l ; for ( j = 0 ; j & lt ; l ; ++ j ) zz [ j ] = yy [ j ]; for ( j = 0 ; j & lt ; l \u2212 1 ; j = j + 2 ) ( yy [ j / 2 ] = recipsqrttwo * ( zz [ j ] + zz [ j + 1 ]); yy [( j + l )/ 2 ] = rcipsqrttwo * ( zz [ j ] \u2212 zz [ j + 1 ]); } } return ; } the above program computes the 32 haar transformed values 1 to 32 from the 32 atrial signal potential time - sequenced input values . it does so with only 31 additions , 31 subtractions , and 62 multiplications . it is carefully designed to compute each value in a systematic way , making multiple use of intermediate results . first , the program computes sixteen sums of and sixteen differences between eight adjacent pairs of the 16 time sequenced atrial signal potential values , and the sixteen difference values become the haar transformed values 17 to 32 , reflecting the strength of the highest frequency haar wavelets positioned at 16 different positions in time , corresponding to the wavelets w 17 to w 32 shown in fig3 . all the sum and difference values are scaled by multiplication by 2 \u2212 1 / 2 . next , taking the 16 sums of atrial signal potential values as an intermediate result , the above algorithm generates eight sums of and eight differences between adjacent pairs of these sixteen intermediate values that resulted from the first sixteen additions , and the eight newly - computed difference values become the haar transformed values 9 to 16 , reflecting the strength of the second to the highest frequency values at eight different points in time , corresponding to wavelets w 9 to w 16 in fig3 . all of these eight sum and eight difference values are again scaled by 2 \u2212 1 / 2 so that the wavelets w 9 to w 16 are scaled by \u00bd ( 2 \u2212 2 / 2 or 2 \u2212 1 / 2 times 2 \u2212 1 / 2 ). next , taking these eight sums of sums of adjacent atrial signal potential values as an intermediate result , the above algorithm generates four sum and four difference values , again scaling by 2 \u2212 1 / 2 , and the four difference values become haar transformed values 5 to 8 , reflecting the strength of the middle frequency wavelets at four different points in time , and corresponding to wavelets w 5 to w 8 in fig3 . next , taking these four sums of sums of sums of adjacent atrial signal potential values , the above algorithm generates two sum and two difference values , again scaling by 2 \u2212 1 / 2 , and the two difference values become haar transformed values 3 and 4 , reflecting the strength of the second to the lowest frequency wavelets only two of which encompass all the time domain data , corresponding to the wavelets w 3 and w 4 shown in fig3 . and finally , the algorithm generates the sum of and the difference between the final remaining two sums of sums of sums of sums of atrial signal potential values , again scaling by 2 \u2212 1 / 2 . the difference value is then the haar transformed value 2 , representing the strength of the lowest - frequency wavelet , the one corresponding to the wavelet w 2 in fig3 the wavelet that extends the full length of the time scale . the sum value is then the first , or d . c ., transformed value , representing the average signal potential level over the 32 sampled points in time , which corresponds to the wavelet w 1 in fig3 . another way of viewing this computation is illustrated in fig4 a and 4b . the 32 atrial signal potential values are shown at the top of fig4 a and are identified as c 0 5 through c 31 5 . the superscript \u201c 5 \u201d indicates that these values are processed when the index value \u201c n \u201d in the above computer program is equal to \u201c 5 \u201d\u2014 that is , during the first pass through the data generating intermediary sums and differences . during subsequent passes , the value of n is decremented to 4 , 3 , 2 , and finally to 1 . during each pass through the data , the computer generates sums c n - 1 and differences d n - 1 , as shown in fig4 b , between adjacent pairs of the values c 0 n and c 1 n ; c 2 n and c 3 n ; and so on , so that the number of newly - generated c n - 1 and d n - 1 terms is reduced by half with each computer pass through the intermediary results . at the end of all these computations , the value c 0 0 is the first , or d . c ., haar transformed value ; and the values d 0 0 ; d 0 1 and d 1 1 ; d 0 2 , d 1 2 , d 2 2 , and d 3 2 ; d 0 3 , d 1 3 , . . . and d 7 3 ; and d 0 4 , d 1 4 , and d 15 4 are , respectively , the remaining haar transform values 2 through 32 . fig4 thus illustrates quite succinctly how all the transform computations are carried out , and how the intermediary \u201c c \u201d sum values , such as c 0 4 , are used to compute multiple haar values , such as the values d 0 3 , d 0 2 , d 0 1 , d 0 0 , and c 0 0 all of which are computed from the intermediary value c 0 4 . saving and reusing these intermediary \u201c c \u201d values saves much computational time . fig4 b indicates the precise addition and subtraction operations that are carried out at each level to compute the values in the next lower level , proceeding down through the chart presented in fig4 a . but in any given application to heart waveform analysis , all of these computations may not be needed , and accordingly the number of computations may be reduced much further . since the present invention teaches that only a small number of these haar transformed values need actually be considered , a far less computationally intensive transform can be developed which only generates the intermediate and final transformed values that are actually needed to generate the specific haar transformed values which have proved to be significant in discriminating between sinus tachycardia ( or st ) and non - st conditions . all others need not be computed , and the above program may be reduced to a special algorithm that omits as many sums , differences , and multiplications as possible . for example , if only the transformed values 1 , 5 , 9 , and 24 are significant , then 31 additions are required to compute the first transformed value ( the d . c . value \u2014 the simple sum of all the time domain signal values ); six additions and one subtraction are required to compute the fifth transformed value ( in fig3 the difference between the sums of time domain values under each half of the wavelet w 5 at 314 in fig3 ); two additions and one subtraction are required to compute the ninth transformed value ( in fig3 the difference between the sums of the time domain values under each half of the wavelet w9 at 322 in fig3 ); and only one subtraction is required to compute the 24 th transformed value ( in fig3 the difference between the two time domain values under the respective halves of the wavelet w 24 ( not shown ) which is the same size as , but differently positioned in time than , the wavelet w 18 at 330 ( fig3 ). accordingly , if only the transformed values 1 , 5 , 9 , and 24 actually evaluated , then only 42 additions and subtractions are required . but even this number can be reduced further . if the computational algorithm set forth in the above illustrative computer program is followed as a guide , then the intermediary sums used in computing the haar first , or d . c ., transformed value can be re - used to compute the sums for the haar transformed values 5 and 9 , assuming these intermediary results are saved in the manner described in the above program example . then 31 additions are still required to compute the first transformed value , but only one subtraction is required to compute each of the transformed values 5 , 9 , and 24 , giving a total of additions and subtraction of only 34 operations . and if only transformed values 1 , 5 , and 9 are computed , the number of additions and subtractions is reduced to 33 . and if the first transformed value is omitted and if transformed values 5 , 9 , and 24 are selected , then the 31 additions needed to compute the first transformed value are not required . then the number of computations for transformed value 5 is 8 additions and one subtraction ; the number of computations for transformed value 9 is 2 additions and one subtraction ; and the number of computations for transformed value 24 is still just one subtraction . so the total number of additions and subtractions is just 13 . but even this number can be reduced to 11 when it is realized that the two additions done for the transformed value 9 are also done ( in the above computational algorithm ) when computing the transformed value 5 . so the number of additions and subtractions can be reduced to 11 . also , when computing such a small number of transformed values , the number of multiplications can be reduced as well by postponing the multiplications until a transformed value is actually computed . with reference to fig4 a , instead of dividing every sum and difference by the square root of two ( as shown in fig4 b ), several vertical sums of \u201c c \u201d values in different rows of the fig4 a table can be formed , and then a \u201c d \u201d value can be computed by subtraction , and then the resulting unscaled transformed value can be scaled with a single multiplication by 2 \u2212 j / 2 where j is the number of rows in the table of fig4 a traversed by the one or more sum and the one difference computations . thus , the number of scaling multiplications can sometimes be equal to the number of transformed values that are computed , typically 3 or 4 . accordingly , by using wavelet analysis and transformation , instead of fourier or discrete cosine or other sinusoid analysis and transformation , a highly useful and highly frequency - specific result can be achieved with a very small number of computations , thereby conserving power and battery life , and yet achieving a high degree of precision in recognizing changes in morphology . in our tests , we have selected certain transformed values as being much more significant than other values in distinguishing between st and non - st . we focused upon those transform values whose \u201c+ 1 \u201d and \u201c\u2212 1 \u201d scope included the middle 8 time points most significant to analysis of the peak of the atrial depolarization . working across test data samples obtained from patients who had dual chamber icds , we computed the variance of each transformed value for st and non - st events using the standard statistical formula for computing variance . we then calculated the ratio of the variance of non - st events to the variance of st events , and selected those terms with the highest variance ratios for further consideration . in this manner , we reduced the number of transformed values that were included in the cwa or correlation process while always maintaining or improving the performance achieved . ultimately we settled upon the three or four transformed values having the highest ratios . these were the transform values 1 , 5 , 9 , and 24 . we tested 1 , 5 , and 24 together ; 5 , 9 , and 24 together ; and 1 , 5 , 9 , and 24 together . these selected haar transform values , when used in these combinations , required minimal computations and thereby produced a savings in battery power , and that also gave better performance at distinguishing st from non - st than did all of the transformed values used together . so an increase in accuracy was achieved as well as a decrease in computational complexity by this approach to atrial waveform analysis . [ 0054 ] fig5 for example , illustrates actual plots of digital information illustrating the effect of using only the three coefficients 1 , 5 , and 24 . at 502 , a normal waveform is shown , with amplitude plotted against sample numbers from 1 to 32 . at 504 , an abnormal waveform is shown , again with amplitude plotted against sample numbers . after the haar transformation , at 506 and 508 , the wavelet coefficient amplitude values are indicated for the coefficients 1 to 32 . at 506 , the coefficients generated by the normal waveform 502 are shown , and at 508 , the coefficients generated by the abnormal waveform 504 are shown . one may directly compare these component values and verify that the coefficients 1 , 5 , and 24 at 506 and at 508 vary significantly in amplitude . comparison of this same information among different patients ( not shown in this figure ) also indicated that this variation is relatively constant from one patient to another . at 510 and 512 , using only the coefficients 1 , 5 , and 24 , the heartbeat waveforms are reconstructed by a reverse transformation , the normal reconstructed waveform shown at 510 and the abnormal reconstructed waveform shown at 512 . the marked differences between these two reconstructed waveforms highlights the way in which these three coefficients can signal an abnormal condition with less computation . in a practical system , a small number of transform values are selected , in the manner just described . the algorithm for computing these values is then refined , as explained above , to reduce the number of additions , subtractions , and multiplications to the minimum possible while preserving the accuracy of the computations . we first compute a value \u03c1 , which is the correlation between these values with the corresponding values in a template that represents the values for an average normal population . we then compare this value \u03c1 to a threshold correlation value \u03b2 that is chosen to give optimal results on experimental data . ( see the examples in the tables presented below .) for each waveform analyzed , a test is made to determine whether \u03c1 is greater than \u03b2 . if so , then this waveform is placed into the buffer marked \u201c st \u201d at step 210 . if not , then this waveform is placed into the buffer marked \u201c non - st \u201d at step 211 . finally , after ten waveforms have been analyzed , a count of st and non - st marked waveforms is made to see if the count of non - st waveforms is greater than some threshold value x ( at step 212 ), where x can be , for example , 7 . if more than seven waveforms are non - st , then therapy is delivered at step 214 . the buffers at steps 210 and 211 in fig2 are buffers that hold the last ten decision values . these buffers can be pictured as sliding windows revealing the most recent ten values to permit the decision at 212 to be continuously updated . the buffers thus function as a nonlinear filter preventing irregular values from delivering therapy improperly . in developing a working prototype of this system , we used a development data set consisting of 20 episodes of st taken from 4 patients and 18 episodes of af taken from 7 patients . patient number episodes of st episodes of af 1 2 6 2 12 0 13 0 1 4 0 2 5 5 1 6 0 3 7 0 3 8 0 2 9 1 0 total 20 18 we used this data to optimize our choice of \u03c1 and \u03b2 and also the particular coefficients that we chose to examine . next , we tested our prototype using a set of 16 episodes of st obtained from 4 patients together with 17 episodes of af obtained from 7 patients . patient number episodes of st episodes of af 1 8 0 2 3 1 3 3 0 4 2 0 5 0 4 6 0 4 7 0 3 8 0 3 9 0 1 10 0 1 total 16 17 x out threshold coefficients of 10 ( \u03b2 ) sensitivity specificity 1 , 5 , 9 , & amp ; 24 3 0 . 808 76 % 94 % 1 , 5 , 9 , & amp ; 24 5 0 . 975 88 % 94 % 1 , 5 , 9 , & amp ; 24 6 0 . 976 82 % 94 % 1 , 5 , 9 , & amp ; 24 7 0 . 983 88 % 94 % 1 , 5 , & amp ; 24 3 0 . 815 82 % 94 % 1 , 5 , & amp ; 24 4 0 . 943 88 % 94 % 1 , 5 , & amp ; 24 5 0 . 956 82 % 94 % 1 , 5 , & amp ; 24 7 0 . 991 94 % 94 % 5 , 9 , & amp ; 24 8 0 . 9993 82 % 94 % 5 , 9 , & amp ; 24 9 0 . 9997 76 % 81 % the quality of the selected subset of transformed values for characterizing changes in morphology may be demonstrated graphically , as indicated in fig5 by performing a reverse haar transform using only the three or four or so selected transformed values . as can be seen ( at 510 and 512 in fig5 ), the inverse transformations produce distorted representations of the original atrial waveforms which illustrate how the choice of transform values can emphasize the changes . this is another useful way to assist one in selecting which transformed values are most useful . one feature of the invention is its ability to use a template derived from an average normal population , rather than deriving a patient customized average normal template for each individual patient . prior systems have required each new patient to be monitored while in a normal state , and the captured data was then averaged to form a patient specific normal template . contrary to this , the present invention achieved the results shown above using the same template for all patients . accordingly , the invention may be used with a new patient without the necessity of such preliminary testing of each patient and without a new customized template necessarily having to be created for each patient . one reason why the present invention can function using a template that represents average values for a normal population is because the specific coefficients selected in the transform domain , in addition to having been selected to emphasize the difference between normal and abnormal values , are also preferably chosen to minimize the differences between different patients . in some cases , values that were good at distinguishing between normal and abnormal rhythms for one patient were not as good at doing so for some other patient . these values are preferably not selected for use in implementing the present invention . for example , the transform coefficients may be selected such that all ( or most of ) the normal values of a particular coefficient gathered from many patients had the same sign ( positive or negative ) and similar amplitudes ( or if they varied in sign , they were near to zero ). in addition , the abnormal coefficients are significantly different in value from the normal coefficients for each particular patient . [ 0068 ] fig6 at 600 , illustrates one way in which this may be done . first , at step 602 , one gathers a large number of st and non - st data samples . next , all of this data is transformed , generating coefficients for every waveform set of data ( step 604 ). the variance of each coefficient is then computed first for the st samples ( step 606 ) and then for the non - st samples . then , for each coefficient , the ratio of the variance of the non - st samples to that of the st samples is computed ( step 608 ). finally , those coefficients whose variance ratio was large may be selected a coefficients for use in creating a population template and in testing patients ( step 612 ). in addition , at step 614 , the number of coefficients retained may be further reduced to reduce the number of computations that need to be performed , as explained above . in the version of the invention that was used to generate the data shown above , the population template was computed as follows , using the 20 st episodes in the development data set ( also used in the first of the two tables presented above ). this is illustrated at 700 in fig7 . first , the coefficients were selected as was explained above and as shown in fig6 . next , for each patient episode of st ( step 702 ), the recorded heartbeats were broken up into groups of ten consecutive heartbeats ( step 704 ). for each group ( step 706 ), the selected coefficients of the haar transform were computed ( step 708 ) for each of the ten heartbeats and the median value was then selected ( step 710 ) for each coefficient to form a proto - template . if these median values correlated well with the coefficient values for each of the ten heartbeats , the proto - template was retained ( step 714 ). otherwise , it was discarded . in this manner , a whole bunch of proto - templates were obtained from each patient episode . next , median values of the coefficients from all of the proto - templates ( step 716 ) were computed ( step 718 ). these values were then used as the population template for distinguishing st from non - st events ( step 720 ). while the preferred embodiment of the invention has been described above , it is to be understood that numerous modifications and changes will occur to those who are skilled in the art to which the invention pertains . accordingly , the following claims annexed to and forming a part of this specification are intended to define the true spirit and scope of the invention \u2014 that is , what is new and what is desired to be secured by letters patent of the united states .", "category": "Human Necessities"}
|
{"patent": "the present invention is directed towards improving the quality and functionality of both implantable cardioverter defibrillators ( icds ) and implantable atrial defibrillators ( iads ). all such devices need to have a method and mechanism for accurately distinguishing sinus tachycardia ( st ) form non - st , such as atrial fibrillation and atrial flutter . this method and mechanism should be conservative , biased towards not initiating therapy unless non - st is clearly indicated . it should also require as few machine computational cycles as possible to reduce icd and iad battery drain . the principle underlying the present invention is that st and non - st can be distinguished by studying and measuring the atrial electrogram morphology to detect aberrant conduction in the atria . thus , by comparing normal st waveforms to a series of newly - occurring waveforms , st is indicated by abnormal atrial waveform morphology which signals aberrant conduction within the atria improper electrical patterns of depolarization . this will show up in the signal potentials captured by a bipolar atrial lead . with reference to fig1 the present invention can be implemented in an icd 100 ( which could be an iad ) that is implanted in the chest of a patient who suffers from unpleasant and possibly life - threatening irregularities in heart operation due to improper electrical stimulation of the heart muscle cells . during normal heart action , the heart &# 39 ; s electrical impulses originate in the sino - atrial node as an action potential that is transmitted smoothly to all portions of the atria , causing contraction of the atrial chambers . the electrical impulse continues in its path to a cluster of conduction fibrils known as the atrioventricular node . after a delay of about one - tenth of a second , an action potential flows over the ventricles and causes them to contract in synchronism with and following shortly after the atrial contraction . in this manner , the heart pumps blood to the lungs and from the lungs to the body . a bipolar lead 102 is implanted within the atria 104 of a heart 103 to measure the electrical potential between nearby cells of the atria . as the action potential actively passes from cell to cell past the two electrodes of the lead 102 , an oscillating signal potential is developed across the two electrodes and is conveyed by the bipolar lead 102 to the icd 100 where the signal is sampled about 1000 times per second and is digitized , with the sequential samples stored within a ram memory within the icd 100 for further analysis . another bipolar lead 105 may extend to the ventricle ( not shown ) of the heart 103 to measure the action potentials generated within the ventricles . these same leads , or other leads , may be used by the icd for administering various types of therapy , such as pacing or defibrillation shock therapy . the data samples collected from within the atria are now analyzed . first , with reference to data gathered from the ventricles , if some event in the ventricle , such as a mis - timed depolarization event that partially overlaps the atria &# 39 ; s p wave , could leak across to the atria and distort the measurement of the potentials for a given atrial depolarization , then the data for that particular heartbeat is discarded and is not used in further analyses . such distortion can also be found simply because a given heartbeat set of data gathered from the atria is itself badly distorted , and this approach must be taken in the case of an iad having no ventricular lead . the remaining data is retained for further processing . in preparation for further processing , the data for individual heartbeats is identified and gathered . the position within the gathered data of the negative spike to the atrial p depolarization waveform is determined and is selected as the center of the waveform for each beat . then , since data close to this p - wave is to be analyzed , and since data further away may be distorted to some degree by ventricular activity or by variations in the p - to - p interval , 32 sequential data samples are selected for each heartbeat such that the negative spike of the p - wave becomes signal potential sample number 16 of the 32 sequential samples selected for further analyses . these samples for a normal heartbeat appear at 302 in fig3 and at 502 in fig5 where the amplitude of the atrial signal potential is plotted against time over the 32 sampled values , with the negative p - wave spike positioned as signal sample number 16 . likewise , the samples for an abnormal heartbeat appear at 504 in fig5 . straightforward correlation ( cwa ) of the 32 samples against a template containing a typical or normal pattern could be performed at this juncture , as taught on page 563 ( equation ( 1 )) of the article by thorne , cited in the introductory portion of this specification . but the performance of such a cross correlation at repeated regular intervals would generate numerous computations and would drain the icd &# 39 ; s battery more rapidly than is desirable . in addition , provision would have to be made for a full template waveform that can be stored within the icd to be used as a comparison reference . in addition , such templates generally need to be patient specific , and they are more susceptible to normal changes in shape . accordingly , to reduce the mathematical complexity of the cross correlation computations , it is desirable to reduce the number of data points from 32 down to some much lower number through the use of some form of transformation into a different set of values . for example , the data could be transformed using the karhunen - loeve transformation described in the morris article cited at the beginning of this specification . but that transformation is fairly computationally intensive . computationally , a fast fourier transformation would be more efficient , but any such transformation into the frequency domain , where variable frequency sine and cosine wave amplitudes result from the transformation , does not match itself well to the morphology of atrial heart waveforms , which tend to be formed from large , ringing , spike - like representations of p - wave depolarization events travelling as moving action potential fronts past the pair of electrodes attached to the bipolar lead , and not as steady harmonics extending across time . accordingly , fourier - class transforms do not adequately reduce the number of significant data values that must be considered . and in addition , when the \u201c window \u201d size for a fourier analysis is selected , if it is wide , then the frequency values are not time - specific , but represent average signal harmonic content over the entire window time duration . and if the window is narrow , then the harmonics are too spread out in frequency , and the frequency - domain data generated by the transform does not resolve things finely enough in the frequency domain . for all of the above reasons , the present invention analyzes the waveforms using a discrete wavelet transform . this has the advantage that while low frequency wavelets are broad in time , high frequency wavelets have a much narrower timeslot focus . thus , for example , with 32 data samples taken sequentially over time , a wavelet transformation generates a d . c . wavelet that spans the entire set of 32 points ; a first reversing wavelet that also spans the entire set of points ; two second harmonic wavelets that each focus upon only one - half of the 32 data points ; four third harmonic wavelets that each focus upon only one - fourth of the 32 data points ; eight fourth harmonic wavelets that each focus upon four of the 32 data points ; and 16 fifth harmonic wavelets that each focuses upon only two of the 32 data points . accordingly , the higher - frequency wavelet amplitudes are quite time specific , unlike fourier sinusoidal amplitudes , while the lower - frequency and d . c . wavelet amplitudes give one broad information about the whole set of 32 points . and like the fourier transform , the discrete wavelet transform preserves all of the information of the original atrial signal , such that the transformation is fully reversible . 32 discrete wavelet transform values may be reverse transformed back into 32 values indicating the atrial signal potentials at 32 sequential points in time . in essence , the haar function is performing multiple digital filtering operations , at variable positions and utilizing varying - width windowing functions , to develop component values that represent varying types of heartbeat activity , at varying frequencies , spread over varying widths , and located at varying positions . these component values can be studied to see which subset of these component values are good at distinguishing st from non - st , or ( more generally ) which subset of these component values are good at distinguishing one type of heartbeat waveform from another . further study of such a selected subset can further determine which of the subset of components are relatively invariant from one individual &# 39 ; s heart to another . one preferably selects component values that are suitable from both of these two perspectives . the particular wavelet transformation to choose can be tailored to the nature of the data , with the wavelets chosen to resemble somewhat the values to be found within the data so that some transformed values are of much larger amplitude than others , and such that the low amplitude transformed values may be disregarded . of particular importance with an atrial p waveform of the type being analyzed here are wavelets having frequency values roughly comparable to and timed to coincide with the ringing of the atrial depolarization . this tailoring can minimize the number of transformed data values that are significant and that are candidates for full participation in the correlation analyses to determine if there has been a substantial change in waveform morphology . another factor in selecting a particular wavelet transformation is reducing the number of computations that must be performed to carry out the transformation . yet another factor is selecting a wavelet transformation where some of the transformed component values vary from normal to abnormal heart rhythms in much the same over a population of individuals so that a generic template of these component values may be derived that can be used with many individuals , rather than with just one individual . these factors suggest that the haar transformation would be a suitable candidate for use in analysis of atrial waveforms and possibly other waveforms as well . with reference to fig3 in the case of 32 voltage values sampled over time , the 32 applicable haar discrete transform wavelets appear as shown in this figure . other discrete wavelet transforms based upon wavelets having triangular or other shapes may also prove usable . the haar transform wavelets , in particular , have the shape of a square waveform , as will be described , and this can simplify the computations required . in fig3 time increases from left to right . a scale 304 indicates the 32 points in time at which the atrial waveform is sampled , and an analog atrial waveform ( before digitization ) is shown at 302 . the p wave negative notch 303 is shown centered at the 16 th timeslot so that the signal potential of the atria waveform sampled at this point in time becomes the 16 th data value in the set of 32 data values that are to be subjected to the haar transformation . a haar wavelet is simply one cycle of a square waveform that starts at zero , then swings positive ( to \u201c+ 1 \u201d), then swings negative ( to \u201c\u2212 1 \u201d), and then swings back to zero . the wavelet w 2 , shown at 308 in fig3 for example , is at \u201c+ 1 \u201d at points in time 1 to 16 and is at \u201c\u2212 1 \u201d at points in time 17 to 32 . the shorter waveform w 6 , shown at 316 in fig3 is at \u201c+ 1 \u201d at points in time 9 to 12 and is at \u201c\u2212 1 \u201d at points in time 13 to 16 , and is at \u201c 0 \u201d at all other points in time . the waveform w 1 , shown at 306 in fig3 is so slow to fluctuate in time that its \u201c\u2212 1 \u201d portion is off of the chart ( in fig3 ) to the right , and is ignored ; and accordingly , it has the value \u201c+ 1 \u201d for all 32 of the points in time shown in fig3 . it thus represents the average , or d . c ., component of the atrial signal potential . the lowest frequency wavelets w 1 and w 2 encompass the entire set of 32 points in time . the higher frequency wavelets w 3 and w 4 each encompasses only half of the points in time , and the two wavelets taken together encompass all the points in time . the four wavelets w 5 , w 6 , w 7 , and w 8 each encompasses only eight points in time , and the four wavelets taken together encompass all points in time . the eight wavelets w 9 . . . w 16 each encompasses only four points in time , and together all eight wavelets encompass all points in time . and finally , the wavelets w 17 . . . w 32 each encompass only two points in time , but the sixteen wavelets together encompass all points in time . thus , each wavelet has a characteristic time span and position as well as a characteristic frequency , with higher - frequency wavelets having a narrower and more specific time span and position than lower - frequency wavelets . this is advantageous with an impulse - type , ringing signal such as the atrial depolarization waveform considered here , since many higher - frequency transform values that correspond to haar wavelets positioned in time away from the p waveform or that do not correspond to its frequency of ringing may be low in amplitude such that they may safely be ignored during the analysis steps , thereby reducing the data that must be processed during the correlation steps . ( the above , while presented in the context of the haar wavelet , is also applicable to other wavelet shapes that may used to perform a discrete wavelet transform ( dwt )). each of the 32 haar transformed values is computed as follows : multiply each of the 32 sampled and digitized voltage values representing the analog atrial waveform 302 by the correspondingly - positioned -( in - time ) amplitude values of one of the 32 haar wavelets ( shown in fig3 ); then sum the resulting products ; and then multiply the resulting sums by a discrete wavelet transform scaling factor 2 \u2212 j / 2 , where j equals 1 for the wavelets w 17 to w 32 , 2 for w 9 to w 16 , 3 for w 5 to w 8 , 4 for w 3 to w 4 , and 5 for w 1 and w 2 . for example , and referring to fig3 : the first wavelet w 1 is always + 1 , so the atrial signal potential values are simply summed and then multiplied by 2 \u2212 5 / 2 ; the second wavelet is + 1 for time points 1 to 16 and \u2212 1 for time points 17 to 32 , so the first 16 values are summed , the second 16 values are summed , the difference between the first and second sums is computed , and the result is multiplied by 2 \u2212 5 / 2 ; and so on until all the transformed values have been computed , one for each haar wavelet . the 32 time - sequential atrial signal potential values are thus transformed into 32 haar wavelet amplitude values which may be called \u201c transformed values \u201d and which may be assigned the numbers 1 to 32 corresponding to the subscript numbers of the haar wavelets to which they each correspond and whose amplitudes they represent . the above description of how to compute the haar wavelet transformed values is accurate , but it is not the most efficient way to proceed in the case of haar wavelets . like the fourier transform , which has a corresponding fast fourier transform that saves intermediate results and thereby avoids re - computing them and thus reduces substantially the number of computations , the haar transformation also may be carried out in a manner that saves intermediate sums and re - uses them to reduce substantially the number of computations . this is described below at the point where fig4 is described , since fig4 illustrates graphically how this can be done . the illustrative program listing presented below is also an implementation of the fast haar wavelet transformation algorithm . the haar transformation can readily be carried out by any digital computer . as an example of how the haar transformation can be carried out on 32 values , the following program , written in the language c , is illustrative of many possible programs that may be written . in the exemplary program that follows , a 32 - element array yy contains 32 data values representing the 32 sampled atrial signal potential values representing the fluctuations over time of the signal supplied by the bipolar lead 102 . the analog atrial signal is sampled , digitized , broken into separate beat data sets , pre - processed ( to remove waveforms distorted by ventricular activity ), centered ( with the negative depolarization spike at data point 16 ), and fed into the subroutine presented below . this subroutine is compiled ( or assembled , if rewritten in assembly language ) and installed in the rom of the icd &# 39 ; s embedded microprocessor . this subroutine returns , contained within the same array yy , the 32 transformed haar wavelet amplitude values described above . ( the constant value \u201c recipsqrttwo \u201d is the reciprocal of the square root of two ). an illustrative version of the subroutine for computing all 32 of the haar transformed values in an efficient manner is presented here : void haar ( double yy []) { int i , j , l ; double zz [ 32 ]; for ( i = 5 ; i & gt ; 0 ; i \u2212\u2212) { l = 1 ; for ( j = 1 ; j & lt ;= i ; ++ j ) l = 2 * l ; for ( j = 0 ; j & lt ; l ; ++ j ) zz [ j ] = yy [ j ]; for ( j = 0 ; j & lt ; l \u2212 1 ; j = j + 2 ) ( yy [ j / 2 ] = recipsqrttwo * ( zz [ j ] + zz [ j + 1 ]); yy [( j + l )/ 2 ] = rcipsqrttwo * ( zz [ j ] \u2212 zz [ j + 1 ]); } } return ; } the above program computes the 32 haar transformed values 1 to 32 from the 32 atrial signal potential time - sequenced input values . it does so with only 31 additions , 31 subtractions , and 62 multiplications . it is carefully designed to compute each value in a systematic way , making multiple use of intermediate results . first , the program computes sixteen sums of and sixteen differences between eight adjacent pairs of the 16 time sequenced atrial signal potential values , and the sixteen difference values become the haar transformed values 17 to 32 , reflecting the strength of the highest frequency haar wavelets positioned at 16 different positions in time , corresponding to the wavelets w 17 to w 32 shown in fig3 . all the sum and difference values are scaled by multiplication by 2 \u2212 1 / 2 . next , taking the 16 sums of atrial signal potential values as an intermediate result , the above algorithm generates eight sums of and eight differences between adjacent pairs of these sixteen intermediate values that resulted from the first sixteen additions , and the eight newly - computed difference values become the haar transformed values 9 to 16 , reflecting the strength of the second to the highest frequency values at eight different points in time , corresponding to wavelets w 9 to w 16 in fig3 . all of these eight sum and eight difference values are again scaled by 2 \u2212 1 / 2 so that the wavelets w 9 to w 16 are scaled by \u00bd ( 2 \u2212 2 / 2 or 2 \u2212 1 / 2 times 2 \u2212 1 / 2 ). next , taking these eight sums of sums of adjacent atrial signal potential values as an intermediate result , the above algorithm generates four sum and four difference values , again scaling by 2 \u2212 1 / 2 , and the four difference values become haar transformed values 5 to 8 , reflecting the strength of the middle frequency wavelets at four different points in time , and corresponding to wavelets w 5 to w 8 in fig3 . next , taking these four sums of sums of sums of adjacent atrial signal potential values , the above algorithm generates two sum and two difference values , again scaling by 2 \u2212 1 / 2 , and the two difference values become haar transformed values 3 and 4 , reflecting the strength of the second to the lowest frequency wavelets only two of which encompass all the time domain data , corresponding to the wavelets w 3 and w 4 shown in fig3 . and finally , the algorithm generates the sum of and the difference between the final remaining two sums of sums of sums of sums of atrial signal potential values , again scaling by 2 \u2212 1 / 2 . the difference value is then the haar transformed value 2 , representing the strength of the lowest - frequency wavelet , the one corresponding to the wavelet w 2 in fig3 the wavelet that extends the full length of the time scale . the sum value is then the first , or d . c ., transformed value , representing the average signal potential level over the 32 sampled points in time , which corresponds to the wavelet w 1 in fig3 . another way of viewing this computation is illustrated in fig4 a and 4b . the 32 atrial signal potential values are shown at the top of fig4 a and are identified as c 0 5 through c 31 5 . the superscript \u201c 5 \u201d indicates that these values are processed when the index value \u201c n \u201d in the above computer program is equal to \u201c 5 \u201d\u2014 that is , during the first pass through the data generating intermediary sums and differences . during subsequent passes , the value of n is decremented to 4 , 3 , 2 , and finally to 1 . during each pass through the data , the computer generates sums c n - 1 and differences d n - 1 , as shown in fig4 b , between adjacent pairs of the values c 0 n and c 1 n ; c 2 n and c 3 n ; and so on , so that the number of newly - generated c n - 1 and d n - 1 terms is reduced by half with each computer pass through the intermediary results . at the end of all these computations , the value c 0 0 is the first , or d . c ., haar transformed value ; and the values d 0 0 ; d 0 1 and d 1 1 ; d 0 2 , d 1 2 , d 2 2 , and d 3 2 ; d 0 3 , d 1 3 , . . . and d 7 3 ; and d 0 4 , d 1 4 , and d 15 4 are , respectively , the remaining haar transform values 2 through 32 . fig4 thus illustrates quite succinctly how all the transform computations are carried out , and how the intermediary \u201c c \u201d sum values , such as c 0 4 , are used to compute multiple haar values , such as the values d 0 3 , d 0 2 , d 0 1 , d 0 0 , and c 0 0 all of which are computed from the intermediary value c 0 4 . saving and reusing these intermediary \u201c c \u201d values saves much computational time . fig4 b indicates the precise addition and subtraction operations that are carried out at each level to compute the values in the next lower level , proceeding down through the chart presented in fig4 a . but in any given application to heart waveform analysis , all of these computations may not be needed , and accordingly the number of computations may be reduced much further . since the present invention teaches that only a small number of these haar transformed values need actually be considered , a far less computationally intensive transform can be developed which only generates the intermediate and final transformed values that are actually needed to generate the specific haar transformed values which have proved to be significant in discriminating between sinus tachycardia ( or st ) and non - st conditions . all others need not be computed , and the above program may be reduced to a special algorithm that omits as many sums , differences , and multiplications as possible . for example , if only the transformed values 1 , 5 , 9 , and 24 are significant , then 31 additions are required to compute the first transformed value ( the d . c . value \u2014 the simple sum of all the time domain signal values ); six additions and one subtraction are required to compute the fifth transformed value ( in fig3 the difference between the sums of time domain values under each half of the wavelet w 5 at 314 in fig3 ); two additions and one subtraction are required to compute the ninth transformed value ( in fig3 the difference between the sums of the time domain values under each half of the wavelet w9 at 322 in fig3 ); and only one subtraction is required to compute the 24 th transformed value ( in fig3 the difference between the two time domain values under the respective halves of the wavelet w 24 ( not shown ) which is the same size as , but differently positioned in time than , the wavelet w 18 at 330 ( fig3 ). accordingly , if only the transformed values 1 , 5 , 9 , and 24 actually evaluated , then only 42 additions and subtractions are required . but even this number can be reduced further . if the computational algorithm set forth in the above illustrative computer program is followed as a guide , then the intermediary sums used in computing the haar first , or d . c ., transformed value can be re - used to compute the sums for the haar transformed values 5 and 9 , assuming these intermediary results are saved in the manner described in the above program example . then 31 additions are still required to compute the first transformed value , but only one subtraction is required to compute each of the transformed values 5 , 9 , and 24 , giving a total of additions and subtraction of only 34 operations . and if only transformed values 1 , 5 , and 9 are computed , the number of additions and subtractions is reduced to 33 . and if the first transformed value is omitted and if transformed values 5 , 9 , and 24 are selected , then the 31 additions needed to compute the first transformed value are not required . then the number of computations for transformed value 5 is 8 additions and one subtraction ; the number of computations for transformed value 9 is 2 additions and one subtraction ; and the number of computations for transformed value 24 is still just one subtraction . so the total number of additions and subtractions is just 13 . but even this number can be reduced to 11 when it is realized that the two additions done for the transformed value 9 are also done ( in the above computational algorithm ) when computing the transformed value 5 . so the number of additions and subtractions can be reduced to 11 . also , when computing such a small number of transformed values , the number of multiplications can be reduced as well by postponing the multiplications until a transformed value is actually computed . with reference to fig4 a , instead of dividing every sum and difference by the square root of two ( as shown in fig4 b ), several vertical sums of \u201c c \u201d values in different rows of the fig4 a table can be formed , and then a \u201c d \u201d value can be computed by subtraction , and then the resulting unscaled transformed value can be scaled with a single multiplication by 2 \u2212 j / 2 where j is the number of rows in the table of fig4 a traversed by the one or more sum and the one difference computations . thus , the number of scaling multiplications can sometimes be equal to the number of transformed values that are computed , typically 3 or 4 . accordingly , by using wavelet analysis and transformation , instead of fourier or discrete cosine or other sinusoid analysis and transformation , a highly useful and highly frequency - specific result can be achieved with a very small number of computations , thereby conserving power and battery life , and yet achieving a high degree of precision in recognizing changes in morphology . in our tests , we have selected certain transformed values as being much more significant than other values in distinguishing between st and non - st . we focused upon those transform values whose \u201c+ 1 \u201d and \u201c\u2212 1 \u201d scope included the middle 8 time points most significant to analysis of the peak of the atrial depolarization . working across test data samples obtained from patients who had dual chamber icds , we computed the variance of each transformed value for st and non - st events using the standard statistical formula for computing variance . we then calculated the ratio of the variance of non - st events to the variance of st events , and selected those terms with the highest variance ratios for further consideration . in this manner , we reduced the number of transformed values that were included in the cwa or correlation process while always maintaining or improving the performance achieved . ultimately we settled upon the three or four transformed values having the highest ratios . these were the transform values 1 , 5 , 9 , and 24 . we tested 1 , 5 , and 24 together ; 5 , 9 , and 24 together ; and 1 , 5 , 9 , and 24 together . these selected haar transform values , when used in these combinations , required minimal computations and thereby produced a savings in battery power , and that also gave better performance at distinguishing st from non - st than did all of the transformed values used together . so an increase in accuracy was achieved as well as a decrease in computational complexity by this approach to atrial waveform analysis . [ 0054 ] fig5 for example , illustrates actual plots of digital information illustrating the effect of using only the three coefficients 1 , 5 , and 24 . at 502 , a normal waveform is shown , with amplitude plotted against sample numbers from 1 to 32 . at 504 , an abnormal waveform is shown , again with amplitude plotted against sample numbers . after the haar transformation , at 506 and 508 , the wavelet coefficient amplitude values are indicated for the coefficients 1 to 32 . at 506 , the coefficients generated by the normal waveform 502 are shown , and at 508 , the coefficients generated by the abnormal waveform 504 are shown . one may directly compare these component values and verify that the coefficients 1 , 5 , and 24 at 506 and at 508 vary significantly in amplitude . comparison of this same information among different patients ( not shown in this figure ) also indicated that this variation is relatively constant from one patient to another . at 510 and 512 , using only the coefficients 1 , 5 , and 24 , the heartbeat waveforms are reconstructed by a reverse transformation , the normal reconstructed waveform shown at 510 and the abnormal reconstructed waveform shown at 512 . the marked differences between these two reconstructed waveforms highlights the way in which these three coefficients can signal an abnormal condition with less computation . in a practical system , a small number of transform values are selected , in the manner just described . the algorithm for computing these values is then refined , as explained above , to reduce the number of additions , subtractions , and multiplications to the minimum possible while preserving the accuracy of the computations . we first compute a value \u03c1 , which is the correlation between these values with the corresponding values in a template that represents the values for an average normal population . we then compare this value \u03c1 to a threshold correlation value \u03b2 that is chosen to give optimal results on experimental data . ( see the examples in the tables presented below .) for each waveform analyzed , a test is made to determine whether \u03c1 is greater than \u03b2 . if so , then this waveform is placed into the buffer marked \u201c st \u201d at step 210 . if not , then this waveform is placed into the buffer marked \u201c non - st \u201d at step 211 . finally , after ten waveforms have been analyzed , a count of st and non - st marked waveforms is made to see if the count of non - st waveforms is greater than some threshold value x ( at step 212 ), where x can be , for example , 7 . if more than seven waveforms are non - st , then therapy is delivered at step 214 . the buffers at steps 210 and 211 in fig2 are buffers that hold the last ten decision values . these buffers can be pictured as sliding windows revealing the most recent ten values to permit the decision at 212 to be continuously updated . the buffers thus function as a nonlinear filter preventing irregular values from delivering therapy improperly . in developing a working prototype of this system , we used a development data set consisting of 20 episodes of st taken from 4 patients and 18 episodes of af taken from 7 patients . patient number episodes of st episodes of af 1 2 6 2 12 0 13 0 1 4 0 2 5 5 1 6 0 3 7 0 3 8 0 2 9 1 0 total 20 18 we used this data to optimize our choice of \u03c1 and \u03b2 and also the particular coefficients that we chose to examine . next , we tested our prototype using a set of 16 episodes of st obtained from 4 patients together with 17 episodes of af obtained from 7 patients . patient number episodes of st episodes of af 1 8 0 2 3 1 3 3 0 4 2 0 5 0 4 6 0 4 7 0 3 8 0 3 9 0 1 10 0 1 total 16 17 x out threshold coefficients of 10 ( \u03b2 ) sensitivity specificity 1 , 5 , 9 , & amp ; 24 3 0 . 808 76 % 94 % 1 , 5 , 9 , & amp ; 24 5 0 . 975 88 % 94 % 1 , 5 , 9 , & amp ; 24 6 0 . 976 82 % 94 % 1 , 5 , 9 , & amp ; 24 7 0 . 983 88 % 94 % 1 , 5 , & amp ; 24 3 0 . 815 82 % 94 % 1 , 5 , & amp ; 24 4 0 . 943 88 % 94 % 1 , 5 , & amp ; 24 5 0 . 956 82 % 94 % 1 , 5 , & amp ; 24 7 0 . 991 94 % 94 % 5 , 9 , & amp ; 24 8 0 . 9993 82 % 94 % 5 , 9 , & amp ; 24 9 0 . 9997 76 % 81 % the quality of the selected subset of transformed values for characterizing changes in morphology may be demonstrated graphically , as indicated in fig5 by performing a reverse haar transform using only the three or four or so selected transformed values . as can be seen ( at 510 and 512 in fig5 ), the inverse transformations produce distorted representations of the original atrial waveforms which illustrate how the choice of transform values can emphasize the changes . this is another useful way to assist one in selecting which transformed values are most useful . one feature of the invention is its ability to use a template derived from an average normal population , rather than deriving a patient customized average normal template for each individual patient . prior systems have required each new patient to be monitored while in a normal state , and the captured data was then averaged to form a patient specific normal template . contrary to this , the present invention achieved the results shown above using the same template for all patients . accordingly , the invention may be used with a new patient without the necessity of such preliminary testing of each patient and without a new customized template necessarily having to be created for each patient . one reason why the present invention can function using a template that represents average values for a normal population is because the specific coefficients selected in the transform domain , in addition to having been selected to emphasize the difference between normal and abnormal values , are also preferably chosen to minimize the differences between different patients . in some cases , values that were good at distinguishing between normal and abnormal rhythms for one patient were not as good at doing so for some other patient . these values are preferably not selected for use in implementing the present invention . for example , the transform coefficients may be selected such that all ( or most of ) the normal values of a particular coefficient gathered from many patients had the same sign ( positive or negative ) and similar amplitudes ( or if they varied in sign , they were near to zero ). in addition , the abnormal coefficients are significantly different in value from the normal coefficients for each particular patient . [ 0068 ] fig6 at 600 , illustrates one way in which this may be done . first , at step 602 , one gathers a large number of st and non - st data samples . next , all of this data is transformed , generating coefficients for every waveform set of data ( step 604 ). the variance of each coefficient is then computed first for the st samples ( step 606 ) and then for the non - st samples . then , for each coefficient , the ratio of the variance of the non - st samples to that of the st samples is computed ( step 608 ). finally , those coefficients whose variance ratio was large may be selected a coefficients for use in creating a population template and in testing patients ( step 612 ). in addition , at step 614 , the number of coefficients retained may be further reduced to reduce the number of computations that need to be performed , as explained above . in the version of the invention that was used to generate the data shown above , the population template was computed as follows , using the 20 st episodes in the development data set ( also used in the first of the two tables presented above ). this is illustrated at 700 in fig7 . first , the coefficients were selected as was explained above and as shown in fig6 . next , for each patient episode of st ( step 702 ), the recorded heartbeats were broken up into groups of ten consecutive heartbeats ( step 704 ). for each group ( step 706 ), the selected coefficients of the haar transform were computed ( step 708 ) for each of the ten heartbeats and the median value was then selected ( step 710 ) for each coefficient to form a proto - template . if these median values correlated well with the coefficient values for each of the ten heartbeats , the proto - template was retained ( step 714 ). otherwise , it was discarded . in this manner , a whole bunch of proto - templates were obtained from each patient episode . next , median values of the coefficients from all of the proto - templates ( step 716 ) were computed ( step 718 ). these values were then used as the population template for distinguishing st from non - st events ( step 720 ). while the preferred embodiment of the invention has been described above , it is to be understood that numerous modifications and changes will occur to those who are skilled in the art to which the invention pertains . accordingly , the following claims annexed to and forming a part of this specification are intended to define the true spirit and scope of the invention \u2014 that is , what is new and what is desired to be secured by letters patent of the united states .", "category": "Physics"}
|
Is the category the most suitable category for the given patent?
| 0.25 |
1cd5305994f1692ec7ec09368a2a725ff3c3dd15c21bf56c12e04e558cdb5fbc
| 0.037842 | 0.033203 | 0.046143 | 0.212891 | 0.255859 | 0.675781 |
null |
{"category": "Human Necessities", "patent": "the present invention is directed towards improving the quality and functionality of both implantable cardioverter defibrillators ( icds ) and implantable atrial defibrillators ( iads ). all such devices need to have a method and mechanism for accurately distinguishing sinus tachycardia ( st ) form non - st , such as atrial fibrillation and atrial flutter . this method and mechanism should be conservative , biased towards not initiating therapy unless non - st is clearly indicated . it should also require as few machine computational cycles as possible to reduce icd and iad battery drain . the principle underlying the present invention is that st and non - st can be distinguished by studying and measuring the atrial electrogram morphology to detect aberrant conduction in the atria . thus , by comparing normal st waveforms to a series of newly - occurring waveforms , st is indicated by abnormal atrial waveform morphology which signals aberrant conduction within the atria improper electrical patterns of depolarization . this will show up in the signal potentials captured by a bipolar atrial lead . with reference to fig1 the present invention can be implemented in an icd 100 ( which could be an iad ) that is implanted in the chest of a patient who suffers from unpleasant and possibly life - threatening irregularities in heart operation due to improper electrical stimulation of the heart muscle cells . during normal heart action , the heart &# 39 ; s electrical impulses originate in the sino - atrial node as an action potential that is transmitted smoothly to all portions of the atria , causing contraction of the atrial chambers . the electrical impulse continues in its path to a cluster of conduction fibrils known as the atrioventricular node . after a delay of about one - tenth of a second , an action potential flows over the ventricles and causes them to contract in synchronism with and following shortly after the atrial contraction . in this manner , the heart pumps blood to the lungs and from the lungs to the body . a bipolar lead 102 is implanted within the atria 104 of a heart 103 to measure the electrical potential between nearby cells of the atria . as the action potential actively passes from cell to cell past the two electrodes of the lead 102 , an oscillating signal potential is developed across the two electrodes and is conveyed by the bipolar lead 102 to the icd 100 where the signal is sampled about 1000 times per second and is digitized , with the sequential samples stored within a ram memory within the icd 100 for further analysis . another bipolar lead 105 may extend to the ventricle ( not shown ) of the heart 103 to measure the action potentials generated within the ventricles . these same leads , or other leads , may be used by the icd for administering various types of therapy , such as pacing or defibrillation shock therapy . the data samples collected from within the atria are now analyzed . first , with reference to data gathered from the ventricles , if some event in the ventricle , such as a mis - timed depolarization event that partially overlaps the atria &# 39 ; s p wave , could leak across to the atria and distort the measurement of the potentials for a given atrial depolarization , then the data for that particular heartbeat is discarded and is not used in further analyses . such distortion can also be found simply because a given heartbeat set of data gathered from the atria is itself badly distorted , and this approach must be taken in the case of an iad having no ventricular lead . the remaining data is retained for further processing . in preparation for further processing , the data for individual heartbeats is identified and gathered . the position within the gathered data of the negative spike to the atrial p depolarization waveform is determined and is selected as the center of the waveform for each beat . then , since data close to this p - wave is to be analyzed , and since data further away may be distorted to some degree by ventricular activity or by variations in the p - to - p interval , 32 sequential data samples are selected for each heartbeat such that the negative spike of the p - wave becomes signal potential sample number 16 of the 32 sequential samples selected for further analyses . these samples for a normal heartbeat appear at 302 in fig3 and at 502 in fig5 where the amplitude of the atrial signal potential is plotted against time over the 32 sampled values , with the negative p - wave spike positioned as signal sample number 16 . likewise , the samples for an abnormal heartbeat appear at 504 in fig5 . straightforward correlation ( cwa ) of the 32 samples against a template containing a typical or normal pattern could be performed at this juncture , as taught on page 563 ( equation ( 1 )) of the article by thorne , cited in the introductory portion of this specification . but the performance of such a cross correlation at repeated regular intervals would generate numerous computations and would drain the icd &# 39 ; s battery more rapidly than is desirable . in addition , provision would have to be made for a full template waveform that can be stored within the icd to be used as a comparison reference . in addition , such templates generally need to be patient specific , and they are more susceptible to normal changes in shape . accordingly , to reduce the mathematical complexity of the cross correlation computations , it is desirable to reduce the number of data points from 32 down to some much lower number through the use of some form of transformation into a different set of values . for example , the data could be transformed using the karhunen - loeve transformation described in the morris article cited at the beginning of this specification . but that transformation is fairly computationally intensive . computationally , a fast fourier transformation would be more efficient , but any such transformation into the frequency domain , where variable frequency sine and cosine wave amplitudes result from the transformation , does not match itself well to the morphology of atrial heart waveforms , which tend to be formed from large , ringing , spike - like representations of p - wave depolarization events travelling as moving action potential fronts past the pair of electrodes attached to the bipolar lead , and not as steady harmonics extending across time . accordingly , fourier - class transforms do not adequately reduce the number of significant data values that must be considered . and in addition , when the \u201c window \u201d size for a fourier analysis is selected , if it is wide , then the frequency values are not time - specific , but represent average signal harmonic content over the entire window time duration . and if the window is narrow , then the harmonics are too spread out in frequency , and the frequency - domain data generated by the transform does not resolve things finely enough in the frequency domain . for all of the above reasons , the present invention analyzes the waveforms using a discrete wavelet transform . this has the advantage that while low frequency wavelets are broad in time , high frequency wavelets have a much narrower timeslot focus . thus , for example , with 32 data samples taken sequentially over time , a wavelet transformation generates a d . c . wavelet that spans the entire set of 32 points ; a first reversing wavelet that also spans the entire set of points ; two second harmonic wavelets that each focus upon only one - half of the 32 data points ; four third harmonic wavelets that each focus upon only one - fourth of the 32 data points ; eight fourth harmonic wavelets that each focus upon four of the 32 data points ; and 16 fifth harmonic wavelets that each focuses upon only two of the 32 data points . accordingly , the higher - frequency wavelet amplitudes are quite time specific , unlike fourier sinusoidal amplitudes , while the lower - frequency and d . c . wavelet amplitudes give one broad information about the whole set of 32 points . and like the fourier transform , the discrete wavelet transform preserves all of the information of the original atrial signal , such that the transformation is fully reversible . 32 discrete wavelet transform values may be reverse transformed back into 32 values indicating the atrial signal potentials at 32 sequential points in time . in essence , the haar function is performing multiple digital filtering operations , at variable positions and utilizing varying - width windowing functions , to develop component values that represent varying types of heartbeat activity , at varying frequencies , spread over varying widths , and located at varying positions . these component values can be studied to see which subset of these component values are good at distinguishing st from non - st , or ( more generally ) which subset of these component values are good at distinguishing one type of heartbeat waveform from another . further study of such a selected subset can further determine which of the subset of components are relatively invariant from one individual &# 39 ; s heart to another . one preferably selects component values that are suitable from both of these two perspectives . the particular wavelet transformation to choose can be tailored to the nature of the data , with the wavelets chosen to resemble somewhat the values to be found within the data so that some transformed values are of much larger amplitude than others , and such that the low amplitude transformed values may be disregarded . of particular importance with an atrial p waveform of the type being analyzed here are wavelets having frequency values roughly comparable to and timed to coincide with the ringing of the atrial depolarization . this tailoring can minimize the number of transformed data values that are significant and that are candidates for full participation in the correlation analyses to determine if there has been a substantial change in waveform morphology . another factor in selecting a particular wavelet transformation is reducing the number of computations that must be performed to carry out the transformation . yet another factor is selecting a wavelet transformation where some of the transformed component values vary from normal to abnormal heart rhythms in much the same over a population of individuals so that a generic template of these component values may be derived that can be used with many individuals , rather than with just one individual . these factors suggest that the haar transformation would be a suitable candidate for use in analysis of atrial waveforms and possibly other waveforms as well . with reference to fig3 in the case of 32 voltage values sampled over time , the 32 applicable haar discrete transform wavelets appear as shown in this figure . other discrete wavelet transforms based upon wavelets having triangular or other shapes may also prove usable . the haar transform wavelets , in particular , have the shape of a square waveform , as will be described , and this can simplify the computations required . in fig3 time increases from left to right . a scale 304 indicates the 32 points in time at which the atrial waveform is sampled , and an analog atrial waveform ( before digitization ) is shown at 302 . the p wave negative notch 303 is shown centered at the 16 th timeslot so that the signal potential of the atria waveform sampled at this point in time becomes the 16 th data value in the set of 32 data values that are to be subjected to the haar transformation . a haar wavelet is simply one cycle of a square waveform that starts at zero , then swings positive ( to \u201c+ 1 \u201d), then swings negative ( to \u201c\u2212 1 \u201d), and then swings back to zero . the wavelet w 2 , shown at 308 in fig3 for example , is at \u201c+ 1 \u201d at points in time 1 to 16 and is at \u201c\u2212 1 \u201d at points in time 17 to 32 . the shorter waveform w 6 , shown at 316 in fig3 is at \u201c+ 1 \u201d at points in time 9 to 12 and is at \u201c\u2212 1 \u201d at points in time 13 to 16 , and is at \u201c 0 \u201d at all other points in time . the waveform w 1 , shown at 306 in fig3 is so slow to fluctuate in time that its \u201c\u2212 1 \u201d portion is off of the chart ( in fig3 ) to the right , and is ignored ; and accordingly , it has the value \u201c+ 1 \u201d for all 32 of the points in time shown in fig3 . it thus represents the average , or d . c ., component of the atrial signal potential . the lowest frequency wavelets w 1 and w 2 encompass the entire set of 32 points in time . the higher frequency wavelets w 3 and w 4 each encompasses only half of the points in time , and the two wavelets taken together encompass all the points in time . the four wavelets w 5 , w 6 , w 7 , and w 8 each encompasses only eight points in time , and the four wavelets taken together encompass all points in time . the eight wavelets w 9 . . . w 16 each encompasses only four points in time , and together all eight wavelets encompass all points in time . and finally , the wavelets w 17 . . . w 32 each encompass only two points in time , but the sixteen wavelets together encompass all points in time . thus , each wavelet has a characteristic time span and position as well as a characteristic frequency , with higher - frequency wavelets having a narrower and more specific time span and position than lower - frequency wavelets . this is advantageous with an impulse - type , ringing signal such as the atrial depolarization waveform considered here , since many higher - frequency transform values that correspond to haar wavelets positioned in time away from the p waveform or that do not correspond to its frequency of ringing may be low in amplitude such that they may safely be ignored during the analysis steps , thereby reducing the data that must be processed during the correlation steps . ( the above , while presented in the context of the haar wavelet , is also applicable to other wavelet shapes that may used to perform a discrete wavelet transform ( dwt )). each of the 32 haar transformed values is computed as follows : multiply each of the 32 sampled and digitized voltage values representing the analog atrial waveform 302 by the correspondingly - positioned -( in - time ) amplitude values of one of the 32 haar wavelets ( shown in fig3 ); then sum the resulting products ; and then multiply the resulting sums by a discrete wavelet transform scaling factor 2 \u2212 j / 2 , where j equals 1 for the wavelets w 17 to w 32 , 2 for w 9 to w 16 , 3 for w 5 to w 8 , 4 for w 3 to w 4 , and 5 for w 1 and w 2 . for example , and referring to fig3 : the first wavelet w 1 is always + 1 , so the atrial signal potential values are simply summed and then multiplied by 2 \u2212 5 / 2 ; the second wavelet is + 1 for time points 1 to 16 and \u2212 1 for time points 17 to 32 , so the first 16 values are summed , the second 16 values are summed , the difference between the first and second sums is computed , and the result is multiplied by 2 \u2212 5 / 2 ; and so on until all the transformed values have been computed , one for each haar wavelet . the 32 time - sequential atrial signal potential values are thus transformed into 32 haar wavelet amplitude values which may be called \u201c transformed values \u201d and which may be assigned the numbers 1 to 32 corresponding to the subscript numbers of the haar wavelets to which they each correspond and whose amplitudes they represent . the above description of how to compute the haar wavelet transformed values is accurate , but it is not the most efficient way to proceed in the case of haar wavelets . like the fourier transform , which has a corresponding fast fourier transform that saves intermediate results and thereby avoids re - computing them and thus reduces substantially the number of computations , the haar transformation also may be carried out in a manner that saves intermediate sums and re - uses them to reduce substantially the number of computations . this is described below at the point where fig4 is described , since fig4 illustrates graphically how this can be done . the illustrative program listing presented below is also an implementation of the fast haar wavelet transformation algorithm . the haar transformation can readily be carried out by any digital computer . as an example of how the haar transformation can be carried out on 32 values , the following program , written in the language c , is illustrative of many possible programs that may be written . in the exemplary program that follows , a 32 - element array yy contains 32 data values representing the 32 sampled atrial signal potential values representing the fluctuations over time of the signal supplied by the bipolar lead 102 . the analog atrial signal is sampled , digitized , broken into separate beat data sets , pre - processed ( to remove waveforms distorted by ventricular activity ), centered ( with the negative depolarization spike at data point 16 ), and fed into the subroutine presented below . this subroutine is compiled ( or assembled , if rewritten in assembly language ) and installed in the rom of the icd &# 39 ; s embedded microprocessor . this subroutine returns , contained within the same array yy , the 32 transformed haar wavelet amplitude values described above . ( the constant value \u201c recipsqrttwo \u201d is the reciprocal of the square root of two ). an illustrative version of the subroutine for computing all 32 of the haar transformed values in an efficient manner is presented here : void haar ( double yy []) { int i , j , l ; double zz [ 32 ]; for ( i = 5 ; i & gt ; 0 ; i \u2212\u2212) { l = 1 ; for ( j = 1 ; j & lt ;= i ; ++ j ) l = 2 * l ; for ( j = 0 ; j & lt ; l ; ++ j ) zz [ j ] = yy [ j ]; for ( j = 0 ; j & lt ; l \u2212 1 ; j = j + 2 ) ( yy [ j / 2 ] = recipsqrttwo * ( zz [ j ] + zz [ j + 1 ]); yy [( j + l )/ 2 ] = rcipsqrttwo * ( zz [ j ] \u2212 zz [ j + 1 ]); } } return ; } the above program computes the 32 haar transformed values 1 to 32 from the 32 atrial signal potential time - sequenced input values . it does so with only 31 additions , 31 subtractions , and 62 multiplications . it is carefully designed to compute each value in a systematic way , making multiple use of intermediate results . first , the program computes sixteen sums of and sixteen differences between eight adjacent pairs of the 16 time sequenced atrial signal potential values , and the sixteen difference values become the haar transformed values 17 to 32 , reflecting the strength of the highest frequency haar wavelets positioned at 16 different positions in time , corresponding to the wavelets w 17 to w 32 shown in fig3 . all the sum and difference values are scaled by multiplication by 2 \u2212 1 / 2 . next , taking the 16 sums of atrial signal potential values as an intermediate result , the above algorithm generates eight sums of and eight differences between adjacent pairs of these sixteen intermediate values that resulted from the first sixteen additions , and the eight newly - computed difference values become the haar transformed values 9 to 16 , reflecting the strength of the second to the highest frequency values at eight different points in time , corresponding to wavelets w 9 to w 16 in fig3 . all of these eight sum and eight difference values are again scaled by 2 \u2212 1 / 2 so that the wavelets w 9 to w 16 are scaled by \u00bd ( 2 \u2212 2 / 2 or 2 \u2212 1 / 2 times 2 \u2212 1 / 2 ). next , taking these eight sums of sums of adjacent atrial signal potential values as an intermediate result , the above algorithm generates four sum and four difference values , again scaling by 2 \u2212 1 / 2 , and the four difference values become haar transformed values 5 to 8 , reflecting the strength of the middle frequency wavelets at four different points in time , and corresponding to wavelets w 5 to w 8 in fig3 . next , taking these four sums of sums of sums of adjacent atrial signal potential values , the above algorithm generates two sum and two difference values , again scaling by 2 \u2212 1 / 2 , and the two difference values become haar transformed values 3 and 4 , reflecting the strength of the second to the lowest frequency wavelets only two of which encompass all the time domain data , corresponding to the wavelets w 3 and w 4 shown in fig3 . and finally , the algorithm generates the sum of and the difference between the final remaining two sums of sums of sums of sums of atrial signal potential values , again scaling by 2 \u2212 1 / 2 . the difference value is then the haar transformed value 2 , representing the strength of the lowest - frequency wavelet , the one corresponding to the wavelet w 2 in fig3 the wavelet that extends the full length of the time scale . the sum value is then the first , or d . c ., transformed value , representing the average signal potential level over the 32 sampled points in time , which corresponds to the wavelet w 1 in fig3 . another way of viewing this computation is illustrated in fig4 a and 4b . the 32 atrial signal potential values are shown at the top of fig4 a and are identified as c 0 5 through c 31 5 . the superscript \u201c 5 \u201d indicates that these values are processed when the index value \u201c n \u201d in the above computer program is equal to \u201c 5 \u201d\u2014 that is , during the first pass through the data generating intermediary sums and differences . during subsequent passes , the value of n is decremented to 4 , 3 , 2 , and finally to 1 . during each pass through the data , the computer generates sums c n - 1 and differences d n - 1 , as shown in fig4 b , between adjacent pairs of the values c 0 n and c 1 n ; c 2 n and c 3 n ; and so on , so that the number of newly - generated c n - 1 and d n - 1 terms is reduced by half with each computer pass through the intermediary results . at the end of all these computations , the value c 0 0 is the first , or d . c ., haar transformed value ; and the values d 0 0 ; d 0 1 and d 1 1 ; d 0 2 , d 1 2 , d 2 2 , and d 3 2 ; d 0 3 , d 1 3 , . . . and d 7 3 ; and d 0 4 , d 1 4 , and d 15 4 are , respectively , the remaining haar transform values 2 through 32 . fig4 thus illustrates quite succinctly how all the transform computations are carried out , and how the intermediary \u201c c \u201d sum values , such as c 0 4 , are used to compute multiple haar values , such as the values d 0 3 , d 0 2 , d 0 1 , d 0 0 , and c 0 0 all of which are computed from the intermediary value c 0 4 . saving and reusing these intermediary \u201c c \u201d values saves much computational time . fig4 b indicates the precise addition and subtraction operations that are carried out at each level to compute the values in the next lower level , proceeding down through the chart presented in fig4 a . but in any given application to heart waveform analysis , all of these computations may not be needed , and accordingly the number of computations may be reduced much further . since the present invention teaches that only a small number of these haar transformed values need actually be considered , a far less computationally intensive transform can be developed which only generates the intermediate and final transformed values that are actually needed to generate the specific haar transformed values which have proved to be significant in discriminating between sinus tachycardia ( or st ) and non - st conditions . all others need not be computed , and the above program may be reduced to a special algorithm that omits as many sums , differences , and multiplications as possible . for example , if only the transformed values 1 , 5 , 9 , and 24 are significant , then 31 additions are required to compute the first transformed value ( the d . c . value \u2014 the simple sum of all the time domain signal values ); six additions and one subtraction are required to compute the fifth transformed value ( in fig3 the difference between the sums of time domain values under each half of the wavelet w 5 at 314 in fig3 ); two additions and one subtraction are required to compute the ninth transformed value ( in fig3 the difference between the sums of the time domain values under each half of the wavelet w9 at 322 in fig3 ); and only one subtraction is required to compute the 24 th transformed value ( in fig3 the difference between the two time domain values under the respective halves of the wavelet w 24 ( not shown ) which is the same size as , but differently positioned in time than , the wavelet w 18 at 330 ( fig3 ). accordingly , if only the transformed values 1 , 5 , 9 , and 24 actually evaluated , then only 42 additions and subtractions are required . but even this number can be reduced further . if the computational algorithm set forth in the above illustrative computer program is followed as a guide , then the intermediary sums used in computing the haar first , or d . c ., transformed value can be re - used to compute the sums for the haar transformed values 5 and 9 , assuming these intermediary results are saved in the manner described in the above program example . then 31 additions are still required to compute the first transformed value , but only one subtraction is required to compute each of the transformed values 5 , 9 , and 24 , giving a total of additions and subtraction of only 34 operations . and if only transformed values 1 , 5 , and 9 are computed , the number of additions and subtractions is reduced to 33 . and if the first transformed value is omitted and if transformed values 5 , 9 , and 24 are selected , then the 31 additions needed to compute the first transformed value are not required . then the number of computations for transformed value 5 is 8 additions and one subtraction ; the number of computations for transformed value 9 is 2 additions and one subtraction ; and the number of computations for transformed value 24 is still just one subtraction . so the total number of additions and subtractions is just 13 . but even this number can be reduced to 11 when it is realized that the two additions done for the transformed value 9 are also done ( in the above computational algorithm ) when computing the transformed value 5 . so the number of additions and subtractions can be reduced to 11 . also , when computing such a small number of transformed values , the number of multiplications can be reduced as well by postponing the multiplications until a transformed value is actually computed . with reference to fig4 a , instead of dividing every sum and difference by the square root of two ( as shown in fig4 b ), several vertical sums of \u201c c \u201d values in different rows of the fig4 a table can be formed , and then a \u201c d \u201d value can be computed by subtraction , and then the resulting unscaled transformed value can be scaled with a single multiplication by 2 \u2212 j / 2 where j is the number of rows in the table of fig4 a traversed by the one or more sum and the one difference computations . thus , the number of scaling multiplications can sometimes be equal to the number of transformed values that are computed , typically 3 or 4 . accordingly , by using wavelet analysis and transformation , instead of fourier or discrete cosine or other sinusoid analysis and transformation , a highly useful and highly frequency - specific result can be achieved with a very small number of computations , thereby conserving power and battery life , and yet achieving a high degree of precision in recognizing changes in morphology . in our tests , we have selected certain transformed values as being much more significant than other values in distinguishing between st and non - st . we focused upon those transform values whose \u201c+ 1 \u201d and \u201c\u2212 1 \u201d scope included the middle 8 time points most significant to analysis of the peak of the atrial depolarization . working across test data samples obtained from patients who had dual chamber icds , we computed the variance of each transformed value for st and non - st events using the standard statistical formula for computing variance . we then calculated the ratio of the variance of non - st events to the variance of st events , and selected those terms with the highest variance ratios for further consideration . in this manner , we reduced the number of transformed values that were included in the cwa or correlation process while always maintaining or improving the performance achieved . ultimately we settled upon the three or four transformed values having the highest ratios . these were the transform values 1 , 5 , 9 , and 24 . we tested 1 , 5 , and 24 together ; 5 , 9 , and 24 together ; and 1 , 5 , 9 , and 24 together . these selected haar transform values , when used in these combinations , required minimal computations and thereby produced a savings in battery power , and that also gave better performance at distinguishing st from non - st than did all of the transformed values used together . so an increase in accuracy was achieved as well as a decrease in computational complexity by this approach to atrial waveform analysis . [ 0054 ] fig5 for example , illustrates actual plots of digital information illustrating the effect of using only the three coefficients 1 , 5 , and 24 . at 502 , a normal waveform is shown , with amplitude plotted against sample numbers from 1 to 32 . at 504 , an abnormal waveform is shown , again with amplitude plotted against sample numbers . after the haar transformation , at 506 and 508 , the wavelet coefficient amplitude values are indicated for the coefficients 1 to 32 . at 506 , the coefficients generated by the normal waveform 502 are shown , and at 508 , the coefficients generated by the abnormal waveform 504 are shown . one may directly compare these component values and verify that the coefficients 1 , 5 , and 24 at 506 and at 508 vary significantly in amplitude . comparison of this same information among different patients ( not shown in this figure ) also indicated that this variation is relatively constant from one patient to another . at 510 and 512 , using only the coefficients 1 , 5 , and 24 , the heartbeat waveforms are reconstructed by a reverse transformation , the normal reconstructed waveform shown at 510 and the abnormal reconstructed waveform shown at 512 . the marked differences between these two reconstructed waveforms highlights the way in which these three coefficients can signal an abnormal condition with less computation . in a practical system , a small number of transform values are selected , in the manner just described . the algorithm for computing these values is then refined , as explained above , to reduce the number of additions , subtractions , and multiplications to the minimum possible while preserving the accuracy of the computations . we first compute a value \u03c1 , which is the correlation between these values with the corresponding values in a template that represents the values for an average normal population . we then compare this value \u03c1 to a threshold correlation value \u03b2 that is chosen to give optimal results on experimental data . ( see the examples in the tables presented below .) for each waveform analyzed , a test is made to determine whether \u03c1 is greater than \u03b2 . if so , then this waveform is placed into the buffer marked \u201c st \u201d at step 210 . if not , then this waveform is placed into the buffer marked \u201c non - st \u201d at step 211 . finally , after ten waveforms have been analyzed , a count of st and non - st marked waveforms is made to see if the count of non - st waveforms is greater than some threshold value x ( at step 212 ), where x can be , for example , 7 . if more than seven waveforms are non - st , then therapy is delivered at step 214 . the buffers at steps 210 and 211 in fig2 are buffers that hold the last ten decision values . these buffers can be pictured as sliding windows revealing the most recent ten values to permit the decision at 212 to be continuously updated . the buffers thus function as a nonlinear filter preventing irregular values from delivering therapy improperly . in developing a working prototype of this system , we used a development data set consisting of 20 episodes of st taken from 4 patients and 18 episodes of af taken from 7 patients . patient number episodes of st episodes of af 1 2 6 2 12 0 13 0 1 4 0 2 5 5 1 6 0 3 7 0 3 8 0 2 9 1 0 total 20 18 we used this data to optimize our choice of \u03c1 and \u03b2 and also the particular coefficients that we chose to examine . next , we tested our prototype using a set of 16 episodes of st obtained from 4 patients together with 17 episodes of af obtained from 7 patients . patient number episodes of st episodes of af 1 8 0 2 3 1 3 3 0 4 2 0 5 0 4 6 0 4 7 0 3 8 0 3 9 0 1 10 0 1 total 16 17 x out threshold coefficients of 10 ( \u03b2 ) sensitivity specificity 1 , 5 , 9 , & amp ; 24 3 0 . 808 76 % 94 % 1 , 5 , 9 , & amp ; 24 5 0 . 975 88 % 94 % 1 , 5 , 9 , & amp ; 24 6 0 . 976 82 % 94 % 1 , 5 , 9 , & amp ; 24 7 0 . 983 88 % 94 % 1 , 5 , & amp ; 24 3 0 . 815 82 % 94 % 1 , 5 , & amp ; 24 4 0 . 943 88 % 94 % 1 , 5 , & amp ; 24 5 0 . 956 82 % 94 % 1 , 5 , & amp ; 24 7 0 . 991 94 % 94 % 5 , 9 , & amp ; 24 8 0 . 9993 82 % 94 % 5 , 9 , & amp ; 24 9 0 . 9997 76 % 81 % the quality of the selected subset of transformed values for characterizing changes in morphology may be demonstrated graphically , as indicated in fig5 by performing a reverse haar transform using only the three or four or so selected transformed values . as can be seen ( at 510 and 512 in fig5 ), the inverse transformations produce distorted representations of the original atrial waveforms which illustrate how the choice of transform values can emphasize the changes . this is another useful way to assist one in selecting which transformed values are most useful . one feature of the invention is its ability to use a template derived from an average normal population , rather than deriving a patient customized average normal template for each individual patient . prior systems have required each new patient to be monitored while in a normal state , and the captured data was then averaged to form a patient specific normal template . contrary to this , the present invention achieved the results shown above using the same template for all patients . accordingly , the invention may be used with a new patient without the necessity of such preliminary testing of each patient and without a new customized template necessarily having to be created for each patient . one reason why the present invention can function using a template that represents average values for a normal population is because the specific coefficients selected in the transform domain , in addition to having been selected to emphasize the difference between normal and abnormal values , are also preferably chosen to minimize the differences between different patients . in some cases , values that were good at distinguishing between normal and abnormal rhythms for one patient were not as good at doing so for some other patient . these values are preferably not selected for use in implementing the present invention . for example , the transform coefficients may be selected such that all ( or most of ) the normal values of a particular coefficient gathered from many patients had the same sign ( positive or negative ) and similar amplitudes ( or if they varied in sign , they were near to zero ). in addition , the abnormal coefficients are significantly different in value from the normal coefficients for each particular patient . [ 0068 ] fig6 at 600 , illustrates one way in which this may be done . first , at step 602 , one gathers a large number of st and non - st data samples . next , all of this data is transformed , generating coefficients for every waveform set of data ( step 604 ). the variance of each coefficient is then computed first for the st samples ( step 606 ) and then for the non - st samples . then , for each coefficient , the ratio of the variance of the non - st samples to that of the st samples is computed ( step 608 ). finally , those coefficients whose variance ratio was large may be selected a coefficients for use in creating a population template and in testing patients ( step 612 ). in addition , at step 614 , the number of coefficients retained may be further reduced to reduce the number of computations that need to be performed , as explained above . in the version of the invention that was used to generate the data shown above , the population template was computed as follows , using the 20 st episodes in the development data set ( also used in the first of the two tables presented above ). this is illustrated at 700 in fig7 . first , the coefficients were selected as was explained above and as shown in fig6 . next , for each patient episode of st ( step 702 ), the recorded heartbeats were broken up into groups of ten consecutive heartbeats ( step 704 ). for each group ( step 706 ), the selected coefficients of the haar transform were computed ( step 708 ) for each of the ten heartbeats and the median value was then selected ( step 710 ) for each coefficient to form a proto - template . if these median values correlated well with the coefficient values for each of the ten heartbeats , the proto - template was retained ( step 714 ). otherwise , it was discarded . in this manner , a whole bunch of proto - templates were obtained from each patient episode . next , median values of the coefficients from all of the proto - templates ( step 716 ) were computed ( step 718 ). these values were then used as the population template for distinguishing st from non - st events ( step 720 ). while the preferred embodiment of the invention has been described above , it is to be understood that numerous modifications and changes will occur to those who are skilled in the art to which the invention pertains . accordingly , the following claims annexed to and forming a part of this specification are intended to define the true spirit and scope of the invention \u2014 that is , what is new and what is desired to be secured by letters patent of the united states ."}
|
{"category": "Electricity", "patent": "the present invention is directed towards improving the quality and functionality of both implantable cardioverter defibrillators ( icds ) and implantable atrial defibrillators ( iads ). all such devices need to have a method and mechanism for accurately distinguishing sinus tachycardia ( st ) form non - st , such as atrial fibrillation and atrial flutter . this method and mechanism should be conservative , biased towards not initiating therapy unless non - st is clearly indicated . it should also require as few machine computational cycles as possible to reduce icd and iad battery drain . the principle underlying the present invention is that st and non - st can be distinguished by studying and measuring the atrial electrogram morphology to detect aberrant conduction in the atria . thus , by comparing normal st waveforms to a series of newly - occurring waveforms , st is indicated by abnormal atrial waveform morphology which signals aberrant conduction within the atria improper electrical patterns of depolarization . this will show up in the signal potentials captured by a bipolar atrial lead . with reference to fig1 the present invention can be implemented in an icd 100 ( which could be an iad ) that is implanted in the chest of a patient who suffers from unpleasant and possibly life - threatening irregularities in heart operation due to improper electrical stimulation of the heart muscle cells . during normal heart action , the heart &# 39 ; s electrical impulses originate in the sino - atrial node as an action potential that is transmitted smoothly to all portions of the atria , causing contraction of the atrial chambers . the electrical impulse continues in its path to a cluster of conduction fibrils known as the atrioventricular node . after a delay of about one - tenth of a second , an action potential flows over the ventricles and causes them to contract in synchronism with and following shortly after the atrial contraction . in this manner , the heart pumps blood to the lungs and from the lungs to the body . a bipolar lead 102 is implanted within the atria 104 of a heart 103 to measure the electrical potential between nearby cells of the atria . as the action potential actively passes from cell to cell past the two electrodes of the lead 102 , an oscillating signal potential is developed across the two electrodes and is conveyed by the bipolar lead 102 to the icd 100 where the signal is sampled about 1000 times per second and is digitized , with the sequential samples stored within a ram memory within the icd 100 for further analysis . another bipolar lead 105 may extend to the ventricle ( not shown ) of the heart 103 to measure the action potentials generated within the ventricles . these same leads , or other leads , may be used by the icd for administering various types of therapy , such as pacing or defibrillation shock therapy . the data samples collected from within the atria are now analyzed . first , with reference to data gathered from the ventricles , if some event in the ventricle , such as a mis - timed depolarization event that partially overlaps the atria &# 39 ; s p wave , could leak across to the atria and distort the measurement of the potentials for a given atrial depolarization , then the data for that particular heartbeat is discarded and is not used in further analyses . such distortion can also be found simply because a given heartbeat set of data gathered from the atria is itself badly distorted , and this approach must be taken in the case of an iad having no ventricular lead . the remaining data is retained for further processing . in preparation for further processing , the data for individual heartbeats is identified and gathered . the position within the gathered data of the negative spike to the atrial p depolarization waveform is determined and is selected as the center of the waveform for each beat . then , since data close to this p - wave is to be analyzed , and since data further away may be distorted to some degree by ventricular activity or by variations in the p - to - p interval , 32 sequential data samples are selected for each heartbeat such that the negative spike of the p - wave becomes signal potential sample number 16 of the 32 sequential samples selected for further analyses . these samples for a normal heartbeat appear at 302 in fig3 and at 502 in fig5 where the amplitude of the atrial signal potential is plotted against time over the 32 sampled values , with the negative p - wave spike positioned as signal sample number 16 . likewise , the samples for an abnormal heartbeat appear at 504 in fig5 . straightforward correlation ( cwa ) of the 32 samples against a template containing a typical or normal pattern could be performed at this juncture , as taught on page 563 ( equation ( 1 )) of the article by thorne , cited in the introductory portion of this specification . but the performance of such a cross correlation at repeated regular intervals would generate numerous computations and would drain the icd &# 39 ; s battery more rapidly than is desirable . in addition , provision would have to be made for a full template waveform that can be stored within the icd to be used as a comparison reference . in addition , such templates generally need to be patient specific , and they are more susceptible to normal changes in shape . accordingly , to reduce the mathematical complexity of the cross correlation computations , it is desirable to reduce the number of data points from 32 down to some much lower number through the use of some form of transformation into a different set of values . for example , the data could be transformed using the karhunen - loeve transformation described in the morris article cited at the beginning of this specification . but that transformation is fairly computationally intensive . computationally , a fast fourier transformation would be more efficient , but any such transformation into the frequency domain , where variable frequency sine and cosine wave amplitudes result from the transformation , does not match itself well to the morphology of atrial heart waveforms , which tend to be formed from large , ringing , spike - like representations of p - wave depolarization events travelling as moving action potential fronts past the pair of electrodes attached to the bipolar lead , and not as steady harmonics extending across time . accordingly , fourier - class transforms do not adequately reduce the number of significant data values that must be considered . and in addition , when the \u201c window \u201d size for a fourier analysis is selected , if it is wide , then the frequency values are not time - specific , but represent average signal harmonic content over the entire window time duration . and if the window is narrow , then the harmonics are too spread out in frequency , and the frequency - domain data generated by the transform does not resolve things finely enough in the frequency domain . for all of the above reasons , the present invention analyzes the waveforms using a discrete wavelet transform . this has the advantage that while low frequency wavelets are broad in time , high frequency wavelets have a much narrower timeslot focus . thus , for example , with 32 data samples taken sequentially over time , a wavelet transformation generates a d . c . wavelet that spans the entire set of 32 points ; a first reversing wavelet that also spans the entire set of points ; two second harmonic wavelets that each focus upon only one - half of the 32 data points ; four third harmonic wavelets that each focus upon only one - fourth of the 32 data points ; eight fourth harmonic wavelets that each focus upon four of the 32 data points ; and 16 fifth harmonic wavelets that each focuses upon only two of the 32 data points . accordingly , the higher - frequency wavelet amplitudes are quite time specific , unlike fourier sinusoidal amplitudes , while the lower - frequency and d . c . wavelet amplitudes give one broad information about the whole set of 32 points . and like the fourier transform , the discrete wavelet transform preserves all of the information of the original atrial signal , such that the transformation is fully reversible . 32 discrete wavelet transform values may be reverse transformed back into 32 values indicating the atrial signal potentials at 32 sequential points in time . in essence , the haar function is performing multiple digital filtering operations , at variable positions and utilizing varying - width windowing functions , to develop component values that represent varying types of heartbeat activity , at varying frequencies , spread over varying widths , and located at varying positions . these component values can be studied to see which subset of these component values are good at distinguishing st from non - st , or ( more generally ) which subset of these component values are good at distinguishing one type of heartbeat waveform from another . further study of such a selected subset can further determine which of the subset of components are relatively invariant from one individual &# 39 ; s heart to another . one preferably selects component values that are suitable from both of these two perspectives . the particular wavelet transformation to choose can be tailored to the nature of the data , with the wavelets chosen to resemble somewhat the values to be found within the data so that some transformed values are of much larger amplitude than others , and such that the low amplitude transformed values may be disregarded . of particular importance with an atrial p waveform of the type being analyzed here are wavelets having frequency values roughly comparable to and timed to coincide with the ringing of the atrial depolarization . this tailoring can minimize the number of transformed data values that are significant and that are candidates for full participation in the correlation analyses to determine if there has been a substantial change in waveform morphology . another factor in selecting a particular wavelet transformation is reducing the number of computations that must be performed to carry out the transformation . yet another factor is selecting a wavelet transformation where some of the transformed component values vary from normal to abnormal heart rhythms in much the same over a population of individuals so that a generic template of these component values may be derived that can be used with many individuals , rather than with just one individual . these factors suggest that the haar transformation would be a suitable candidate for use in analysis of atrial waveforms and possibly other waveforms as well . with reference to fig3 in the case of 32 voltage values sampled over time , the 32 applicable haar discrete transform wavelets appear as shown in this figure . other discrete wavelet transforms based upon wavelets having triangular or other shapes may also prove usable . the haar transform wavelets , in particular , have the shape of a square waveform , as will be described , and this can simplify the computations required . in fig3 time increases from left to right . a scale 304 indicates the 32 points in time at which the atrial waveform is sampled , and an analog atrial waveform ( before digitization ) is shown at 302 . the p wave negative notch 303 is shown centered at the 16 th timeslot so that the signal potential of the atria waveform sampled at this point in time becomes the 16 th data value in the set of 32 data values that are to be subjected to the haar transformation . a haar wavelet is simply one cycle of a square waveform that starts at zero , then swings positive ( to \u201c+ 1 \u201d), then swings negative ( to \u201c\u2212 1 \u201d), and then swings back to zero . the wavelet w 2 , shown at 308 in fig3 for example , is at \u201c+ 1 \u201d at points in time 1 to 16 and is at \u201c\u2212 1 \u201d at points in time 17 to 32 . the shorter waveform w 6 , shown at 316 in fig3 is at \u201c+ 1 \u201d at points in time 9 to 12 and is at \u201c\u2212 1 \u201d at points in time 13 to 16 , and is at \u201c 0 \u201d at all other points in time . the waveform w 1 , shown at 306 in fig3 is so slow to fluctuate in time that its \u201c\u2212 1 \u201d portion is off of the chart ( in fig3 ) to the right , and is ignored ; and accordingly , it has the value \u201c+ 1 \u201d for all 32 of the points in time shown in fig3 . it thus represents the average , or d . c ., component of the atrial signal potential . the lowest frequency wavelets w 1 and w 2 encompass the entire set of 32 points in time . the higher frequency wavelets w 3 and w 4 each encompasses only half of the points in time , and the two wavelets taken together encompass all the points in time . the four wavelets w 5 , w 6 , w 7 , and w 8 each encompasses only eight points in time , and the four wavelets taken together encompass all points in time . the eight wavelets w 9 . . . w 16 each encompasses only four points in time , and together all eight wavelets encompass all points in time . and finally , the wavelets w 17 . . . w 32 each encompass only two points in time , but the sixteen wavelets together encompass all points in time . thus , each wavelet has a characteristic time span and position as well as a characteristic frequency , with higher - frequency wavelets having a narrower and more specific time span and position than lower - frequency wavelets . this is advantageous with an impulse - type , ringing signal such as the atrial depolarization waveform considered here , since many higher - frequency transform values that correspond to haar wavelets positioned in time away from the p waveform or that do not correspond to its frequency of ringing may be low in amplitude such that they may safely be ignored during the analysis steps , thereby reducing the data that must be processed during the correlation steps . ( the above , while presented in the context of the haar wavelet , is also applicable to other wavelet shapes that may used to perform a discrete wavelet transform ( dwt )). each of the 32 haar transformed values is computed as follows : multiply each of the 32 sampled and digitized voltage values representing the analog atrial waveform 302 by the correspondingly - positioned -( in - time ) amplitude values of one of the 32 haar wavelets ( shown in fig3 ); then sum the resulting products ; and then multiply the resulting sums by a discrete wavelet transform scaling factor 2 \u2212 j / 2 , where j equals 1 for the wavelets w 17 to w 32 , 2 for w 9 to w 16 , 3 for w 5 to w 8 , 4 for w 3 to w 4 , and 5 for w 1 and w 2 . for example , and referring to fig3 : the first wavelet w 1 is always + 1 , so the atrial signal potential values are simply summed and then multiplied by 2 \u2212 5 / 2 ; the second wavelet is + 1 for time points 1 to 16 and \u2212 1 for time points 17 to 32 , so the first 16 values are summed , the second 16 values are summed , the difference between the first and second sums is computed , and the result is multiplied by 2 \u2212 5 / 2 ; and so on until all the transformed values have been computed , one for each haar wavelet . the 32 time - sequential atrial signal potential values are thus transformed into 32 haar wavelet amplitude values which may be called \u201c transformed values \u201d and which may be assigned the numbers 1 to 32 corresponding to the subscript numbers of the haar wavelets to which they each correspond and whose amplitudes they represent . the above description of how to compute the haar wavelet transformed values is accurate , but it is not the most efficient way to proceed in the case of haar wavelets . like the fourier transform , which has a corresponding fast fourier transform that saves intermediate results and thereby avoids re - computing them and thus reduces substantially the number of computations , the haar transformation also may be carried out in a manner that saves intermediate sums and re - uses them to reduce substantially the number of computations . this is described below at the point where fig4 is described , since fig4 illustrates graphically how this can be done . the illustrative program listing presented below is also an implementation of the fast haar wavelet transformation algorithm . the haar transformation can readily be carried out by any digital computer . as an example of how the haar transformation can be carried out on 32 values , the following program , written in the language c , is illustrative of many possible programs that may be written . in the exemplary program that follows , a 32 - element array yy contains 32 data values representing the 32 sampled atrial signal potential values representing the fluctuations over time of the signal supplied by the bipolar lead 102 . the analog atrial signal is sampled , digitized , broken into separate beat data sets , pre - processed ( to remove waveforms distorted by ventricular activity ), centered ( with the negative depolarization spike at data point 16 ), and fed into the subroutine presented below . this subroutine is compiled ( or assembled , if rewritten in assembly language ) and installed in the rom of the icd &# 39 ; s embedded microprocessor . this subroutine returns , contained within the same array yy , the 32 transformed haar wavelet amplitude values described above . ( the constant value \u201c recipsqrttwo \u201d is the reciprocal of the square root of two ). an illustrative version of the subroutine for computing all 32 of the haar transformed values in an efficient manner is presented here : void haar ( double yy []) { int i , j , l ; double zz [ 32 ]; for ( i = 5 ; i & gt ; 0 ; i \u2212\u2212) { l = 1 ; for ( j = 1 ; j & lt ;= i ; ++ j ) l = 2 * l ; for ( j = 0 ; j & lt ; l ; ++ j ) zz [ j ] = yy [ j ]; for ( j = 0 ; j & lt ; l \u2212 1 ; j = j + 2 ) ( yy [ j / 2 ] = recipsqrttwo * ( zz [ j ] + zz [ j + 1 ]); yy [( j + l )/ 2 ] = rcipsqrttwo * ( zz [ j ] \u2212 zz [ j + 1 ]); } } return ; } the above program computes the 32 haar transformed values 1 to 32 from the 32 atrial signal potential time - sequenced input values . it does so with only 31 additions , 31 subtractions , and 62 multiplications . it is carefully designed to compute each value in a systematic way , making multiple use of intermediate results . first , the program computes sixteen sums of and sixteen differences between eight adjacent pairs of the 16 time sequenced atrial signal potential values , and the sixteen difference values become the haar transformed values 17 to 32 , reflecting the strength of the highest frequency haar wavelets positioned at 16 different positions in time , corresponding to the wavelets w 17 to w 32 shown in fig3 . all the sum and difference values are scaled by multiplication by 2 \u2212 1 / 2 . next , taking the 16 sums of atrial signal potential values as an intermediate result , the above algorithm generates eight sums of and eight differences between adjacent pairs of these sixteen intermediate values that resulted from the first sixteen additions , and the eight newly - computed difference values become the haar transformed values 9 to 16 , reflecting the strength of the second to the highest frequency values at eight different points in time , corresponding to wavelets w 9 to w 16 in fig3 . all of these eight sum and eight difference values are again scaled by 2 \u2212 1 / 2 so that the wavelets w 9 to w 16 are scaled by \u00bd ( 2 \u2212 2 / 2 or 2 \u2212 1 / 2 times 2 \u2212 1 / 2 ). next , taking these eight sums of sums of adjacent atrial signal potential values as an intermediate result , the above algorithm generates four sum and four difference values , again scaling by 2 \u2212 1 / 2 , and the four difference values become haar transformed values 5 to 8 , reflecting the strength of the middle frequency wavelets at four different points in time , and corresponding to wavelets w 5 to w 8 in fig3 . next , taking these four sums of sums of sums of adjacent atrial signal potential values , the above algorithm generates two sum and two difference values , again scaling by 2 \u2212 1 / 2 , and the two difference values become haar transformed values 3 and 4 , reflecting the strength of the second to the lowest frequency wavelets only two of which encompass all the time domain data , corresponding to the wavelets w 3 and w 4 shown in fig3 . and finally , the algorithm generates the sum of and the difference between the final remaining two sums of sums of sums of sums of atrial signal potential values , again scaling by 2 \u2212 1 / 2 . the difference value is then the haar transformed value 2 , representing the strength of the lowest - frequency wavelet , the one corresponding to the wavelet w 2 in fig3 the wavelet that extends the full length of the time scale . the sum value is then the first , or d . c ., transformed value , representing the average signal potential level over the 32 sampled points in time , which corresponds to the wavelet w 1 in fig3 . another way of viewing this computation is illustrated in fig4 a and 4b . the 32 atrial signal potential values are shown at the top of fig4 a and are identified as c 0 5 through c 31 5 . the superscript \u201c 5 \u201d indicates that these values are processed when the index value \u201c n \u201d in the above computer program is equal to \u201c 5 \u201d\u2014 that is , during the first pass through the data generating intermediary sums and differences . during subsequent passes , the value of n is decremented to 4 , 3 , 2 , and finally to 1 . during each pass through the data , the computer generates sums c n - 1 and differences d n - 1 , as shown in fig4 b , between adjacent pairs of the values c 0 n and c 1 n ; c 2 n and c 3 n ; and so on , so that the number of newly - generated c n - 1 and d n - 1 terms is reduced by half with each computer pass through the intermediary results . at the end of all these computations , the value c 0 0 is the first , or d . c ., haar transformed value ; and the values d 0 0 ; d 0 1 and d 1 1 ; d 0 2 , d 1 2 , d 2 2 , and d 3 2 ; d 0 3 , d 1 3 , . . . and d 7 3 ; and d 0 4 , d 1 4 , and d 15 4 are , respectively , the remaining haar transform values 2 through 32 . fig4 thus illustrates quite succinctly how all the transform computations are carried out , and how the intermediary \u201c c \u201d sum values , such as c 0 4 , are used to compute multiple haar values , such as the values d 0 3 , d 0 2 , d 0 1 , d 0 0 , and c 0 0 all of which are computed from the intermediary value c 0 4 . saving and reusing these intermediary \u201c c \u201d values saves much computational time . fig4 b indicates the precise addition and subtraction operations that are carried out at each level to compute the values in the next lower level , proceeding down through the chart presented in fig4 a . but in any given application to heart waveform analysis , all of these computations may not be needed , and accordingly the number of computations may be reduced much further . since the present invention teaches that only a small number of these haar transformed values need actually be considered , a far less computationally intensive transform can be developed which only generates the intermediate and final transformed values that are actually needed to generate the specific haar transformed values which have proved to be significant in discriminating between sinus tachycardia ( or st ) and non - st conditions . all others need not be computed , and the above program may be reduced to a special algorithm that omits as many sums , differences , and multiplications as possible . for example , if only the transformed values 1 , 5 , 9 , and 24 are significant , then 31 additions are required to compute the first transformed value ( the d . c . value \u2014 the simple sum of all the time domain signal values ); six additions and one subtraction are required to compute the fifth transformed value ( in fig3 the difference between the sums of time domain values under each half of the wavelet w 5 at 314 in fig3 ); two additions and one subtraction are required to compute the ninth transformed value ( in fig3 the difference between the sums of the time domain values under each half of the wavelet w9 at 322 in fig3 ); and only one subtraction is required to compute the 24 th transformed value ( in fig3 the difference between the two time domain values under the respective halves of the wavelet w 24 ( not shown ) which is the same size as , but differently positioned in time than , the wavelet w 18 at 330 ( fig3 ). accordingly , if only the transformed values 1 , 5 , 9 , and 24 actually evaluated , then only 42 additions and subtractions are required . but even this number can be reduced further . if the computational algorithm set forth in the above illustrative computer program is followed as a guide , then the intermediary sums used in computing the haar first , or d . c ., transformed value can be re - used to compute the sums for the haar transformed values 5 and 9 , assuming these intermediary results are saved in the manner described in the above program example . then 31 additions are still required to compute the first transformed value , but only one subtraction is required to compute each of the transformed values 5 , 9 , and 24 , giving a total of additions and subtraction of only 34 operations . and if only transformed values 1 , 5 , and 9 are computed , the number of additions and subtractions is reduced to 33 . and if the first transformed value is omitted and if transformed values 5 , 9 , and 24 are selected , then the 31 additions needed to compute the first transformed value are not required . then the number of computations for transformed value 5 is 8 additions and one subtraction ; the number of computations for transformed value 9 is 2 additions and one subtraction ; and the number of computations for transformed value 24 is still just one subtraction . so the total number of additions and subtractions is just 13 . but even this number can be reduced to 11 when it is realized that the two additions done for the transformed value 9 are also done ( in the above computational algorithm ) when computing the transformed value 5 . so the number of additions and subtractions can be reduced to 11 . also , when computing such a small number of transformed values , the number of multiplications can be reduced as well by postponing the multiplications until a transformed value is actually computed . with reference to fig4 a , instead of dividing every sum and difference by the square root of two ( as shown in fig4 b ), several vertical sums of \u201c c \u201d values in different rows of the fig4 a table can be formed , and then a \u201c d \u201d value can be computed by subtraction , and then the resulting unscaled transformed value can be scaled with a single multiplication by 2 \u2212 j / 2 where j is the number of rows in the table of fig4 a traversed by the one or more sum and the one difference computations . thus , the number of scaling multiplications can sometimes be equal to the number of transformed values that are computed , typically 3 or 4 . accordingly , by using wavelet analysis and transformation , instead of fourier or discrete cosine or other sinusoid analysis and transformation , a highly useful and highly frequency - specific result can be achieved with a very small number of computations , thereby conserving power and battery life , and yet achieving a high degree of precision in recognizing changes in morphology . in our tests , we have selected certain transformed values as being much more significant than other values in distinguishing between st and non - st . we focused upon those transform values whose \u201c+ 1 \u201d and \u201c\u2212 1 \u201d scope included the middle 8 time points most significant to analysis of the peak of the atrial depolarization . working across test data samples obtained from patients who had dual chamber icds , we computed the variance of each transformed value for st and non - st events using the standard statistical formula for computing variance . we then calculated the ratio of the variance of non - st events to the variance of st events , and selected those terms with the highest variance ratios for further consideration . in this manner , we reduced the number of transformed values that were included in the cwa or correlation process while always maintaining or improving the performance achieved . ultimately we settled upon the three or four transformed values having the highest ratios . these were the transform values 1 , 5 , 9 , and 24 . we tested 1 , 5 , and 24 together ; 5 , 9 , and 24 together ; and 1 , 5 , 9 , and 24 together . these selected haar transform values , when used in these combinations , required minimal computations and thereby produced a savings in battery power , and that also gave better performance at distinguishing st from non - st than did all of the transformed values used together . so an increase in accuracy was achieved as well as a decrease in computational complexity by this approach to atrial waveform analysis . [ 0054 ] fig5 for example , illustrates actual plots of digital information illustrating the effect of using only the three coefficients 1 , 5 , and 24 . at 502 , a normal waveform is shown , with amplitude plotted against sample numbers from 1 to 32 . at 504 , an abnormal waveform is shown , again with amplitude plotted against sample numbers . after the haar transformation , at 506 and 508 , the wavelet coefficient amplitude values are indicated for the coefficients 1 to 32 . at 506 , the coefficients generated by the normal waveform 502 are shown , and at 508 , the coefficients generated by the abnormal waveform 504 are shown . one may directly compare these component values and verify that the coefficients 1 , 5 , and 24 at 506 and at 508 vary significantly in amplitude . comparison of this same information among different patients ( not shown in this figure ) also indicated that this variation is relatively constant from one patient to another . at 510 and 512 , using only the coefficients 1 , 5 , and 24 , the heartbeat waveforms are reconstructed by a reverse transformation , the normal reconstructed waveform shown at 510 and the abnormal reconstructed waveform shown at 512 . the marked differences between these two reconstructed waveforms highlights the way in which these three coefficients can signal an abnormal condition with less computation . in a practical system , a small number of transform values are selected , in the manner just described . the algorithm for computing these values is then refined , as explained above , to reduce the number of additions , subtractions , and multiplications to the minimum possible while preserving the accuracy of the computations . we first compute a value \u03c1 , which is the correlation between these values with the corresponding values in a template that represents the values for an average normal population . we then compare this value \u03c1 to a threshold correlation value \u03b2 that is chosen to give optimal results on experimental data . ( see the examples in the tables presented below .) for each waveform analyzed , a test is made to determine whether \u03c1 is greater than \u03b2 . if so , then this waveform is placed into the buffer marked \u201c st \u201d at step 210 . if not , then this waveform is placed into the buffer marked \u201c non - st \u201d at step 211 . finally , after ten waveforms have been analyzed , a count of st and non - st marked waveforms is made to see if the count of non - st waveforms is greater than some threshold value x ( at step 212 ), where x can be , for example , 7 . if more than seven waveforms are non - st , then therapy is delivered at step 214 . the buffers at steps 210 and 211 in fig2 are buffers that hold the last ten decision values . these buffers can be pictured as sliding windows revealing the most recent ten values to permit the decision at 212 to be continuously updated . the buffers thus function as a nonlinear filter preventing irregular values from delivering therapy improperly . in developing a working prototype of this system , we used a development data set consisting of 20 episodes of st taken from 4 patients and 18 episodes of af taken from 7 patients . patient number episodes of st episodes of af 1 2 6 2 12 0 13 0 1 4 0 2 5 5 1 6 0 3 7 0 3 8 0 2 9 1 0 total 20 18 we used this data to optimize our choice of \u03c1 and \u03b2 and also the particular coefficients that we chose to examine . next , we tested our prototype using a set of 16 episodes of st obtained from 4 patients together with 17 episodes of af obtained from 7 patients . patient number episodes of st episodes of af 1 8 0 2 3 1 3 3 0 4 2 0 5 0 4 6 0 4 7 0 3 8 0 3 9 0 1 10 0 1 total 16 17 x out threshold coefficients of 10 ( \u03b2 ) sensitivity specificity 1 , 5 , 9 , & amp ; 24 3 0 . 808 76 % 94 % 1 , 5 , 9 , & amp ; 24 5 0 . 975 88 % 94 % 1 , 5 , 9 , & amp ; 24 6 0 . 976 82 % 94 % 1 , 5 , 9 , & amp ; 24 7 0 . 983 88 % 94 % 1 , 5 , & amp ; 24 3 0 . 815 82 % 94 % 1 , 5 , & amp ; 24 4 0 . 943 88 % 94 % 1 , 5 , & amp ; 24 5 0 . 956 82 % 94 % 1 , 5 , & amp ; 24 7 0 . 991 94 % 94 % 5 , 9 , & amp ; 24 8 0 . 9993 82 % 94 % 5 , 9 , & amp ; 24 9 0 . 9997 76 % 81 % the quality of the selected subset of transformed values for characterizing changes in morphology may be demonstrated graphically , as indicated in fig5 by performing a reverse haar transform using only the three or four or so selected transformed values . as can be seen ( at 510 and 512 in fig5 ), the inverse transformations produce distorted representations of the original atrial waveforms which illustrate how the choice of transform values can emphasize the changes . this is another useful way to assist one in selecting which transformed values are most useful . one feature of the invention is its ability to use a template derived from an average normal population , rather than deriving a patient customized average normal template for each individual patient . prior systems have required each new patient to be monitored while in a normal state , and the captured data was then averaged to form a patient specific normal template . contrary to this , the present invention achieved the results shown above using the same template for all patients . accordingly , the invention may be used with a new patient without the necessity of such preliminary testing of each patient and without a new customized template necessarily having to be created for each patient . one reason why the present invention can function using a template that represents average values for a normal population is because the specific coefficients selected in the transform domain , in addition to having been selected to emphasize the difference between normal and abnormal values , are also preferably chosen to minimize the differences between different patients . in some cases , values that were good at distinguishing between normal and abnormal rhythms for one patient were not as good at doing so for some other patient . these values are preferably not selected for use in implementing the present invention . for example , the transform coefficients may be selected such that all ( or most of ) the normal values of a particular coefficient gathered from many patients had the same sign ( positive or negative ) and similar amplitudes ( or if they varied in sign , they were near to zero ). in addition , the abnormal coefficients are significantly different in value from the normal coefficients for each particular patient . [ 0068 ] fig6 at 600 , illustrates one way in which this may be done . first , at step 602 , one gathers a large number of st and non - st data samples . next , all of this data is transformed , generating coefficients for every waveform set of data ( step 604 ). the variance of each coefficient is then computed first for the st samples ( step 606 ) and then for the non - st samples . then , for each coefficient , the ratio of the variance of the non - st samples to that of the st samples is computed ( step 608 ). finally , those coefficients whose variance ratio was large may be selected a coefficients for use in creating a population template and in testing patients ( step 612 ). in addition , at step 614 , the number of coefficients retained may be further reduced to reduce the number of computations that need to be performed , as explained above . in the version of the invention that was used to generate the data shown above , the population template was computed as follows , using the 20 st episodes in the development data set ( also used in the first of the two tables presented above ). this is illustrated at 700 in fig7 . first , the coefficients were selected as was explained above and as shown in fig6 . next , for each patient episode of st ( step 702 ), the recorded heartbeats were broken up into groups of ten consecutive heartbeats ( step 704 ). for each group ( step 706 ), the selected coefficients of the haar transform were computed ( step 708 ) for each of the ten heartbeats and the median value was then selected ( step 710 ) for each coefficient to form a proto - template . if these median values correlated well with the coefficient values for each of the ten heartbeats , the proto - template was retained ( step 714 ). otherwise , it was discarded . in this manner , a whole bunch of proto - templates were obtained from each patient episode . next , median values of the coefficients from all of the proto - templates ( step 716 ) were computed ( step 718 ). these values were then used as the population template for distinguishing st from non - st events ( step 720 ). while the preferred embodiment of the invention has been described above , it is to be understood that numerous modifications and changes will occur to those who are skilled in the art to which the invention pertains . accordingly , the following claims annexed to and forming a part of this specification are intended to define the true spirit and scope of the invention \u2014 that is , what is new and what is desired to be secured by letters patent of the united states ."}
|
Does the patent belong in this category?
| 0.25 |
1cd5305994f1692ec7ec09368a2a725ff3c3dd15c21bf56c12e04e558cdb5fbc
| 0.145508 | 0.373047 | 0.172852 | 0.128906 | 0.314453 | 0.847656 |
null |
{"patent": "the present invention is directed towards improving the quality and functionality of both implantable cardioverter defibrillators ( icds ) and implantable atrial defibrillators ( iads ). all such devices need to have a method and mechanism for accurately distinguishing sinus tachycardia ( st ) form non - st , such as atrial fibrillation and atrial flutter . this method and mechanism should be conservative , biased towards not initiating therapy unless non - st is clearly indicated . it should also require as few machine computational cycles as possible to reduce icd and iad battery drain . the principle underlying the present invention is that st and non - st can be distinguished by studying and measuring the atrial electrogram morphology to detect aberrant conduction in the atria . thus , by comparing normal st waveforms to a series of newly - occurring waveforms , st is indicated by abnormal atrial waveform morphology which signals aberrant conduction within the atria improper electrical patterns of depolarization . this will show up in the signal potentials captured by a bipolar atrial lead . with reference to fig1 the present invention can be implemented in an icd 100 ( which could be an iad ) that is implanted in the chest of a patient who suffers from unpleasant and possibly life - threatening irregularities in heart operation due to improper electrical stimulation of the heart muscle cells . during normal heart action , the heart &# 39 ; s electrical impulses originate in the sino - atrial node as an action potential that is transmitted smoothly to all portions of the atria , causing contraction of the atrial chambers . the electrical impulse continues in its path to a cluster of conduction fibrils known as the atrioventricular node . after a delay of about one - tenth of a second , an action potential flows over the ventricles and causes them to contract in synchronism with and following shortly after the atrial contraction . in this manner , the heart pumps blood to the lungs and from the lungs to the body . a bipolar lead 102 is implanted within the atria 104 of a heart 103 to measure the electrical potential between nearby cells of the atria . as the action potential actively passes from cell to cell past the two electrodes of the lead 102 , an oscillating signal potential is developed across the two electrodes and is conveyed by the bipolar lead 102 to the icd 100 where the signal is sampled about 1000 times per second and is digitized , with the sequential samples stored within a ram memory within the icd 100 for further analysis . another bipolar lead 105 may extend to the ventricle ( not shown ) of the heart 103 to measure the action potentials generated within the ventricles . these same leads , or other leads , may be used by the icd for administering various types of therapy , such as pacing or defibrillation shock therapy . the data samples collected from within the atria are now analyzed . first , with reference to data gathered from the ventricles , if some event in the ventricle , such as a mis - timed depolarization event that partially overlaps the atria &# 39 ; s p wave , could leak across to the atria and distort the measurement of the potentials for a given atrial depolarization , then the data for that particular heartbeat is discarded and is not used in further analyses . such distortion can also be found simply because a given heartbeat set of data gathered from the atria is itself badly distorted , and this approach must be taken in the case of an iad having no ventricular lead . the remaining data is retained for further processing . in preparation for further processing , the data for individual heartbeats is identified and gathered . the position within the gathered data of the negative spike to the atrial p depolarization waveform is determined and is selected as the center of the waveform for each beat . then , since data close to this p - wave is to be analyzed , and since data further away may be distorted to some degree by ventricular activity or by variations in the p - to - p interval , 32 sequential data samples are selected for each heartbeat such that the negative spike of the p - wave becomes signal potential sample number 16 of the 32 sequential samples selected for further analyses . these samples for a normal heartbeat appear at 302 in fig3 and at 502 in fig5 where the amplitude of the atrial signal potential is plotted against time over the 32 sampled values , with the negative p - wave spike positioned as signal sample number 16 . likewise , the samples for an abnormal heartbeat appear at 504 in fig5 . straightforward correlation ( cwa ) of the 32 samples against a template containing a typical or normal pattern could be performed at this juncture , as taught on page 563 ( equation ( 1 )) of the article by thorne , cited in the introductory portion of this specification . but the performance of such a cross correlation at repeated regular intervals would generate numerous computations and would drain the icd &# 39 ; s battery more rapidly than is desirable . in addition , provision would have to be made for a full template waveform that can be stored within the icd to be used as a comparison reference . in addition , such templates generally need to be patient specific , and they are more susceptible to normal changes in shape . accordingly , to reduce the mathematical complexity of the cross correlation computations , it is desirable to reduce the number of data points from 32 down to some much lower number through the use of some form of transformation into a different set of values . for example , the data could be transformed using the karhunen - loeve transformation described in the morris article cited at the beginning of this specification . but that transformation is fairly computationally intensive . computationally , a fast fourier transformation would be more efficient , but any such transformation into the frequency domain , where variable frequency sine and cosine wave amplitudes result from the transformation , does not match itself well to the morphology of atrial heart waveforms , which tend to be formed from large , ringing , spike - like representations of p - wave depolarization events travelling as moving action potential fronts past the pair of electrodes attached to the bipolar lead , and not as steady harmonics extending across time . accordingly , fourier - class transforms do not adequately reduce the number of significant data values that must be considered . and in addition , when the \u201c window \u201d size for a fourier analysis is selected , if it is wide , then the frequency values are not time - specific , but represent average signal harmonic content over the entire window time duration . and if the window is narrow , then the harmonics are too spread out in frequency , and the frequency - domain data generated by the transform does not resolve things finely enough in the frequency domain . for all of the above reasons , the present invention analyzes the waveforms using a discrete wavelet transform . this has the advantage that while low frequency wavelets are broad in time , high frequency wavelets have a much narrower timeslot focus . thus , for example , with 32 data samples taken sequentially over time , a wavelet transformation generates a d . c . wavelet that spans the entire set of 32 points ; a first reversing wavelet that also spans the entire set of points ; two second harmonic wavelets that each focus upon only one - half of the 32 data points ; four third harmonic wavelets that each focus upon only one - fourth of the 32 data points ; eight fourth harmonic wavelets that each focus upon four of the 32 data points ; and 16 fifth harmonic wavelets that each focuses upon only two of the 32 data points . accordingly , the higher - frequency wavelet amplitudes are quite time specific , unlike fourier sinusoidal amplitudes , while the lower - frequency and d . c . wavelet amplitudes give one broad information about the whole set of 32 points . and like the fourier transform , the discrete wavelet transform preserves all of the information of the original atrial signal , such that the transformation is fully reversible . 32 discrete wavelet transform values may be reverse transformed back into 32 values indicating the atrial signal potentials at 32 sequential points in time . in essence , the haar function is performing multiple digital filtering operations , at variable positions and utilizing varying - width windowing functions , to develop component values that represent varying types of heartbeat activity , at varying frequencies , spread over varying widths , and located at varying positions . these component values can be studied to see which subset of these component values are good at distinguishing st from non - st , or ( more generally ) which subset of these component values are good at distinguishing one type of heartbeat waveform from another . further study of such a selected subset can further determine which of the subset of components are relatively invariant from one individual &# 39 ; s heart to another . one preferably selects component values that are suitable from both of these two perspectives . the particular wavelet transformation to choose can be tailored to the nature of the data , with the wavelets chosen to resemble somewhat the values to be found within the data so that some transformed values are of much larger amplitude than others , and such that the low amplitude transformed values may be disregarded . of particular importance with an atrial p waveform of the type being analyzed here are wavelets having frequency values roughly comparable to and timed to coincide with the ringing of the atrial depolarization . this tailoring can minimize the number of transformed data values that are significant and that are candidates for full participation in the correlation analyses to determine if there has been a substantial change in waveform morphology . another factor in selecting a particular wavelet transformation is reducing the number of computations that must be performed to carry out the transformation . yet another factor is selecting a wavelet transformation where some of the transformed component values vary from normal to abnormal heart rhythms in much the same over a population of individuals so that a generic template of these component values may be derived that can be used with many individuals , rather than with just one individual . these factors suggest that the haar transformation would be a suitable candidate for use in analysis of atrial waveforms and possibly other waveforms as well . with reference to fig3 in the case of 32 voltage values sampled over time , the 32 applicable haar discrete transform wavelets appear as shown in this figure . other discrete wavelet transforms based upon wavelets having triangular or other shapes may also prove usable . the haar transform wavelets , in particular , have the shape of a square waveform , as will be described , and this can simplify the computations required . in fig3 time increases from left to right . a scale 304 indicates the 32 points in time at which the atrial waveform is sampled , and an analog atrial waveform ( before digitization ) is shown at 302 . the p wave negative notch 303 is shown centered at the 16 th timeslot so that the signal potential of the atria waveform sampled at this point in time becomes the 16 th data value in the set of 32 data values that are to be subjected to the haar transformation . a haar wavelet is simply one cycle of a square waveform that starts at zero , then swings positive ( to \u201c+ 1 \u201d), then swings negative ( to \u201c\u2212 1 \u201d), and then swings back to zero . the wavelet w 2 , shown at 308 in fig3 for example , is at \u201c+ 1 \u201d at points in time 1 to 16 and is at \u201c\u2212 1 \u201d at points in time 17 to 32 . the shorter waveform w 6 , shown at 316 in fig3 is at \u201c+ 1 \u201d at points in time 9 to 12 and is at \u201c\u2212 1 \u201d at points in time 13 to 16 , and is at \u201c 0 \u201d at all other points in time . the waveform w 1 , shown at 306 in fig3 is so slow to fluctuate in time that its \u201c\u2212 1 \u201d portion is off of the chart ( in fig3 ) to the right , and is ignored ; and accordingly , it has the value \u201c+ 1 \u201d for all 32 of the points in time shown in fig3 . it thus represents the average , or d . c ., component of the atrial signal potential . the lowest frequency wavelets w 1 and w 2 encompass the entire set of 32 points in time . the higher frequency wavelets w 3 and w 4 each encompasses only half of the points in time , and the two wavelets taken together encompass all the points in time . the four wavelets w 5 , w 6 , w 7 , and w 8 each encompasses only eight points in time , and the four wavelets taken together encompass all points in time . the eight wavelets w 9 . . . w 16 each encompasses only four points in time , and together all eight wavelets encompass all points in time . and finally , the wavelets w 17 . . . w 32 each encompass only two points in time , but the sixteen wavelets together encompass all points in time . thus , each wavelet has a characteristic time span and position as well as a characteristic frequency , with higher - frequency wavelets having a narrower and more specific time span and position than lower - frequency wavelets . this is advantageous with an impulse - type , ringing signal such as the atrial depolarization waveform considered here , since many higher - frequency transform values that correspond to haar wavelets positioned in time away from the p waveform or that do not correspond to its frequency of ringing may be low in amplitude such that they may safely be ignored during the analysis steps , thereby reducing the data that must be processed during the correlation steps . ( the above , while presented in the context of the haar wavelet , is also applicable to other wavelet shapes that may used to perform a discrete wavelet transform ( dwt )). each of the 32 haar transformed values is computed as follows : multiply each of the 32 sampled and digitized voltage values representing the analog atrial waveform 302 by the correspondingly - positioned -( in - time ) amplitude values of one of the 32 haar wavelets ( shown in fig3 ); then sum the resulting products ; and then multiply the resulting sums by a discrete wavelet transform scaling factor 2 \u2212 j / 2 , where j equals 1 for the wavelets w 17 to w 32 , 2 for w 9 to w 16 , 3 for w 5 to w 8 , 4 for w 3 to w 4 , and 5 for w 1 and w 2 . for example , and referring to fig3 : the first wavelet w 1 is always + 1 , so the atrial signal potential values are simply summed and then multiplied by 2 \u2212 5 / 2 ; the second wavelet is + 1 for time points 1 to 16 and \u2212 1 for time points 17 to 32 , so the first 16 values are summed , the second 16 values are summed , the difference between the first and second sums is computed , and the result is multiplied by 2 \u2212 5 / 2 ; and so on until all the transformed values have been computed , one for each haar wavelet . the 32 time - sequential atrial signal potential values are thus transformed into 32 haar wavelet amplitude values which may be called \u201c transformed values \u201d and which may be assigned the numbers 1 to 32 corresponding to the subscript numbers of the haar wavelets to which they each correspond and whose amplitudes they represent . the above description of how to compute the haar wavelet transformed values is accurate , but it is not the most efficient way to proceed in the case of haar wavelets . like the fourier transform , which has a corresponding fast fourier transform that saves intermediate results and thereby avoids re - computing them and thus reduces substantially the number of computations , the haar transformation also may be carried out in a manner that saves intermediate sums and re - uses them to reduce substantially the number of computations . this is described below at the point where fig4 is described , since fig4 illustrates graphically how this can be done . the illustrative program listing presented below is also an implementation of the fast haar wavelet transformation algorithm . the haar transformation can readily be carried out by any digital computer . as an example of how the haar transformation can be carried out on 32 values , the following program , written in the language c , is illustrative of many possible programs that may be written . in the exemplary program that follows , a 32 - element array yy contains 32 data values representing the 32 sampled atrial signal potential values representing the fluctuations over time of the signal supplied by the bipolar lead 102 . the analog atrial signal is sampled , digitized , broken into separate beat data sets , pre - processed ( to remove waveforms distorted by ventricular activity ), centered ( with the negative depolarization spike at data point 16 ), and fed into the subroutine presented below . this subroutine is compiled ( or assembled , if rewritten in assembly language ) and installed in the rom of the icd &# 39 ; s embedded microprocessor . this subroutine returns , contained within the same array yy , the 32 transformed haar wavelet amplitude values described above . ( the constant value \u201c recipsqrttwo \u201d is the reciprocal of the square root of two ). an illustrative version of the subroutine for computing all 32 of the haar transformed values in an efficient manner is presented here : void haar ( double yy []) { int i , j , l ; double zz [ 32 ]; for ( i = 5 ; i & gt ; 0 ; i \u2212\u2212) { l = 1 ; for ( j = 1 ; j & lt ;= i ; ++ j ) l = 2 * l ; for ( j = 0 ; j & lt ; l ; ++ j ) zz [ j ] = yy [ j ]; for ( j = 0 ; j & lt ; l \u2212 1 ; j = j + 2 ) ( yy [ j / 2 ] = recipsqrttwo * ( zz [ j ] + zz [ j + 1 ]); yy [( j + l )/ 2 ] = rcipsqrttwo * ( zz [ j ] \u2212 zz [ j + 1 ]); } } return ; } the above program computes the 32 haar transformed values 1 to 32 from the 32 atrial signal potential time - sequenced input values . it does so with only 31 additions , 31 subtractions , and 62 multiplications . it is carefully designed to compute each value in a systematic way , making multiple use of intermediate results . first , the program computes sixteen sums of and sixteen differences between eight adjacent pairs of the 16 time sequenced atrial signal potential values , and the sixteen difference values become the haar transformed values 17 to 32 , reflecting the strength of the highest frequency haar wavelets positioned at 16 different positions in time , corresponding to the wavelets w 17 to w 32 shown in fig3 . all the sum and difference values are scaled by multiplication by 2 \u2212 1 / 2 . next , taking the 16 sums of atrial signal potential values as an intermediate result , the above algorithm generates eight sums of and eight differences between adjacent pairs of these sixteen intermediate values that resulted from the first sixteen additions , and the eight newly - computed difference values become the haar transformed values 9 to 16 , reflecting the strength of the second to the highest frequency values at eight different points in time , corresponding to wavelets w 9 to w 16 in fig3 . all of these eight sum and eight difference values are again scaled by 2 \u2212 1 / 2 so that the wavelets w 9 to w 16 are scaled by \u00bd ( 2 \u2212 2 / 2 or 2 \u2212 1 / 2 times 2 \u2212 1 / 2 ). next , taking these eight sums of sums of adjacent atrial signal potential values as an intermediate result , the above algorithm generates four sum and four difference values , again scaling by 2 \u2212 1 / 2 , and the four difference values become haar transformed values 5 to 8 , reflecting the strength of the middle frequency wavelets at four different points in time , and corresponding to wavelets w 5 to w 8 in fig3 . next , taking these four sums of sums of sums of adjacent atrial signal potential values , the above algorithm generates two sum and two difference values , again scaling by 2 \u2212 1 / 2 , and the two difference values become haar transformed values 3 and 4 , reflecting the strength of the second to the lowest frequency wavelets only two of which encompass all the time domain data , corresponding to the wavelets w 3 and w 4 shown in fig3 . and finally , the algorithm generates the sum of and the difference between the final remaining two sums of sums of sums of sums of atrial signal potential values , again scaling by 2 \u2212 1 / 2 . the difference value is then the haar transformed value 2 , representing the strength of the lowest - frequency wavelet , the one corresponding to the wavelet w 2 in fig3 the wavelet that extends the full length of the time scale . the sum value is then the first , or d . c ., transformed value , representing the average signal potential level over the 32 sampled points in time , which corresponds to the wavelet w 1 in fig3 . another way of viewing this computation is illustrated in fig4 a and 4b . the 32 atrial signal potential values are shown at the top of fig4 a and are identified as c 0 5 through c 31 5 . the superscript \u201c 5 \u201d indicates that these values are processed when the index value \u201c n \u201d in the above computer program is equal to \u201c 5 \u201d\u2014 that is , during the first pass through the data generating intermediary sums and differences . during subsequent passes , the value of n is decremented to 4 , 3 , 2 , and finally to 1 . during each pass through the data , the computer generates sums c n - 1 and differences d n - 1 , as shown in fig4 b , between adjacent pairs of the values c 0 n and c 1 n ; c 2 n and c 3 n ; and so on , so that the number of newly - generated c n - 1 and d n - 1 terms is reduced by half with each computer pass through the intermediary results . at the end of all these computations , the value c 0 0 is the first , or d . c ., haar transformed value ; and the values d 0 0 ; d 0 1 and d 1 1 ; d 0 2 , d 1 2 , d 2 2 , and d 3 2 ; d 0 3 , d 1 3 , . . . and d 7 3 ; and d 0 4 , d 1 4 , and d 15 4 are , respectively , the remaining haar transform values 2 through 32 . fig4 thus illustrates quite succinctly how all the transform computations are carried out , and how the intermediary \u201c c \u201d sum values , such as c 0 4 , are used to compute multiple haar values , such as the values d 0 3 , d 0 2 , d 0 1 , d 0 0 , and c 0 0 all of which are computed from the intermediary value c 0 4 . saving and reusing these intermediary \u201c c \u201d values saves much computational time . fig4 b indicates the precise addition and subtraction operations that are carried out at each level to compute the values in the next lower level , proceeding down through the chart presented in fig4 a . but in any given application to heart waveform analysis , all of these computations may not be needed , and accordingly the number of computations may be reduced much further . since the present invention teaches that only a small number of these haar transformed values need actually be considered , a far less computationally intensive transform can be developed which only generates the intermediate and final transformed values that are actually needed to generate the specific haar transformed values which have proved to be significant in discriminating between sinus tachycardia ( or st ) and non - st conditions . all others need not be computed , and the above program may be reduced to a special algorithm that omits as many sums , differences , and multiplications as possible . for example , if only the transformed values 1 , 5 , 9 , and 24 are significant , then 31 additions are required to compute the first transformed value ( the d . c . value \u2014 the simple sum of all the time domain signal values ); six additions and one subtraction are required to compute the fifth transformed value ( in fig3 the difference between the sums of time domain values under each half of the wavelet w 5 at 314 in fig3 ); two additions and one subtraction are required to compute the ninth transformed value ( in fig3 the difference between the sums of the time domain values under each half of the wavelet w9 at 322 in fig3 ); and only one subtraction is required to compute the 24 th transformed value ( in fig3 the difference between the two time domain values under the respective halves of the wavelet w 24 ( not shown ) which is the same size as , but differently positioned in time than , the wavelet w 18 at 330 ( fig3 ). accordingly , if only the transformed values 1 , 5 , 9 , and 24 actually evaluated , then only 42 additions and subtractions are required . but even this number can be reduced further . if the computational algorithm set forth in the above illustrative computer program is followed as a guide , then the intermediary sums used in computing the haar first , or d . c ., transformed value can be re - used to compute the sums for the haar transformed values 5 and 9 , assuming these intermediary results are saved in the manner described in the above program example . then 31 additions are still required to compute the first transformed value , but only one subtraction is required to compute each of the transformed values 5 , 9 , and 24 , giving a total of additions and subtraction of only 34 operations . and if only transformed values 1 , 5 , and 9 are computed , the number of additions and subtractions is reduced to 33 . and if the first transformed value is omitted and if transformed values 5 , 9 , and 24 are selected , then the 31 additions needed to compute the first transformed value are not required . then the number of computations for transformed value 5 is 8 additions and one subtraction ; the number of computations for transformed value 9 is 2 additions and one subtraction ; and the number of computations for transformed value 24 is still just one subtraction . so the total number of additions and subtractions is just 13 . but even this number can be reduced to 11 when it is realized that the two additions done for the transformed value 9 are also done ( in the above computational algorithm ) when computing the transformed value 5 . so the number of additions and subtractions can be reduced to 11 . also , when computing such a small number of transformed values , the number of multiplications can be reduced as well by postponing the multiplications until a transformed value is actually computed . with reference to fig4 a , instead of dividing every sum and difference by the square root of two ( as shown in fig4 b ), several vertical sums of \u201c c \u201d values in different rows of the fig4 a table can be formed , and then a \u201c d \u201d value can be computed by subtraction , and then the resulting unscaled transformed value can be scaled with a single multiplication by 2 \u2212 j / 2 where j is the number of rows in the table of fig4 a traversed by the one or more sum and the one difference computations . thus , the number of scaling multiplications can sometimes be equal to the number of transformed values that are computed , typically 3 or 4 . accordingly , by using wavelet analysis and transformation , instead of fourier or discrete cosine or other sinusoid analysis and transformation , a highly useful and highly frequency - specific result can be achieved with a very small number of computations , thereby conserving power and battery life , and yet achieving a high degree of precision in recognizing changes in morphology . in our tests , we have selected certain transformed values as being much more significant than other values in distinguishing between st and non - st . we focused upon those transform values whose \u201c+ 1 \u201d and \u201c\u2212 1 \u201d scope included the middle 8 time points most significant to analysis of the peak of the atrial depolarization . working across test data samples obtained from patients who had dual chamber icds , we computed the variance of each transformed value for st and non - st events using the standard statistical formula for computing variance . we then calculated the ratio of the variance of non - st events to the variance of st events , and selected those terms with the highest variance ratios for further consideration . in this manner , we reduced the number of transformed values that were included in the cwa or correlation process while always maintaining or improving the performance achieved . ultimately we settled upon the three or four transformed values having the highest ratios . these were the transform values 1 , 5 , 9 , and 24 . we tested 1 , 5 , and 24 together ; 5 , 9 , and 24 together ; and 1 , 5 , 9 , and 24 together . these selected haar transform values , when used in these combinations , required minimal computations and thereby produced a savings in battery power , and that also gave better performance at distinguishing st from non - st than did all of the transformed values used together . so an increase in accuracy was achieved as well as a decrease in computational complexity by this approach to atrial waveform analysis . [ 0054 ] fig5 for example , illustrates actual plots of digital information illustrating the effect of using only the three coefficients 1 , 5 , and 24 . at 502 , a normal waveform is shown , with amplitude plotted against sample numbers from 1 to 32 . at 504 , an abnormal waveform is shown , again with amplitude plotted against sample numbers . after the haar transformation , at 506 and 508 , the wavelet coefficient amplitude values are indicated for the coefficients 1 to 32 . at 506 , the coefficients generated by the normal waveform 502 are shown , and at 508 , the coefficients generated by the abnormal waveform 504 are shown . one may directly compare these component values and verify that the coefficients 1 , 5 , and 24 at 506 and at 508 vary significantly in amplitude . comparison of this same information among different patients ( not shown in this figure ) also indicated that this variation is relatively constant from one patient to another . at 510 and 512 , using only the coefficients 1 , 5 , and 24 , the heartbeat waveforms are reconstructed by a reverse transformation , the normal reconstructed waveform shown at 510 and the abnormal reconstructed waveform shown at 512 . the marked differences between these two reconstructed waveforms highlights the way in which these three coefficients can signal an abnormal condition with less computation . in a practical system , a small number of transform values are selected , in the manner just described . the algorithm for computing these values is then refined , as explained above , to reduce the number of additions , subtractions , and multiplications to the minimum possible while preserving the accuracy of the computations . we first compute a value \u03c1 , which is the correlation between these values with the corresponding values in a template that represents the values for an average normal population . we then compare this value \u03c1 to a threshold correlation value \u03b2 that is chosen to give optimal results on experimental data . ( see the examples in the tables presented below .) for each waveform analyzed , a test is made to determine whether \u03c1 is greater than \u03b2 . if so , then this waveform is placed into the buffer marked \u201c st \u201d at step 210 . if not , then this waveform is placed into the buffer marked \u201c non - st \u201d at step 211 . finally , after ten waveforms have been analyzed , a count of st and non - st marked waveforms is made to see if the count of non - st waveforms is greater than some threshold value x ( at step 212 ), where x can be , for example , 7 . if more than seven waveforms are non - st , then therapy is delivered at step 214 . the buffers at steps 210 and 211 in fig2 are buffers that hold the last ten decision values . these buffers can be pictured as sliding windows revealing the most recent ten values to permit the decision at 212 to be continuously updated . the buffers thus function as a nonlinear filter preventing irregular values from delivering therapy improperly . in developing a working prototype of this system , we used a development data set consisting of 20 episodes of st taken from 4 patients and 18 episodes of af taken from 7 patients . patient number episodes of st episodes of af 1 2 6 2 12 0 13 0 1 4 0 2 5 5 1 6 0 3 7 0 3 8 0 2 9 1 0 total 20 18 we used this data to optimize our choice of \u03c1 and \u03b2 and also the particular coefficients that we chose to examine . next , we tested our prototype using a set of 16 episodes of st obtained from 4 patients together with 17 episodes of af obtained from 7 patients . patient number episodes of st episodes of af 1 8 0 2 3 1 3 3 0 4 2 0 5 0 4 6 0 4 7 0 3 8 0 3 9 0 1 10 0 1 total 16 17 x out threshold coefficients of 10 ( \u03b2 ) sensitivity specificity 1 , 5 , 9 , & amp ; 24 3 0 . 808 76 % 94 % 1 , 5 , 9 , & amp ; 24 5 0 . 975 88 % 94 % 1 , 5 , 9 , & amp ; 24 6 0 . 976 82 % 94 % 1 , 5 , 9 , & amp ; 24 7 0 . 983 88 % 94 % 1 , 5 , & amp ; 24 3 0 . 815 82 % 94 % 1 , 5 , & amp ; 24 4 0 . 943 88 % 94 % 1 , 5 , & amp ; 24 5 0 . 956 82 % 94 % 1 , 5 , & amp ; 24 7 0 . 991 94 % 94 % 5 , 9 , & amp ; 24 8 0 . 9993 82 % 94 % 5 , 9 , & amp ; 24 9 0 . 9997 76 % 81 % the quality of the selected subset of transformed values for characterizing changes in morphology may be demonstrated graphically , as indicated in fig5 by performing a reverse haar transform using only the three or four or so selected transformed values . as can be seen ( at 510 and 512 in fig5 ), the inverse transformations produce distorted representations of the original atrial waveforms which illustrate how the choice of transform values can emphasize the changes . this is another useful way to assist one in selecting which transformed values are most useful . one feature of the invention is its ability to use a template derived from an average normal population , rather than deriving a patient customized average normal template for each individual patient . prior systems have required each new patient to be monitored while in a normal state , and the captured data was then averaged to form a patient specific normal template . contrary to this , the present invention achieved the results shown above using the same template for all patients . accordingly , the invention may be used with a new patient without the necessity of such preliminary testing of each patient and without a new customized template necessarily having to be created for each patient . one reason why the present invention can function using a template that represents average values for a normal population is because the specific coefficients selected in the transform domain , in addition to having been selected to emphasize the difference between normal and abnormal values , are also preferably chosen to minimize the differences between different patients . in some cases , values that were good at distinguishing between normal and abnormal rhythms for one patient were not as good at doing so for some other patient . these values are preferably not selected for use in implementing the present invention . for example , the transform coefficients may be selected such that all ( or most of ) the normal values of a particular coefficient gathered from many patients had the same sign ( positive or negative ) and similar amplitudes ( or if they varied in sign , they were near to zero ). in addition , the abnormal coefficients are significantly different in value from the normal coefficients for each particular patient . [ 0068 ] fig6 at 600 , illustrates one way in which this may be done . first , at step 602 , one gathers a large number of st and non - st data samples . next , all of this data is transformed , generating coefficients for every waveform set of data ( step 604 ). the variance of each coefficient is then computed first for the st samples ( step 606 ) and then for the non - st samples . then , for each coefficient , the ratio of the variance of the non - st samples to that of the st samples is computed ( step 608 ). finally , those coefficients whose variance ratio was large may be selected a coefficients for use in creating a population template and in testing patients ( step 612 ). in addition , at step 614 , the number of coefficients retained may be further reduced to reduce the number of computations that need to be performed , as explained above . in the version of the invention that was used to generate the data shown above , the population template was computed as follows , using the 20 st episodes in the development data set ( also used in the first of the two tables presented above ). this is illustrated at 700 in fig7 . first , the coefficients were selected as was explained above and as shown in fig6 . next , for each patient episode of st ( step 702 ), the recorded heartbeats were broken up into groups of ten consecutive heartbeats ( step 704 ). for each group ( step 706 ), the selected coefficients of the haar transform were computed ( step 708 ) for each of the ten heartbeats and the median value was then selected ( step 710 ) for each coefficient to form a proto - template . if these median values correlated well with the coefficient values for each of the ten heartbeats , the proto - template was retained ( step 714 ). otherwise , it was discarded . in this manner , a whole bunch of proto - templates were obtained from each patient episode . next , median values of the coefficients from all of the proto - templates ( step 716 ) were computed ( step 718 ). these values were then used as the population template for distinguishing st from non - st events ( step 720 ). while the preferred embodiment of the invention has been described above , it is to be understood that numerous modifications and changes will occur to those who are skilled in the art to which the invention pertains . accordingly , the following claims annexed to and forming a part of this specification are intended to define the true spirit and scope of the invention \u2014 that is , what is new and what is desired to be secured by letters patent of the united states .", "category": "Human Necessities"}
|
{"patent": "the present invention is directed towards improving the quality and functionality of both implantable cardioverter defibrillators ( icds ) and implantable atrial defibrillators ( iads ). all such devices need to have a method and mechanism for accurately distinguishing sinus tachycardia ( st ) form non - st , such as atrial fibrillation and atrial flutter . this method and mechanism should be conservative , biased towards not initiating therapy unless non - st is clearly indicated . it should also require as few machine computational cycles as possible to reduce icd and iad battery drain . the principle underlying the present invention is that st and non - st can be distinguished by studying and measuring the atrial electrogram morphology to detect aberrant conduction in the atria . thus , by comparing normal st waveforms to a series of newly - occurring waveforms , st is indicated by abnormal atrial waveform morphology which signals aberrant conduction within the atria improper electrical patterns of depolarization . this will show up in the signal potentials captured by a bipolar atrial lead . with reference to fig1 the present invention can be implemented in an icd 100 ( which could be an iad ) that is implanted in the chest of a patient who suffers from unpleasant and possibly life - threatening irregularities in heart operation due to improper electrical stimulation of the heart muscle cells . during normal heart action , the heart &# 39 ; s electrical impulses originate in the sino - atrial node as an action potential that is transmitted smoothly to all portions of the atria , causing contraction of the atrial chambers . the electrical impulse continues in its path to a cluster of conduction fibrils known as the atrioventricular node . after a delay of about one - tenth of a second , an action potential flows over the ventricles and causes them to contract in synchronism with and following shortly after the atrial contraction . in this manner , the heart pumps blood to the lungs and from the lungs to the body . a bipolar lead 102 is implanted within the atria 104 of a heart 103 to measure the electrical potential between nearby cells of the atria . as the action potential actively passes from cell to cell past the two electrodes of the lead 102 , an oscillating signal potential is developed across the two electrodes and is conveyed by the bipolar lead 102 to the icd 100 where the signal is sampled about 1000 times per second and is digitized , with the sequential samples stored within a ram memory within the icd 100 for further analysis . another bipolar lead 105 may extend to the ventricle ( not shown ) of the heart 103 to measure the action potentials generated within the ventricles . these same leads , or other leads , may be used by the icd for administering various types of therapy , such as pacing or defibrillation shock therapy . the data samples collected from within the atria are now analyzed . first , with reference to data gathered from the ventricles , if some event in the ventricle , such as a mis - timed depolarization event that partially overlaps the atria &# 39 ; s p wave , could leak across to the atria and distort the measurement of the potentials for a given atrial depolarization , then the data for that particular heartbeat is discarded and is not used in further analyses . such distortion can also be found simply because a given heartbeat set of data gathered from the atria is itself badly distorted , and this approach must be taken in the case of an iad having no ventricular lead . the remaining data is retained for further processing . in preparation for further processing , the data for individual heartbeats is identified and gathered . the position within the gathered data of the negative spike to the atrial p depolarization waveform is determined and is selected as the center of the waveform for each beat . then , since data close to this p - wave is to be analyzed , and since data further away may be distorted to some degree by ventricular activity or by variations in the p - to - p interval , 32 sequential data samples are selected for each heartbeat such that the negative spike of the p - wave becomes signal potential sample number 16 of the 32 sequential samples selected for further analyses . these samples for a normal heartbeat appear at 302 in fig3 and at 502 in fig5 where the amplitude of the atrial signal potential is plotted against time over the 32 sampled values , with the negative p - wave spike positioned as signal sample number 16 . likewise , the samples for an abnormal heartbeat appear at 504 in fig5 . straightforward correlation ( cwa ) of the 32 samples against a template containing a typical or normal pattern could be performed at this juncture , as taught on page 563 ( equation ( 1 )) of the article by thorne , cited in the introductory portion of this specification . but the performance of such a cross correlation at repeated regular intervals would generate numerous computations and would drain the icd &# 39 ; s battery more rapidly than is desirable . in addition , provision would have to be made for a full template waveform that can be stored within the icd to be used as a comparison reference . in addition , such templates generally need to be patient specific , and they are more susceptible to normal changes in shape . accordingly , to reduce the mathematical complexity of the cross correlation computations , it is desirable to reduce the number of data points from 32 down to some much lower number through the use of some form of transformation into a different set of values . for example , the data could be transformed using the karhunen - loeve transformation described in the morris article cited at the beginning of this specification . but that transformation is fairly computationally intensive . computationally , a fast fourier transformation would be more efficient , but any such transformation into the frequency domain , where variable frequency sine and cosine wave amplitudes result from the transformation , does not match itself well to the morphology of atrial heart waveforms , which tend to be formed from large , ringing , spike - like representations of p - wave depolarization events travelling as moving action potential fronts past the pair of electrodes attached to the bipolar lead , and not as steady harmonics extending across time . accordingly , fourier - class transforms do not adequately reduce the number of significant data values that must be considered . and in addition , when the \u201c window \u201d size for a fourier analysis is selected , if it is wide , then the frequency values are not time - specific , but represent average signal harmonic content over the entire window time duration . and if the window is narrow , then the harmonics are too spread out in frequency , and the frequency - domain data generated by the transform does not resolve things finely enough in the frequency domain . for all of the above reasons , the present invention analyzes the waveforms using a discrete wavelet transform . this has the advantage that while low frequency wavelets are broad in time , high frequency wavelets have a much narrower timeslot focus . thus , for example , with 32 data samples taken sequentially over time , a wavelet transformation generates a d . c . wavelet that spans the entire set of 32 points ; a first reversing wavelet that also spans the entire set of points ; two second harmonic wavelets that each focus upon only one - half of the 32 data points ; four third harmonic wavelets that each focus upon only one - fourth of the 32 data points ; eight fourth harmonic wavelets that each focus upon four of the 32 data points ; and 16 fifth harmonic wavelets that each focuses upon only two of the 32 data points . accordingly , the higher - frequency wavelet amplitudes are quite time specific , unlike fourier sinusoidal amplitudes , while the lower - frequency and d . c . wavelet amplitudes give one broad information about the whole set of 32 points . and like the fourier transform , the discrete wavelet transform preserves all of the information of the original atrial signal , such that the transformation is fully reversible . 32 discrete wavelet transform values may be reverse transformed back into 32 values indicating the atrial signal potentials at 32 sequential points in time . in essence , the haar function is performing multiple digital filtering operations , at variable positions and utilizing varying - width windowing functions , to develop component values that represent varying types of heartbeat activity , at varying frequencies , spread over varying widths , and located at varying positions . these component values can be studied to see which subset of these component values are good at distinguishing st from non - st , or ( more generally ) which subset of these component values are good at distinguishing one type of heartbeat waveform from another . further study of such a selected subset can further determine which of the subset of components are relatively invariant from one individual &# 39 ; s heart to another . one preferably selects component values that are suitable from both of these two perspectives . the particular wavelet transformation to choose can be tailored to the nature of the data , with the wavelets chosen to resemble somewhat the values to be found within the data so that some transformed values are of much larger amplitude than others , and such that the low amplitude transformed values may be disregarded . of particular importance with an atrial p waveform of the type being analyzed here are wavelets having frequency values roughly comparable to and timed to coincide with the ringing of the atrial depolarization . this tailoring can minimize the number of transformed data values that are significant and that are candidates for full participation in the correlation analyses to determine if there has been a substantial change in waveform morphology . another factor in selecting a particular wavelet transformation is reducing the number of computations that must be performed to carry out the transformation . yet another factor is selecting a wavelet transformation where some of the transformed component values vary from normal to abnormal heart rhythms in much the same over a population of individuals so that a generic template of these component values may be derived that can be used with many individuals , rather than with just one individual . these factors suggest that the haar transformation would be a suitable candidate for use in analysis of atrial waveforms and possibly other waveforms as well . with reference to fig3 in the case of 32 voltage values sampled over time , the 32 applicable haar discrete transform wavelets appear as shown in this figure . other discrete wavelet transforms based upon wavelets having triangular or other shapes may also prove usable . the haar transform wavelets , in particular , have the shape of a square waveform , as will be described , and this can simplify the computations required . in fig3 time increases from left to right . a scale 304 indicates the 32 points in time at which the atrial waveform is sampled , and an analog atrial waveform ( before digitization ) is shown at 302 . the p wave negative notch 303 is shown centered at the 16 th timeslot so that the signal potential of the atria waveform sampled at this point in time becomes the 16 th data value in the set of 32 data values that are to be subjected to the haar transformation . a haar wavelet is simply one cycle of a square waveform that starts at zero , then swings positive ( to \u201c+ 1 \u201d), then swings negative ( to \u201c\u2212 1 \u201d), and then swings back to zero . the wavelet w 2 , shown at 308 in fig3 for example , is at \u201c+ 1 \u201d at points in time 1 to 16 and is at \u201c\u2212 1 \u201d at points in time 17 to 32 . the shorter waveform w 6 , shown at 316 in fig3 is at \u201c+ 1 \u201d at points in time 9 to 12 and is at \u201c\u2212 1 \u201d at points in time 13 to 16 , and is at \u201c 0 \u201d at all other points in time . the waveform w 1 , shown at 306 in fig3 is so slow to fluctuate in time that its \u201c\u2212 1 \u201d portion is off of the chart ( in fig3 ) to the right , and is ignored ; and accordingly , it has the value \u201c+ 1 \u201d for all 32 of the points in time shown in fig3 . it thus represents the average , or d . c ., component of the atrial signal potential . the lowest frequency wavelets w 1 and w 2 encompass the entire set of 32 points in time . the higher frequency wavelets w 3 and w 4 each encompasses only half of the points in time , and the two wavelets taken together encompass all the points in time . the four wavelets w 5 , w 6 , w 7 , and w 8 each encompasses only eight points in time , and the four wavelets taken together encompass all points in time . the eight wavelets w 9 . . . w 16 each encompasses only four points in time , and together all eight wavelets encompass all points in time . and finally , the wavelets w 17 . . . w 32 each encompass only two points in time , but the sixteen wavelets together encompass all points in time . thus , each wavelet has a characteristic time span and position as well as a characteristic frequency , with higher - frequency wavelets having a narrower and more specific time span and position than lower - frequency wavelets . this is advantageous with an impulse - type , ringing signal such as the atrial depolarization waveform considered here , since many higher - frequency transform values that correspond to haar wavelets positioned in time away from the p waveform or that do not correspond to its frequency of ringing may be low in amplitude such that they may safely be ignored during the analysis steps , thereby reducing the data that must be processed during the correlation steps . ( the above , while presented in the context of the haar wavelet , is also applicable to other wavelet shapes that may used to perform a discrete wavelet transform ( dwt )). each of the 32 haar transformed values is computed as follows : multiply each of the 32 sampled and digitized voltage values representing the analog atrial waveform 302 by the correspondingly - positioned -( in - time ) amplitude values of one of the 32 haar wavelets ( shown in fig3 ); then sum the resulting products ; and then multiply the resulting sums by a discrete wavelet transform scaling factor 2 \u2212 j / 2 , where j equals 1 for the wavelets w 17 to w 32 , 2 for w 9 to w 16 , 3 for w 5 to w 8 , 4 for w 3 to w 4 , and 5 for w 1 and w 2 . for example , and referring to fig3 : the first wavelet w 1 is always + 1 , so the atrial signal potential values are simply summed and then multiplied by 2 \u2212 5 / 2 ; the second wavelet is + 1 for time points 1 to 16 and \u2212 1 for time points 17 to 32 , so the first 16 values are summed , the second 16 values are summed , the difference between the first and second sums is computed , and the result is multiplied by 2 \u2212 5 / 2 ; and so on until all the transformed values have been computed , one for each haar wavelet . the 32 time - sequential atrial signal potential values are thus transformed into 32 haar wavelet amplitude values which may be called \u201c transformed values \u201d and which may be assigned the numbers 1 to 32 corresponding to the subscript numbers of the haar wavelets to which they each correspond and whose amplitudes they represent . the above description of how to compute the haar wavelet transformed values is accurate , but it is not the most efficient way to proceed in the case of haar wavelets . like the fourier transform , which has a corresponding fast fourier transform that saves intermediate results and thereby avoids re - computing them and thus reduces substantially the number of computations , the haar transformation also may be carried out in a manner that saves intermediate sums and re - uses them to reduce substantially the number of computations . this is described below at the point where fig4 is described , since fig4 illustrates graphically how this can be done . the illustrative program listing presented below is also an implementation of the fast haar wavelet transformation algorithm . the haar transformation can readily be carried out by any digital computer . as an example of how the haar transformation can be carried out on 32 values , the following program , written in the language c , is illustrative of many possible programs that may be written . in the exemplary program that follows , a 32 - element array yy contains 32 data values representing the 32 sampled atrial signal potential values representing the fluctuations over time of the signal supplied by the bipolar lead 102 . the analog atrial signal is sampled , digitized , broken into separate beat data sets , pre - processed ( to remove waveforms distorted by ventricular activity ), centered ( with the negative depolarization spike at data point 16 ), and fed into the subroutine presented below . this subroutine is compiled ( or assembled , if rewritten in assembly language ) and installed in the rom of the icd &# 39 ; s embedded microprocessor . this subroutine returns , contained within the same array yy , the 32 transformed haar wavelet amplitude values described above . ( the constant value \u201c recipsqrttwo \u201d is the reciprocal of the square root of two ). an illustrative version of the subroutine for computing all 32 of the haar transformed values in an efficient manner is presented here : void haar ( double yy []) { int i , j , l ; double zz [ 32 ]; for ( i = 5 ; i & gt ; 0 ; i \u2212\u2212) { l = 1 ; for ( j = 1 ; j & lt ;= i ; ++ j ) l = 2 * l ; for ( j = 0 ; j & lt ; l ; ++ j ) zz [ j ] = yy [ j ]; for ( j = 0 ; j & lt ; l \u2212 1 ; j = j + 2 ) ( yy [ j / 2 ] = recipsqrttwo * ( zz [ j ] + zz [ j + 1 ]); yy [( j + l )/ 2 ] = rcipsqrttwo * ( zz [ j ] \u2212 zz [ j + 1 ]); } } return ; } the above program computes the 32 haar transformed values 1 to 32 from the 32 atrial signal potential time - sequenced input values . it does so with only 31 additions , 31 subtractions , and 62 multiplications . it is carefully designed to compute each value in a systematic way , making multiple use of intermediate results . first , the program computes sixteen sums of and sixteen differences between eight adjacent pairs of the 16 time sequenced atrial signal potential values , and the sixteen difference values become the haar transformed values 17 to 32 , reflecting the strength of the highest frequency haar wavelets positioned at 16 different positions in time , corresponding to the wavelets w 17 to w 32 shown in fig3 . all the sum and difference values are scaled by multiplication by 2 \u2212 1 / 2 . next , taking the 16 sums of atrial signal potential values as an intermediate result , the above algorithm generates eight sums of and eight differences between adjacent pairs of these sixteen intermediate values that resulted from the first sixteen additions , and the eight newly - computed difference values become the haar transformed values 9 to 16 , reflecting the strength of the second to the highest frequency values at eight different points in time , corresponding to wavelets w 9 to w 16 in fig3 . all of these eight sum and eight difference values are again scaled by 2 \u2212 1 / 2 so that the wavelets w 9 to w 16 are scaled by \u00bd ( 2 \u2212 2 / 2 or 2 \u2212 1 / 2 times 2 \u2212 1 / 2 ). next , taking these eight sums of sums of adjacent atrial signal potential values as an intermediate result , the above algorithm generates four sum and four difference values , again scaling by 2 \u2212 1 / 2 , and the four difference values become haar transformed values 5 to 8 , reflecting the strength of the middle frequency wavelets at four different points in time , and corresponding to wavelets w 5 to w 8 in fig3 . next , taking these four sums of sums of sums of adjacent atrial signal potential values , the above algorithm generates two sum and two difference values , again scaling by 2 \u2212 1 / 2 , and the two difference values become haar transformed values 3 and 4 , reflecting the strength of the second to the lowest frequency wavelets only two of which encompass all the time domain data , corresponding to the wavelets w 3 and w 4 shown in fig3 . and finally , the algorithm generates the sum of and the difference between the final remaining two sums of sums of sums of sums of atrial signal potential values , again scaling by 2 \u2212 1 / 2 . the difference value is then the haar transformed value 2 , representing the strength of the lowest - frequency wavelet , the one corresponding to the wavelet w 2 in fig3 the wavelet that extends the full length of the time scale . the sum value is then the first , or d . c ., transformed value , representing the average signal potential level over the 32 sampled points in time , which corresponds to the wavelet w 1 in fig3 . another way of viewing this computation is illustrated in fig4 a and 4b . the 32 atrial signal potential values are shown at the top of fig4 a and are identified as c 0 5 through c 31 5 . the superscript \u201c 5 \u201d indicates that these values are processed when the index value \u201c n \u201d in the above computer program is equal to \u201c 5 \u201d\u2014 that is , during the first pass through the data generating intermediary sums and differences . during subsequent passes , the value of n is decremented to 4 , 3 , 2 , and finally to 1 . during each pass through the data , the computer generates sums c n - 1 and differences d n - 1 , as shown in fig4 b , between adjacent pairs of the values c 0 n and c 1 n ; c 2 n and c 3 n ; and so on , so that the number of newly - generated c n - 1 and d n - 1 terms is reduced by half with each computer pass through the intermediary results . at the end of all these computations , the value c 0 0 is the first , or d . c ., haar transformed value ; and the values d 0 0 ; d 0 1 and d 1 1 ; d 0 2 , d 1 2 , d 2 2 , and d 3 2 ; d 0 3 , d 1 3 , . . . and d 7 3 ; and d 0 4 , d 1 4 , and d 15 4 are , respectively , the remaining haar transform values 2 through 32 . fig4 thus illustrates quite succinctly how all the transform computations are carried out , and how the intermediary \u201c c \u201d sum values , such as c 0 4 , are used to compute multiple haar values , such as the values d 0 3 , d 0 2 , d 0 1 , d 0 0 , and c 0 0 all of which are computed from the intermediary value c 0 4 . saving and reusing these intermediary \u201c c \u201d values saves much computational time . fig4 b indicates the precise addition and subtraction operations that are carried out at each level to compute the values in the next lower level , proceeding down through the chart presented in fig4 a . but in any given application to heart waveform analysis , all of these computations may not be needed , and accordingly the number of computations may be reduced much further . since the present invention teaches that only a small number of these haar transformed values need actually be considered , a far less computationally intensive transform can be developed which only generates the intermediate and final transformed values that are actually needed to generate the specific haar transformed values which have proved to be significant in discriminating between sinus tachycardia ( or st ) and non - st conditions . all others need not be computed , and the above program may be reduced to a special algorithm that omits as many sums , differences , and multiplications as possible . for example , if only the transformed values 1 , 5 , 9 , and 24 are significant , then 31 additions are required to compute the first transformed value ( the d . c . value \u2014 the simple sum of all the time domain signal values ); six additions and one subtraction are required to compute the fifth transformed value ( in fig3 the difference between the sums of time domain values under each half of the wavelet w 5 at 314 in fig3 ); two additions and one subtraction are required to compute the ninth transformed value ( in fig3 the difference between the sums of the time domain values under each half of the wavelet w9 at 322 in fig3 ); and only one subtraction is required to compute the 24 th transformed value ( in fig3 the difference between the two time domain values under the respective halves of the wavelet w 24 ( not shown ) which is the same size as , but differently positioned in time than , the wavelet w 18 at 330 ( fig3 ). accordingly , if only the transformed values 1 , 5 , 9 , and 24 actually evaluated , then only 42 additions and subtractions are required . but even this number can be reduced further . if the computational algorithm set forth in the above illustrative computer program is followed as a guide , then the intermediary sums used in computing the haar first , or d . c ., transformed value can be re - used to compute the sums for the haar transformed values 5 and 9 , assuming these intermediary results are saved in the manner described in the above program example . then 31 additions are still required to compute the first transformed value , but only one subtraction is required to compute each of the transformed values 5 , 9 , and 24 , giving a total of additions and subtraction of only 34 operations . and if only transformed values 1 , 5 , and 9 are computed , the number of additions and subtractions is reduced to 33 . and if the first transformed value is omitted and if transformed values 5 , 9 , and 24 are selected , then the 31 additions needed to compute the first transformed value are not required . then the number of computations for transformed value 5 is 8 additions and one subtraction ; the number of computations for transformed value 9 is 2 additions and one subtraction ; and the number of computations for transformed value 24 is still just one subtraction . so the total number of additions and subtractions is just 13 . but even this number can be reduced to 11 when it is realized that the two additions done for the transformed value 9 are also done ( in the above computational algorithm ) when computing the transformed value 5 . so the number of additions and subtractions can be reduced to 11 . also , when computing such a small number of transformed values , the number of multiplications can be reduced as well by postponing the multiplications until a transformed value is actually computed . with reference to fig4 a , instead of dividing every sum and difference by the square root of two ( as shown in fig4 b ), several vertical sums of \u201c c \u201d values in different rows of the fig4 a table can be formed , and then a \u201c d \u201d value can be computed by subtraction , and then the resulting unscaled transformed value can be scaled with a single multiplication by 2 \u2212 j / 2 where j is the number of rows in the table of fig4 a traversed by the one or more sum and the one difference computations . thus , the number of scaling multiplications can sometimes be equal to the number of transformed values that are computed , typically 3 or 4 . accordingly , by using wavelet analysis and transformation , instead of fourier or discrete cosine or other sinusoid analysis and transformation , a highly useful and highly frequency - specific result can be achieved with a very small number of computations , thereby conserving power and battery life , and yet achieving a high degree of precision in recognizing changes in morphology . in our tests , we have selected certain transformed values as being much more significant than other values in distinguishing between st and non - st . we focused upon those transform values whose \u201c+ 1 \u201d and \u201c\u2212 1 \u201d scope included the middle 8 time points most significant to analysis of the peak of the atrial depolarization . working across test data samples obtained from patients who had dual chamber icds , we computed the variance of each transformed value for st and non - st events using the standard statistical formula for computing variance . we then calculated the ratio of the variance of non - st events to the variance of st events , and selected those terms with the highest variance ratios for further consideration . in this manner , we reduced the number of transformed values that were included in the cwa or correlation process while always maintaining or improving the performance achieved . ultimately we settled upon the three or four transformed values having the highest ratios . these were the transform values 1 , 5 , 9 , and 24 . we tested 1 , 5 , and 24 together ; 5 , 9 , and 24 together ; and 1 , 5 , 9 , and 24 together . these selected haar transform values , when used in these combinations , required minimal computations and thereby produced a savings in battery power , and that also gave better performance at distinguishing st from non - st than did all of the transformed values used together . so an increase in accuracy was achieved as well as a decrease in computational complexity by this approach to atrial waveform analysis . [ 0054 ] fig5 for example , illustrates actual plots of digital information illustrating the effect of using only the three coefficients 1 , 5 , and 24 . at 502 , a normal waveform is shown , with amplitude plotted against sample numbers from 1 to 32 . at 504 , an abnormal waveform is shown , again with amplitude plotted against sample numbers . after the haar transformation , at 506 and 508 , the wavelet coefficient amplitude values are indicated for the coefficients 1 to 32 . at 506 , the coefficients generated by the normal waveform 502 are shown , and at 508 , the coefficients generated by the abnormal waveform 504 are shown . one may directly compare these component values and verify that the coefficients 1 , 5 , and 24 at 506 and at 508 vary significantly in amplitude . comparison of this same information among different patients ( not shown in this figure ) also indicated that this variation is relatively constant from one patient to another . at 510 and 512 , using only the coefficients 1 , 5 , and 24 , the heartbeat waveforms are reconstructed by a reverse transformation , the normal reconstructed waveform shown at 510 and the abnormal reconstructed waveform shown at 512 . the marked differences between these two reconstructed waveforms highlights the way in which these three coefficients can signal an abnormal condition with less computation . in a practical system , a small number of transform values are selected , in the manner just described . the algorithm for computing these values is then refined , as explained above , to reduce the number of additions , subtractions , and multiplications to the minimum possible while preserving the accuracy of the computations . we first compute a value \u03c1 , which is the correlation between these values with the corresponding values in a template that represents the values for an average normal population . we then compare this value \u03c1 to a threshold correlation value \u03b2 that is chosen to give optimal results on experimental data . ( see the examples in the tables presented below .) for each waveform analyzed , a test is made to determine whether \u03c1 is greater than \u03b2 . if so , then this waveform is placed into the buffer marked \u201c st \u201d at step 210 . if not , then this waveform is placed into the buffer marked \u201c non - st \u201d at step 211 . finally , after ten waveforms have been analyzed , a count of st and non - st marked waveforms is made to see if the count of non - st waveforms is greater than some threshold value x ( at step 212 ), where x can be , for example , 7 . if more than seven waveforms are non - st , then therapy is delivered at step 214 . the buffers at steps 210 and 211 in fig2 are buffers that hold the last ten decision values . these buffers can be pictured as sliding windows revealing the most recent ten values to permit the decision at 212 to be continuously updated . the buffers thus function as a nonlinear filter preventing irregular values from delivering therapy improperly . in developing a working prototype of this system , we used a development data set consisting of 20 episodes of st taken from 4 patients and 18 episodes of af taken from 7 patients . patient number episodes of st episodes of af 1 2 6 2 12 0 13 0 1 4 0 2 5 5 1 6 0 3 7 0 3 8 0 2 9 1 0 total 20 18 we used this data to optimize our choice of \u03c1 and \u03b2 and also the particular coefficients that we chose to examine . next , we tested our prototype using a set of 16 episodes of st obtained from 4 patients together with 17 episodes of af obtained from 7 patients . patient number episodes of st episodes of af 1 8 0 2 3 1 3 3 0 4 2 0 5 0 4 6 0 4 7 0 3 8 0 3 9 0 1 10 0 1 total 16 17 x out threshold coefficients of 10 ( \u03b2 ) sensitivity specificity 1 , 5 , 9 , & amp ; 24 3 0 . 808 76 % 94 % 1 , 5 , 9 , & amp ; 24 5 0 . 975 88 % 94 % 1 , 5 , 9 , & amp ; 24 6 0 . 976 82 % 94 % 1 , 5 , 9 , & amp ; 24 7 0 . 983 88 % 94 % 1 , 5 , & amp ; 24 3 0 . 815 82 % 94 % 1 , 5 , & amp ; 24 4 0 . 943 88 % 94 % 1 , 5 , & amp ; 24 5 0 . 956 82 % 94 % 1 , 5 , & amp ; 24 7 0 . 991 94 % 94 % 5 , 9 , & amp ; 24 8 0 . 9993 82 % 94 % 5 , 9 , & amp ; 24 9 0 . 9997 76 % 81 % the quality of the selected subset of transformed values for characterizing changes in morphology may be demonstrated graphically , as indicated in fig5 by performing a reverse haar transform using only the three or four or so selected transformed values . as can be seen ( at 510 and 512 in fig5 ), the inverse transformations produce distorted representations of the original atrial waveforms which illustrate how the choice of transform values can emphasize the changes . this is another useful way to assist one in selecting which transformed values are most useful . one feature of the invention is its ability to use a template derived from an average normal population , rather than deriving a patient customized average normal template for each individual patient . prior systems have required each new patient to be monitored while in a normal state , and the captured data was then averaged to form a patient specific normal template . contrary to this , the present invention achieved the results shown above using the same template for all patients . accordingly , the invention may be used with a new patient without the necessity of such preliminary testing of each patient and without a new customized template necessarily having to be created for each patient . one reason why the present invention can function using a template that represents average values for a normal population is because the specific coefficients selected in the transform domain , in addition to having been selected to emphasize the difference between normal and abnormal values , are also preferably chosen to minimize the differences between different patients . in some cases , values that were good at distinguishing between normal and abnormal rhythms for one patient were not as good at doing so for some other patient . these values are preferably not selected for use in implementing the present invention . for example , the transform coefficients may be selected such that all ( or most of ) the normal values of a particular coefficient gathered from many patients had the same sign ( positive or negative ) and similar amplitudes ( or if they varied in sign , they were near to zero ). in addition , the abnormal coefficients are significantly different in value from the normal coefficients for each particular patient . [ 0068 ] fig6 at 600 , illustrates one way in which this may be done . first , at step 602 , one gathers a large number of st and non - st data samples . next , all of this data is transformed , generating coefficients for every waveform set of data ( step 604 ). the variance of each coefficient is then computed first for the st samples ( step 606 ) and then for the non - st samples . then , for each coefficient , the ratio of the variance of the non - st samples to that of the st samples is computed ( step 608 ). finally , those coefficients whose variance ratio was large may be selected a coefficients for use in creating a population template and in testing patients ( step 612 ). in addition , at step 614 , the number of coefficients retained may be further reduced to reduce the number of computations that need to be performed , as explained above . in the version of the invention that was used to generate the data shown above , the population template was computed as follows , using the 20 st episodes in the development data set ( also used in the first of the two tables presented above ). this is illustrated at 700 in fig7 . first , the coefficients were selected as was explained above and as shown in fig6 . next , for each patient episode of st ( step 702 ), the recorded heartbeats were broken up into groups of ten consecutive heartbeats ( step 704 ). for each group ( step 706 ), the selected coefficients of the haar transform were computed ( step 708 ) for each of the ten heartbeats and the median value was then selected ( step 710 ) for each coefficient to form a proto - template . if these median values correlated well with the coefficient values for each of the ten heartbeats , the proto - template was retained ( step 714 ). otherwise , it was discarded . in this manner , a whole bunch of proto - templates were obtained from each patient episode . next , median values of the coefficients from all of the proto - templates ( step 716 ) were computed ( step 718 ). these values were then used as the population template for distinguishing st from non - st events ( step 720 ). while the preferred embodiment of the invention has been described above , it is to be understood that numerous modifications and changes will occur to those who are skilled in the art to which the invention pertains . accordingly , the following claims annexed to and forming a part of this specification are intended to define the true spirit and scope of the invention \u2014 that is , what is new and what is desired to be secured by letters patent of the united states .", "category": "General tagging of new or cross-sectional technology"}
|
Does the patent belong in this category?
| 0.25 |
1cd5305994f1692ec7ec09368a2a725ff3c3dd15c21bf56c12e04e558cdb5fbc
| 0.112793 | 0.183594 | 0.777344 | 0.832031 | 0.597656 | 0.241211 |
null |
{"patent": "in fig1 is shown a preferred embodiment of a wave powered generator incorporating the subject invention . a central barge 12 is positioned between and in spaced parallel relationship to a pair of elongated floats 14 and 15 . the floats 14 and 15 are each connected to the central positioned barge by a pair of pins 16 , 17 and 18 , 19 which are pivotally connected at opposite ends by ball - type sockets 20 . by such connections the floats are maintained more or less in parallel relationship to the barge but are permitted to rise and fall , sway , rock and pitch relative to each other and within the constraints of the rigid pins 16 , 17 , 18 and 19 . it is this relative movement caused primarily by the wave and wind action acting on the floats and barge that is used to generate power . the relative motions are generally described in the previously identified patent application . each float 14 and 15 includes a mast structure 21 held in position by guy wires 22 and supporting a plurality of wind vanes 24 which are controlled in a manner to enhance the relative motion between the floats and the barge . the wind vanes are supported in a manner ( not shown ) so as to pivot between a position extending in a vertical plane so as to catch the wind and a horizontal position so as not to catch the wind . preferably these wind current vanes are each made of a rigid material which is strong enough to withstand the force of the wind and also will hold up in the sea spray environment . additionally there is provided a mechanism ( not shown ) for actuating the vanes on each float in unison much in the manner that a venetian blind is moved between open and closed positions . a wind pressing on the closed vanes , i . e . vertically positioned vanes , will cause the float to heel over and the opening of the vanes will allow the float to move back to the normal position thereby creating a rocking motion . a plurality of water current vanes 27 ( fig1 and 2 ) are provided under each float 14 and 15 . these vanes are supported on vertically extending supports 28 and are rotatable ( preferably in unison ) about a longitudinal axis . by specific control of the vanes ( in a manner not shown ) the vanes can be manipulated so as to cause a rocking motion of the floats and , by cycling such vane movement with the wind vane movement previously described , the floats can be caused to rock relative to the barge . to make use of this rocking motion in the generation of power , there are provided a plurality of elongated members or cables 30 , 31 , 32 and 34 having one end fixed to the various locations on the floats . the cables 32 and 34 extend diagonally upward and downward , respectively , to upper and lower pulleys 35 and 36 located at the ends of vertically extending watertight cylinders 37 in the barge . similarly the cables 30 and 31 extend substantially horizontally over pulleys 38 and 39 positioned at opposite ends of the cylinders 37 and 38 . the cylinders 37 and 38 each function in a similar manner such that only one will be described . however , pairs of such cylinders preferably are positioned at the corners of the float 12 and , if desires , more cylinders can be spaced along the barge . as shown primarily in fig3 wherein a pair of the cylinders are shown in cross - section , a piston 40 is supported therein by the cables 32 and 34 connected to opposite sides . the cables pass through watertight seals 41 at the end of the cylinder . at the lower end of the cylinder is a port 42 connecting to a water inlet 44 extending beneath the float . a one - way valve 45 allows water to be drawn through the inlet 44 and into the cylinder when the piston 40 is moved upward . a separate such inlet with one - way valve is connected with a port 46 positioned at the top of the cylinder ( but for simplification is not shown ). thus as the piston moves up and down the evacuated side of the cylinder is always refilled and maintained full by water drawn in from the sea . in the embodiment shown , upper movement of the piston 40 will force water out through the port 46 , down through the conduit 47 , up through the conduit 48 and into an accumulator 49 . this accumulator is a closed chamber having an air space 50 so that the pumping of pressured water therein pressurizes the air . subsequent downward movement of the piston 40 will force water through the horizontal conduit 51 and upwards through the conduit 48 into the accumulator . stop valves 52 and 54 in the conduits 51 and 47 , respectively , prevent a back flow of water when water is being forced out of the other end of the cylinder . with pressured water flowing into the accumulator , the air pressure buildup results in a constant flow of water through the outlet 55 in the accumulator and upwards through the conduit 56 to a turbine 57 which when rotated turns a shaft 58 . coupled to the shaft is a drive belt 59 leading to a power generator 60 . thus pressured water forced into the accumulator 49 will result in the turning of the turbine to subsequently drive the power generator . by use of the accumulator there is a buffet provided which in essence stores energy in the form of air pressure in the accumulator to supply a constant source of pressured fluid to the turbine . similarly the cylinder - piston combination at the other corner of the barge functions to supply pressure fluid to the accumulator . the cylinder 37a and the connections with the conduits are marked with a similar numbered prefix and the suffix &# 34 ; a &# 34 ; when the function is identical . thus the piston 40a is pulled up and down by the similar diagonal cables 32a and 34a . these cables pass over pulleys 35a and 36a which run through seals 41a and into the cylinder . as the piston 40a is forced up and down , water is drawn into the inlet 44a in the same manner as previously described and another inlet ( not shown ) for passage into the cylinder . the cylinder in turn forces fluid through the horizontal pipes 61 and 62 to pass into the same accumulator 49 in the same manner as previously described . thus there is a parallel flow of fluid into the accumulator as both floats move relative to the barge . similarly there are positioned in parallel to the cylinders 37 and 37a other cylinders connecting with the horizontal cables 30 and 31 , which cylinders function in the same manner to supply pressured water either to the accumulator 49 or to another accumulator ( not shown ). thus as described there is provided a pump and valve system for transmitting the fluid from the pump to the turbine for driving the generator 60 . by use of the closed pressure system , the energy input to the turbine is smoothed out so as not to be as cyclic as might occur otherwise due to the rocking motion of the floats . water is readily available from the surrounding medium and the power can be generated on a more or less constant basis assuming the presence of current and / or wind for moving the float . additionally any number of closed pressure systems can be positioned on the barge so long as there is physical space and flotation and these other systems . all the systems can be caused to turn turbines operating on the same shaft 58 , or turbines independently connected to other generators . in fig4 is shown another embodiment of the invention . shown therein is a barge 64 and float 65 . the barge and float are connected by a pin 66 extending lengthwise through a housing rigidly fixed to the adjacent side of the floating craft . this connecting arrangement allows the craft to pivot under action of the ocean current and wind current but prohibits relatife up and down motion . such an arrangement simplifies the relative motion and makes more rigid the connecting of the craft . cables 66 and 67 are connected between the barge and float in the same manner as described before , with the ends of the cables connected to opposite sides of a piston 40b to pump water to a system in the same manner as previously described .", "category": "Mechanical Engineering; Lightning; Heating; Weapons; Blasting"}
|
{"category": "Human Necessities", "patent": "in fig1 is shown a preferred embodiment of a wave powered generator incorporating the subject invention . a central barge 12 is positioned between and in spaced parallel relationship to a pair of elongated floats 14 and 15 . the floats 14 and 15 are each connected to the central positioned barge by a pair of pins 16 , 17 and 18 , 19 which are pivotally connected at opposite ends by ball - type sockets 20 . by such connections the floats are maintained more or less in parallel relationship to the barge but are permitted to rise and fall , sway , rock and pitch relative to each other and within the constraints of the rigid pins 16 , 17 , 18 and 19 . it is this relative movement caused primarily by the wave and wind action acting on the floats and barge that is used to generate power . the relative motions are generally described in the previously identified patent application . each float 14 and 15 includes a mast structure 21 held in position by guy wires 22 and supporting a plurality of wind vanes 24 which are controlled in a manner to enhance the relative motion between the floats and the barge . the wind vanes are supported in a manner ( not shown ) so as to pivot between a position extending in a vertical plane so as to catch the wind and a horizontal position so as not to catch the wind . preferably these wind current vanes are each made of a rigid material which is strong enough to withstand the force of the wind and also will hold up in the sea spray environment . additionally there is provided a mechanism ( not shown ) for actuating the vanes on each float in unison much in the manner that a venetian blind is moved between open and closed positions . a wind pressing on the closed vanes , i . e . vertically positioned vanes , will cause the float to heel over and the opening of the vanes will allow the float to move back to the normal position thereby creating a rocking motion . a plurality of water current vanes 27 ( fig1 and 2 ) are provided under each float 14 and 15 . these vanes are supported on vertically extending supports 28 and are rotatable ( preferably in unison ) about a longitudinal axis . by specific control of the vanes ( in a manner not shown ) the vanes can be manipulated so as to cause a rocking motion of the floats and , by cycling such vane movement with the wind vane movement previously described , the floats can be caused to rock relative to the barge . to make use of this rocking motion in the generation of power , there are provided a plurality of elongated members or cables 30 , 31 , 32 and 34 having one end fixed to the various locations on the floats . the cables 32 and 34 extend diagonally upward and downward , respectively , to upper and lower pulleys 35 and 36 located at the ends of vertically extending watertight cylinders 37 in the barge . similarly the cables 30 and 31 extend substantially horizontally over pulleys 38 and 39 positioned at opposite ends of the cylinders 37 and 38 . the cylinders 37 and 38 each function in a similar manner such that only one will be described . however , pairs of such cylinders preferably are positioned at the corners of the float 12 and , if desires , more cylinders can be spaced along the barge . as shown primarily in fig3 wherein a pair of the cylinders are shown in cross - section , a piston 40 is supported therein by the cables 32 and 34 connected to opposite sides . the cables pass through watertight seals 41 at the end of the cylinder . at the lower end of the cylinder is a port 42 connecting to a water inlet 44 extending beneath the float . a one - way valve 45 allows water to be drawn through the inlet 44 and into the cylinder when the piston 40 is moved upward . a separate such inlet with one - way valve is connected with a port 46 positioned at the top of the cylinder ( but for simplification is not shown ). thus as the piston moves up and down the evacuated side of the cylinder is always refilled and maintained full by water drawn in from the sea . in the embodiment shown , upper movement of the piston 40 will force water out through the port 46 , down through the conduit 47 , up through the conduit 48 and into an accumulator 49 . this accumulator is a closed chamber having an air space 50 so that the pumping of pressured water therein pressurizes the air . subsequent downward movement of the piston 40 will force water through the horizontal conduit 51 and upwards through the conduit 48 into the accumulator . stop valves 52 and 54 in the conduits 51 and 47 , respectively , prevent a back flow of water when water is being forced out of the other end of the cylinder . with pressured water flowing into the accumulator , the air pressure buildup results in a constant flow of water through the outlet 55 in the accumulator and upwards through the conduit 56 to a turbine 57 which when rotated turns a shaft 58 . coupled to the shaft is a drive belt 59 leading to a power generator 60 . thus pressured water forced into the accumulator 49 will result in the turning of the turbine to subsequently drive the power generator . by use of the accumulator there is a buffet provided which in essence stores energy in the form of air pressure in the accumulator to supply a constant source of pressured fluid to the turbine . similarly the cylinder - piston combination at the other corner of the barge functions to supply pressure fluid to the accumulator . the cylinder 37a and the connections with the conduits are marked with a similar numbered prefix and the suffix &# 34 ; a &# 34 ; when the function is identical . thus the piston 40a is pulled up and down by the similar diagonal cables 32a and 34a . these cables pass over pulleys 35a and 36a which run through seals 41a and into the cylinder . as the piston 40a is forced up and down , water is drawn into the inlet 44a in the same manner as previously described and another inlet ( not shown ) for passage into the cylinder . the cylinder in turn forces fluid through the horizontal pipes 61 and 62 to pass into the same accumulator 49 in the same manner as previously described . thus there is a parallel flow of fluid into the accumulator as both floats move relative to the barge . similarly there are positioned in parallel to the cylinders 37 and 37a other cylinders connecting with the horizontal cables 30 and 31 , which cylinders function in the same manner to supply pressured water either to the accumulator 49 or to another accumulator ( not shown ). thus as described there is provided a pump and valve system for transmitting the fluid from the pump to the turbine for driving the generator 60 . by use of the closed pressure system , the energy input to the turbine is smoothed out so as not to be as cyclic as might occur otherwise due to the rocking motion of the floats . water is readily available from the surrounding medium and the power can be generated on a more or less constant basis assuming the presence of current and / or wind for moving the float . additionally any number of closed pressure systems can be positioned on the barge so long as there is physical space and flotation and these other systems . all the systems can be caused to turn turbines operating on the same shaft 58 , or turbines independently connected to other generators . in fig4 is shown another embodiment of the invention . shown therein is a barge 64 and float 65 . the barge and float are connected by a pin 66 extending lengthwise through a housing rigidly fixed to the adjacent side of the floating craft . this connecting arrangement allows the craft to pivot under action of the ocean current and wind current but prohibits relatife up and down motion . such an arrangement simplifies the relative motion and makes more rigid the connecting of the craft . cables 66 and 67 are connected between the barge and float in the same manner as described before , with the ends of the cables connected to opposite sides of a piston 40b to pump water to a system in the same manner as previously described ."}
|
Is the categorization of this patent accurate?
| 0.25 |
c84b2d917423930170182b72df89909a565c510d41712fe2673b0852d7275aab
| 0.00383 | 0.006897 | 0.019409 | 0.016357 | 0.064453 | 0.004913 |
null |
{"category": "Mechanical Engineering; Lightning; Heating; Weapons; Blasting", "patent": "in fig1 is shown a preferred embodiment of a wave powered generator incorporating the subject invention . a central barge 12 is positioned between and in spaced parallel relationship to a pair of elongated floats 14 and 15 . the floats 14 and 15 are each connected to the central positioned barge by a pair of pins 16 , 17 and 18 , 19 which are pivotally connected at opposite ends by ball - type sockets 20 . by such connections the floats are maintained more or less in parallel relationship to the barge but are permitted to rise and fall , sway , rock and pitch relative to each other and within the constraints of the rigid pins 16 , 17 , 18 and 19 . it is this relative movement caused primarily by the wave and wind action acting on the floats and barge that is used to generate power . the relative motions are generally described in the previously identified patent application . each float 14 and 15 includes a mast structure 21 held in position by guy wires 22 and supporting a plurality of wind vanes 24 which are controlled in a manner to enhance the relative motion between the floats and the barge . the wind vanes are supported in a manner ( not shown ) so as to pivot between a position extending in a vertical plane so as to catch the wind and a horizontal position so as not to catch the wind . preferably these wind current vanes are each made of a rigid material which is strong enough to withstand the force of the wind and also will hold up in the sea spray environment . additionally there is provided a mechanism ( not shown ) for actuating the vanes on each float in unison much in the manner that a venetian blind is moved between open and closed positions . a wind pressing on the closed vanes , i . e . vertically positioned vanes , will cause the float to heel over and the opening of the vanes will allow the float to move back to the normal position thereby creating a rocking motion . a plurality of water current vanes 27 ( fig1 and 2 ) are provided under each float 14 and 15 . these vanes are supported on vertically extending supports 28 and are rotatable ( preferably in unison ) about a longitudinal axis . by specific control of the vanes ( in a manner not shown ) the vanes can be manipulated so as to cause a rocking motion of the floats and , by cycling such vane movement with the wind vane movement previously described , the floats can be caused to rock relative to the barge . to make use of this rocking motion in the generation of power , there are provided a plurality of elongated members or cables 30 , 31 , 32 and 34 having one end fixed to the various locations on the floats . the cables 32 and 34 extend diagonally upward and downward , respectively , to upper and lower pulleys 35 and 36 located at the ends of vertically extending watertight cylinders 37 in the barge . similarly the cables 30 and 31 extend substantially horizontally over pulleys 38 and 39 positioned at opposite ends of the cylinders 37 and 38 . the cylinders 37 and 38 each function in a similar manner such that only one will be described . however , pairs of such cylinders preferably are positioned at the corners of the float 12 and , if desires , more cylinders can be spaced along the barge . as shown primarily in fig3 wherein a pair of the cylinders are shown in cross - section , a piston 40 is supported therein by the cables 32 and 34 connected to opposite sides . the cables pass through watertight seals 41 at the end of the cylinder . at the lower end of the cylinder is a port 42 connecting to a water inlet 44 extending beneath the float . a one - way valve 45 allows water to be drawn through the inlet 44 and into the cylinder when the piston 40 is moved upward . a separate such inlet with one - way valve is connected with a port 46 positioned at the top of the cylinder ( but for simplification is not shown ). thus as the piston moves up and down the evacuated side of the cylinder is always refilled and maintained full by water drawn in from the sea . in the embodiment shown , upper movement of the piston 40 will force water out through the port 46 , down through the conduit 47 , up through the conduit 48 and into an accumulator 49 . this accumulator is a closed chamber having an air space 50 so that the pumping of pressured water therein pressurizes the air . subsequent downward movement of the piston 40 will force water through the horizontal conduit 51 and upwards through the conduit 48 into the accumulator . stop valves 52 and 54 in the conduits 51 and 47 , respectively , prevent a back flow of water when water is being forced out of the other end of the cylinder . with pressured water flowing into the accumulator , the air pressure buildup results in a constant flow of water through the outlet 55 in the accumulator and upwards through the conduit 56 to a turbine 57 which when rotated turns a shaft 58 . coupled to the shaft is a drive belt 59 leading to a power generator 60 . thus pressured water forced into the accumulator 49 will result in the turning of the turbine to subsequently drive the power generator . by use of the accumulator there is a buffet provided which in essence stores energy in the form of air pressure in the accumulator to supply a constant source of pressured fluid to the turbine . similarly the cylinder - piston combination at the other corner of the barge functions to supply pressure fluid to the accumulator . the cylinder 37a and the connections with the conduits are marked with a similar numbered prefix and the suffix &# 34 ; a &# 34 ; when the function is identical . thus the piston 40a is pulled up and down by the similar diagonal cables 32a and 34a . these cables pass over pulleys 35a and 36a which run through seals 41a and into the cylinder . as the piston 40a is forced up and down , water is drawn into the inlet 44a in the same manner as previously described and another inlet ( not shown ) for passage into the cylinder . the cylinder in turn forces fluid through the horizontal pipes 61 and 62 to pass into the same accumulator 49 in the same manner as previously described . thus there is a parallel flow of fluid into the accumulator as both floats move relative to the barge . similarly there are positioned in parallel to the cylinders 37 and 37a other cylinders connecting with the horizontal cables 30 and 31 , which cylinders function in the same manner to supply pressured water either to the accumulator 49 or to another accumulator ( not shown ). thus as described there is provided a pump and valve system for transmitting the fluid from the pump to the turbine for driving the generator 60 . by use of the closed pressure system , the energy input to the turbine is smoothed out so as not to be as cyclic as might occur otherwise due to the rocking motion of the floats . water is readily available from the surrounding medium and the power can be generated on a more or less constant basis assuming the presence of current and / or wind for moving the float . additionally any number of closed pressure systems can be positioned on the barge so long as there is physical space and flotation and these other systems . all the systems can be caused to turn turbines operating on the same shaft 58 , or turbines independently connected to other generators . in fig4 is shown another embodiment of the invention . shown therein is a barge 64 and float 65 . the barge and float are connected by a pin 66 extending lengthwise through a housing rigidly fixed to the adjacent side of the floating craft . this connecting arrangement allows the craft to pivot under action of the ocean current and wind current but prohibits relatife up and down motion . such an arrangement simplifies the relative motion and makes more rigid the connecting of the craft . cables 66 and 67 are connected between the barge and float in the same manner as described before , with the ends of the cables connected to opposite sides of a piston 40b to pump water to a system in the same manner as previously described ."}
|
{"patent": "in fig1 is shown a preferred embodiment of a wave powered generator incorporating the subject invention . a central barge 12 is positioned between and in spaced parallel relationship to a pair of elongated floats 14 and 15 . the floats 14 and 15 are each connected to the central positioned barge by a pair of pins 16 , 17 and 18 , 19 which are pivotally connected at opposite ends by ball - type sockets 20 . by such connections the floats are maintained more or less in parallel relationship to the barge but are permitted to rise and fall , sway , rock and pitch relative to each other and within the constraints of the rigid pins 16 , 17 , 18 and 19 . it is this relative movement caused primarily by the wave and wind action acting on the floats and barge that is used to generate power . the relative motions are generally described in the previously identified patent application . each float 14 and 15 includes a mast structure 21 held in position by guy wires 22 and supporting a plurality of wind vanes 24 which are controlled in a manner to enhance the relative motion between the floats and the barge . the wind vanes are supported in a manner ( not shown ) so as to pivot between a position extending in a vertical plane so as to catch the wind and a horizontal position so as not to catch the wind . preferably these wind current vanes are each made of a rigid material which is strong enough to withstand the force of the wind and also will hold up in the sea spray environment . additionally there is provided a mechanism ( not shown ) for actuating the vanes on each float in unison much in the manner that a venetian blind is moved between open and closed positions . a wind pressing on the closed vanes , i . e . vertically positioned vanes , will cause the float to heel over and the opening of the vanes will allow the float to move back to the normal position thereby creating a rocking motion . a plurality of water current vanes 27 ( fig1 and 2 ) are provided under each float 14 and 15 . these vanes are supported on vertically extending supports 28 and are rotatable ( preferably in unison ) about a longitudinal axis . by specific control of the vanes ( in a manner not shown ) the vanes can be manipulated so as to cause a rocking motion of the floats and , by cycling such vane movement with the wind vane movement previously described , the floats can be caused to rock relative to the barge . to make use of this rocking motion in the generation of power , there are provided a plurality of elongated members or cables 30 , 31 , 32 and 34 having one end fixed to the various locations on the floats . the cables 32 and 34 extend diagonally upward and downward , respectively , to upper and lower pulleys 35 and 36 located at the ends of vertically extending watertight cylinders 37 in the barge . similarly the cables 30 and 31 extend substantially horizontally over pulleys 38 and 39 positioned at opposite ends of the cylinders 37 and 38 . the cylinders 37 and 38 each function in a similar manner such that only one will be described . however , pairs of such cylinders preferably are positioned at the corners of the float 12 and , if desires , more cylinders can be spaced along the barge . as shown primarily in fig3 wherein a pair of the cylinders are shown in cross - section , a piston 40 is supported therein by the cables 32 and 34 connected to opposite sides . the cables pass through watertight seals 41 at the end of the cylinder . at the lower end of the cylinder is a port 42 connecting to a water inlet 44 extending beneath the float . a one - way valve 45 allows water to be drawn through the inlet 44 and into the cylinder when the piston 40 is moved upward . a separate such inlet with one - way valve is connected with a port 46 positioned at the top of the cylinder ( but for simplification is not shown ). thus as the piston moves up and down the evacuated side of the cylinder is always refilled and maintained full by water drawn in from the sea . in the embodiment shown , upper movement of the piston 40 will force water out through the port 46 , down through the conduit 47 , up through the conduit 48 and into an accumulator 49 . this accumulator is a closed chamber having an air space 50 so that the pumping of pressured water therein pressurizes the air . subsequent downward movement of the piston 40 will force water through the horizontal conduit 51 and upwards through the conduit 48 into the accumulator . stop valves 52 and 54 in the conduits 51 and 47 , respectively , prevent a back flow of water when water is being forced out of the other end of the cylinder . with pressured water flowing into the accumulator , the air pressure buildup results in a constant flow of water through the outlet 55 in the accumulator and upwards through the conduit 56 to a turbine 57 which when rotated turns a shaft 58 . coupled to the shaft is a drive belt 59 leading to a power generator 60 . thus pressured water forced into the accumulator 49 will result in the turning of the turbine to subsequently drive the power generator . by use of the accumulator there is a buffet provided which in essence stores energy in the form of air pressure in the accumulator to supply a constant source of pressured fluid to the turbine . similarly the cylinder - piston combination at the other corner of the barge functions to supply pressure fluid to the accumulator . the cylinder 37a and the connections with the conduits are marked with a similar numbered prefix and the suffix &# 34 ; a &# 34 ; when the function is identical . thus the piston 40a is pulled up and down by the similar diagonal cables 32a and 34a . these cables pass over pulleys 35a and 36a which run through seals 41a and into the cylinder . as the piston 40a is forced up and down , water is drawn into the inlet 44a in the same manner as previously described and another inlet ( not shown ) for passage into the cylinder . the cylinder in turn forces fluid through the horizontal pipes 61 and 62 to pass into the same accumulator 49 in the same manner as previously described . thus there is a parallel flow of fluid into the accumulator as both floats move relative to the barge . similarly there are positioned in parallel to the cylinders 37 and 37a other cylinders connecting with the horizontal cables 30 and 31 , which cylinders function in the same manner to supply pressured water either to the accumulator 49 or to another accumulator ( not shown ). thus as described there is provided a pump and valve system for transmitting the fluid from the pump to the turbine for driving the generator 60 . by use of the closed pressure system , the energy input to the turbine is smoothed out so as not to be as cyclic as might occur otherwise due to the rocking motion of the floats . water is readily available from the surrounding medium and the power can be generated on a more or less constant basis assuming the presence of current and / or wind for moving the float . additionally any number of closed pressure systems can be positioned on the barge so long as there is physical space and flotation and these other systems . all the systems can be caused to turn turbines operating on the same shaft 58 , or turbines independently connected to other generators . in fig4 is shown another embodiment of the invention . shown therein is a barge 64 and float 65 . the barge and float are connected by a pin 66 extending lengthwise through a housing rigidly fixed to the adjacent side of the floating craft . this connecting arrangement allows the craft to pivot under action of the ocean current and wind current but prohibits relatife up and down motion . such an arrangement simplifies the relative motion and makes more rigid the connecting of the craft . cables 66 and 67 are connected between the barge and float in the same manner as described before , with the ends of the cables connected to opposite sides of a piston 40b to pump water to a system in the same manner as previously described .", "category": "Performing Operations; Transporting"}
|
Does the category match the content of the patent?
| 0.25 |
c84b2d917423930170182b72df89909a565c510d41712fe2673b0852d7275aab
| 0.012024 | 0.04541 | 0.014038 | 0.090332 | 0.075684 | 0.102539 |
null |
{"category": "Mechanical Engineering; Lightning; Heating; Weapons; Blasting", "patent": "in fig1 is shown a preferred embodiment of a wave powered generator incorporating the subject invention . a central barge 12 is positioned between and in spaced parallel relationship to a pair of elongated floats 14 and 15 . the floats 14 and 15 are each connected to the central positioned barge by a pair of pins 16 , 17 and 18 , 19 which are pivotally connected at opposite ends by ball - type sockets 20 . by such connections the floats are maintained more or less in parallel relationship to the barge but are permitted to rise and fall , sway , rock and pitch relative to each other and within the constraints of the rigid pins 16 , 17 , 18 and 19 . it is this relative movement caused primarily by the wave and wind action acting on the floats and barge that is used to generate power . the relative motions are generally described in the previously identified patent application . each float 14 and 15 includes a mast structure 21 held in position by guy wires 22 and supporting a plurality of wind vanes 24 which are controlled in a manner to enhance the relative motion between the floats and the barge . the wind vanes are supported in a manner ( not shown ) so as to pivot between a position extending in a vertical plane so as to catch the wind and a horizontal position so as not to catch the wind . preferably these wind current vanes are each made of a rigid material which is strong enough to withstand the force of the wind and also will hold up in the sea spray environment . additionally there is provided a mechanism ( not shown ) for actuating the vanes on each float in unison much in the manner that a venetian blind is moved between open and closed positions . a wind pressing on the closed vanes , i . e . vertically positioned vanes , will cause the float to heel over and the opening of the vanes will allow the float to move back to the normal position thereby creating a rocking motion . a plurality of water current vanes 27 ( fig1 and 2 ) are provided under each float 14 and 15 . these vanes are supported on vertically extending supports 28 and are rotatable ( preferably in unison ) about a longitudinal axis . by specific control of the vanes ( in a manner not shown ) the vanes can be manipulated so as to cause a rocking motion of the floats and , by cycling such vane movement with the wind vane movement previously described , the floats can be caused to rock relative to the barge . to make use of this rocking motion in the generation of power , there are provided a plurality of elongated members or cables 30 , 31 , 32 and 34 having one end fixed to the various locations on the floats . the cables 32 and 34 extend diagonally upward and downward , respectively , to upper and lower pulleys 35 and 36 located at the ends of vertically extending watertight cylinders 37 in the barge . similarly the cables 30 and 31 extend substantially horizontally over pulleys 38 and 39 positioned at opposite ends of the cylinders 37 and 38 . the cylinders 37 and 38 each function in a similar manner such that only one will be described . however , pairs of such cylinders preferably are positioned at the corners of the float 12 and , if desires , more cylinders can be spaced along the barge . as shown primarily in fig3 wherein a pair of the cylinders are shown in cross - section , a piston 40 is supported therein by the cables 32 and 34 connected to opposite sides . the cables pass through watertight seals 41 at the end of the cylinder . at the lower end of the cylinder is a port 42 connecting to a water inlet 44 extending beneath the float . a one - way valve 45 allows water to be drawn through the inlet 44 and into the cylinder when the piston 40 is moved upward . a separate such inlet with one - way valve is connected with a port 46 positioned at the top of the cylinder ( but for simplification is not shown ). thus as the piston moves up and down the evacuated side of the cylinder is always refilled and maintained full by water drawn in from the sea . in the embodiment shown , upper movement of the piston 40 will force water out through the port 46 , down through the conduit 47 , up through the conduit 48 and into an accumulator 49 . this accumulator is a closed chamber having an air space 50 so that the pumping of pressured water therein pressurizes the air . subsequent downward movement of the piston 40 will force water through the horizontal conduit 51 and upwards through the conduit 48 into the accumulator . stop valves 52 and 54 in the conduits 51 and 47 , respectively , prevent a back flow of water when water is being forced out of the other end of the cylinder . with pressured water flowing into the accumulator , the air pressure buildup results in a constant flow of water through the outlet 55 in the accumulator and upwards through the conduit 56 to a turbine 57 which when rotated turns a shaft 58 . coupled to the shaft is a drive belt 59 leading to a power generator 60 . thus pressured water forced into the accumulator 49 will result in the turning of the turbine to subsequently drive the power generator . by use of the accumulator there is a buffet provided which in essence stores energy in the form of air pressure in the accumulator to supply a constant source of pressured fluid to the turbine . similarly the cylinder - piston combination at the other corner of the barge functions to supply pressure fluid to the accumulator . the cylinder 37a and the connections with the conduits are marked with a similar numbered prefix and the suffix &# 34 ; a &# 34 ; when the function is identical . thus the piston 40a is pulled up and down by the similar diagonal cables 32a and 34a . these cables pass over pulleys 35a and 36a which run through seals 41a and into the cylinder . as the piston 40a is forced up and down , water is drawn into the inlet 44a in the same manner as previously described and another inlet ( not shown ) for passage into the cylinder . the cylinder in turn forces fluid through the horizontal pipes 61 and 62 to pass into the same accumulator 49 in the same manner as previously described . thus there is a parallel flow of fluid into the accumulator as both floats move relative to the barge . similarly there are positioned in parallel to the cylinders 37 and 37a other cylinders connecting with the horizontal cables 30 and 31 , which cylinders function in the same manner to supply pressured water either to the accumulator 49 or to another accumulator ( not shown ). thus as described there is provided a pump and valve system for transmitting the fluid from the pump to the turbine for driving the generator 60 . by use of the closed pressure system , the energy input to the turbine is smoothed out so as not to be as cyclic as might occur otherwise due to the rocking motion of the floats . water is readily available from the surrounding medium and the power can be generated on a more or less constant basis assuming the presence of current and / or wind for moving the float . additionally any number of closed pressure systems can be positioned on the barge so long as there is physical space and flotation and these other systems . all the systems can be caused to turn turbines operating on the same shaft 58 , or turbines independently connected to other generators . in fig4 is shown another embodiment of the invention . shown therein is a barge 64 and float 65 . the barge and float are connected by a pin 66 extending lengthwise through a housing rigidly fixed to the adjacent side of the floating craft . this connecting arrangement allows the craft to pivot under action of the ocean current and wind current but prohibits relatife up and down motion . such an arrangement simplifies the relative motion and makes more rigid the connecting of the craft . cables 66 and 67 are connected between the barge and float in the same manner as described before , with the ends of the cables connected to opposite sides of a piston 40b to pump water to a system in the same manner as previously described ."}
|
{"category": "Chemistry; Metallurgy", "patent": "in fig1 is shown a preferred embodiment of a wave powered generator incorporating the subject invention . a central barge 12 is positioned between and in spaced parallel relationship to a pair of elongated floats 14 and 15 . the floats 14 and 15 are each connected to the central positioned barge by a pair of pins 16 , 17 and 18 , 19 which are pivotally connected at opposite ends by ball - type sockets 20 . by such connections the floats are maintained more or less in parallel relationship to the barge but are permitted to rise and fall , sway , rock and pitch relative to each other and within the constraints of the rigid pins 16 , 17 , 18 and 19 . it is this relative movement caused primarily by the wave and wind action acting on the floats and barge that is used to generate power . the relative motions are generally described in the previously identified patent application . each float 14 and 15 includes a mast structure 21 held in position by guy wires 22 and supporting a plurality of wind vanes 24 which are controlled in a manner to enhance the relative motion between the floats and the barge . the wind vanes are supported in a manner ( not shown ) so as to pivot between a position extending in a vertical plane so as to catch the wind and a horizontal position so as not to catch the wind . preferably these wind current vanes are each made of a rigid material which is strong enough to withstand the force of the wind and also will hold up in the sea spray environment . additionally there is provided a mechanism ( not shown ) for actuating the vanes on each float in unison much in the manner that a venetian blind is moved between open and closed positions . a wind pressing on the closed vanes , i . e . vertically positioned vanes , will cause the float to heel over and the opening of the vanes will allow the float to move back to the normal position thereby creating a rocking motion . a plurality of water current vanes 27 ( fig1 and 2 ) are provided under each float 14 and 15 . these vanes are supported on vertically extending supports 28 and are rotatable ( preferably in unison ) about a longitudinal axis . by specific control of the vanes ( in a manner not shown ) the vanes can be manipulated so as to cause a rocking motion of the floats and , by cycling such vane movement with the wind vane movement previously described , the floats can be caused to rock relative to the barge . to make use of this rocking motion in the generation of power , there are provided a plurality of elongated members or cables 30 , 31 , 32 and 34 having one end fixed to the various locations on the floats . the cables 32 and 34 extend diagonally upward and downward , respectively , to upper and lower pulleys 35 and 36 located at the ends of vertically extending watertight cylinders 37 in the barge . similarly the cables 30 and 31 extend substantially horizontally over pulleys 38 and 39 positioned at opposite ends of the cylinders 37 and 38 . the cylinders 37 and 38 each function in a similar manner such that only one will be described . however , pairs of such cylinders preferably are positioned at the corners of the float 12 and , if desires , more cylinders can be spaced along the barge . as shown primarily in fig3 wherein a pair of the cylinders are shown in cross - section , a piston 40 is supported therein by the cables 32 and 34 connected to opposite sides . the cables pass through watertight seals 41 at the end of the cylinder . at the lower end of the cylinder is a port 42 connecting to a water inlet 44 extending beneath the float . a one - way valve 45 allows water to be drawn through the inlet 44 and into the cylinder when the piston 40 is moved upward . a separate such inlet with one - way valve is connected with a port 46 positioned at the top of the cylinder ( but for simplification is not shown ). thus as the piston moves up and down the evacuated side of the cylinder is always refilled and maintained full by water drawn in from the sea . in the embodiment shown , upper movement of the piston 40 will force water out through the port 46 , down through the conduit 47 , up through the conduit 48 and into an accumulator 49 . this accumulator is a closed chamber having an air space 50 so that the pumping of pressured water therein pressurizes the air . subsequent downward movement of the piston 40 will force water through the horizontal conduit 51 and upwards through the conduit 48 into the accumulator . stop valves 52 and 54 in the conduits 51 and 47 , respectively , prevent a back flow of water when water is being forced out of the other end of the cylinder . with pressured water flowing into the accumulator , the air pressure buildup results in a constant flow of water through the outlet 55 in the accumulator and upwards through the conduit 56 to a turbine 57 which when rotated turns a shaft 58 . coupled to the shaft is a drive belt 59 leading to a power generator 60 . thus pressured water forced into the accumulator 49 will result in the turning of the turbine to subsequently drive the power generator . by use of the accumulator there is a buffet provided which in essence stores energy in the form of air pressure in the accumulator to supply a constant source of pressured fluid to the turbine . similarly the cylinder - piston combination at the other corner of the barge functions to supply pressure fluid to the accumulator . the cylinder 37a and the connections with the conduits are marked with a similar numbered prefix and the suffix &# 34 ; a &# 34 ; when the function is identical . thus the piston 40a is pulled up and down by the similar diagonal cables 32a and 34a . these cables pass over pulleys 35a and 36a which run through seals 41a and into the cylinder . as the piston 40a is forced up and down , water is drawn into the inlet 44a in the same manner as previously described and another inlet ( not shown ) for passage into the cylinder . the cylinder in turn forces fluid through the horizontal pipes 61 and 62 to pass into the same accumulator 49 in the same manner as previously described . thus there is a parallel flow of fluid into the accumulator as both floats move relative to the barge . similarly there are positioned in parallel to the cylinders 37 and 37a other cylinders connecting with the horizontal cables 30 and 31 , which cylinders function in the same manner to supply pressured water either to the accumulator 49 or to another accumulator ( not shown ). thus as described there is provided a pump and valve system for transmitting the fluid from the pump to the turbine for driving the generator 60 . by use of the closed pressure system , the energy input to the turbine is smoothed out so as not to be as cyclic as might occur otherwise due to the rocking motion of the floats . water is readily available from the surrounding medium and the power can be generated on a more or less constant basis assuming the presence of current and / or wind for moving the float . additionally any number of closed pressure systems can be positioned on the barge so long as there is physical space and flotation and these other systems . all the systems can be caused to turn turbines operating on the same shaft 58 , or turbines independently connected to other generators . in fig4 is shown another embodiment of the invention . shown therein is a barge 64 and float 65 . the barge and float are connected by a pin 66 extending lengthwise through a housing rigidly fixed to the adjacent side of the floating craft . this connecting arrangement allows the craft to pivot under action of the ocean current and wind current but prohibits relatife up and down motion . such an arrangement simplifies the relative motion and makes more rigid the connecting of the craft . cables 66 and 67 are connected between the barge and float in the same manner as described before , with the ends of the cables connected to opposite sides of a piston 40b to pump water to a system in the same manner as previously described ."}
|
Is the categorization of this patent accurate?
| 0.25 |
c84b2d917423930170182b72df89909a565c510d41712fe2673b0852d7275aab
| 0.014038 | 0.012451 | 0.009155 | 0.009399 | 0.119141 | 0.006287 |
null |
{"category": "Mechanical Engineering; Lightning; Heating; Weapons; Blasting", "patent": "in fig1 is shown a preferred embodiment of a wave powered generator incorporating the subject invention . a central barge 12 is positioned between and in spaced parallel relationship to a pair of elongated floats 14 and 15 . the floats 14 and 15 are each connected to the central positioned barge by a pair of pins 16 , 17 and 18 , 19 which are pivotally connected at opposite ends by ball - type sockets 20 . by such connections the floats are maintained more or less in parallel relationship to the barge but are permitted to rise and fall , sway , rock and pitch relative to each other and within the constraints of the rigid pins 16 , 17 , 18 and 19 . it is this relative movement caused primarily by the wave and wind action acting on the floats and barge that is used to generate power . the relative motions are generally described in the previously identified patent application . each float 14 and 15 includes a mast structure 21 held in position by guy wires 22 and supporting a plurality of wind vanes 24 which are controlled in a manner to enhance the relative motion between the floats and the barge . the wind vanes are supported in a manner ( not shown ) so as to pivot between a position extending in a vertical plane so as to catch the wind and a horizontal position so as not to catch the wind . preferably these wind current vanes are each made of a rigid material which is strong enough to withstand the force of the wind and also will hold up in the sea spray environment . additionally there is provided a mechanism ( not shown ) for actuating the vanes on each float in unison much in the manner that a venetian blind is moved between open and closed positions . a wind pressing on the closed vanes , i . e . vertically positioned vanes , will cause the float to heel over and the opening of the vanes will allow the float to move back to the normal position thereby creating a rocking motion . a plurality of water current vanes 27 ( fig1 and 2 ) are provided under each float 14 and 15 . these vanes are supported on vertically extending supports 28 and are rotatable ( preferably in unison ) about a longitudinal axis . by specific control of the vanes ( in a manner not shown ) the vanes can be manipulated so as to cause a rocking motion of the floats and , by cycling such vane movement with the wind vane movement previously described , the floats can be caused to rock relative to the barge . to make use of this rocking motion in the generation of power , there are provided a plurality of elongated members or cables 30 , 31 , 32 and 34 having one end fixed to the various locations on the floats . the cables 32 and 34 extend diagonally upward and downward , respectively , to upper and lower pulleys 35 and 36 located at the ends of vertically extending watertight cylinders 37 in the barge . similarly the cables 30 and 31 extend substantially horizontally over pulleys 38 and 39 positioned at opposite ends of the cylinders 37 and 38 . the cylinders 37 and 38 each function in a similar manner such that only one will be described . however , pairs of such cylinders preferably are positioned at the corners of the float 12 and , if desires , more cylinders can be spaced along the barge . as shown primarily in fig3 wherein a pair of the cylinders are shown in cross - section , a piston 40 is supported therein by the cables 32 and 34 connected to opposite sides . the cables pass through watertight seals 41 at the end of the cylinder . at the lower end of the cylinder is a port 42 connecting to a water inlet 44 extending beneath the float . a one - way valve 45 allows water to be drawn through the inlet 44 and into the cylinder when the piston 40 is moved upward . a separate such inlet with one - way valve is connected with a port 46 positioned at the top of the cylinder ( but for simplification is not shown ). thus as the piston moves up and down the evacuated side of the cylinder is always refilled and maintained full by water drawn in from the sea . in the embodiment shown , upper movement of the piston 40 will force water out through the port 46 , down through the conduit 47 , up through the conduit 48 and into an accumulator 49 . this accumulator is a closed chamber having an air space 50 so that the pumping of pressured water therein pressurizes the air . subsequent downward movement of the piston 40 will force water through the horizontal conduit 51 and upwards through the conduit 48 into the accumulator . stop valves 52 and 54 in the conduits 51 and 47 , respectively , prevent a back flow of water when water is being forced out of the other end of the cylinder . with pressured water flowing into the accumulator , the air pressure buildup results in a constant flow of water through the outlet 55 in the accumulator and upwards through the conduit 56 to a turbine 57 which when rotated turns a shaft 58 . coupled to the shaft is a drive belt 59 leading to a power generator 60 . thus pressured water forced into the accumulator 49 will result in the turning of the turbine to subsequently drive the power generator . by use of the accumulator there is a buffet provided which in essence stores energy in the form of air pressure in the accumulator to supply a constant source of pressured fluid to the turbine . similarly the cylinder - piston combination at the other corner of the barge functions to supply pressure fluid to the accumulator . the cylinder 37a and the connections with the conduits are marked with a similar numbered prefix and the suffix &# 34 ; a &# 34 ; when the function is identical . thus the piston 40a is pulled up and down by the similar diagonal cables 32a and 34a . these cables pass over pulleys 35a and 36a which run through seals 41a and into the cylinder . as the piston 40a is forced up and down , water is drawn into the inlet 44a in the same manner as previously described and another inlet ( not shown ) for passage into the cylinder . the cylinder in turn forces fluid through the horizontal pipes 61 and 62 to pass into the same accumulator 49 in the same manner as previously described . thus there is a parallel flow of fluid into the accumulator as both floats move relative to the barge . similarly there are positioned in parallel to the cylinders 37 and 37a other cylinders connecting with the horizontal cables 30 and 31 , which cylinders function in the same manner to supply pressured water either to the accumulator 49 or to another accumulator ( not shown ). thus as described there is provided a pump and valve system for transmitting the fluid from the pump to the turbine for driving the generator 60 . by use of the closed pressure system , the energy input to the turbine is smoothed out so as not to be as cyclic as might occur otherwise due to the rocking motion of the floats . water is readily available from the surrounding medium and the power can be generated on a more or less constant basis assuming the presence of current and / or wind for moving the float . additionally any number of closed pressure systems can be positioned on the barge so long as there is physical space and flotation and these other systems . all the systems can be caused to turn turbines operating on the same shaft 58 , or turbines independently connected to other generators . in fig4 is shown another embodiment of the invention . shown therein is a barge 64 and float 65 . the barge and float are connected by a pin 66 extending lengthwise through a housing rigidly fixed to the adjacent side of the floating craft . this connecting arrangement allows the craft to pivot under action of the ocean current and wind current but prohibits relatife up and down motion . such an arrangement simplifies the relative motion and makes more rigid the connecting of the craft . cables 66 and 67 are connected between the barge and float in the same manner as described before , with the ends of the cables connected to opposite sides of a piston 40b to pump water to a system in the same manner as previously described ."}
|
{"category": "Textiles; Paper", "patent": "in fig1 is shown a preferred embodiment of a wave powered generator incorporating the subject invention . a central barge 12 is positioned between and in spaced parallel relationship to a pair of elongated floats 14 and 15 . the floats 14 and 15 are each connected to the central positioned barge by a pair of pins 16 , 17 and 18 , 19 which are pivotally connected at opposite ends by ball - type sockets 20 . by such connections the floats are maintained more or less in parallel relationship to the barge but are permitted to rise and fall , sway , rock and pitch relative to each other and within the constraints of the rigid pins 16 , 17 , 18 and 19 . it is this relative movement caused primarily by the wave and wind action acting on the floats and barge that is used to generate power . the relative motions are generally described in the previously identified patent application . each float 14 and 15 includes a mast structure 21 held in position by guy wires 22 and supporting a plurality of wind vanes 24 which are controlled in a manner to enhance the relative motion between the floats and the barge . the wind vanes are supported in a manner ( not shown ) so as to pivot between a position extending in a vertical plane so as to catch the wind and a horizontal position so as not to catch the wind . preferably these wind current vanes are each made of a rigid material which is strong enough to withstand the force of the wind and also will hold up in the sea spray environment . additionally there is provided a mechanism ( not shown ) for actuating the vanes on each float in unison much in the manner that a venetian blind is moved between open and closed positions . a wind pressing on the closed vanes , i . e . vertically positioned vanes , will cause the float to heel over and the opening of the vanes will allow the float to move back to the normal position thereby creating a rocking motion . a plurality of water current vanes 27 ( fig1 and 2 ) are provided under each float 14 and 15 . these vanes are supported on vertically extending supports 28 and are rotatable ( preferably in unison ) about a longitudinal axis . by specific control of the vanes ( in a manner not shown ) the vanes can be manipulated so as to cause a rocking motion of the floats and , by cycling such vane movement with the wind vane movement previously described , the floats can be caused to rock relative to the barge . to make use of this rocking motion in the generation of power , there are provided a plurality of elongated members or cables 30 , 31 , 32 and 34 having one end fixed to the various locations on the floats . the cables 32 and 34 extend diagonally upward and downward , respectively , to upper and lower pulleys 35 and 36 located at the ends of vertically extending watertight cylinders 37 in the barge . similarly the cables 30 and 31 extend substantially horizontally over pulleys 38 and 39 positioned at opposite ends of the cylinders 37 and 38 . the cylinders 37 and 38 each function in a similar manner such that only one will be described . however , pairs of such cylinders preferably are positioned at the corners of the float 12 and , if desires , more cylinders can be spaced along the barge . as shown primarily in fig3 wherein a pair of the cylinders are shown in cross - section , a piston 40 is supported therein by the cables 32 and 34 connected to opposite sides . the cables pass through watertight seals 41 at the end of the cylinder . at the lower end of the cylinder is a port 42 connecting to a water inlet 44 extending beneath the float . a one - way valve 45 allows water to be drawn through the inlet 44 and into the cylinder when the piston 40 is moved upward . a separate such inlet with one - way valve is connected with a port 46 positioned at the top of the cylinder ( but for simplification is not shown ). thus as the piston moves up and down the evacuated side of the cylinder is always refilled and maintained full by water drawn in from the sea . in the embodiment shown , upper movement of the piston 40 will force water out through the port 46 , down through the conduit 47 , up through the conduit 48 and into an accumulator 49 . this accumulator is a closed chamber having an air space 50 so that the pumping of pressured water therein pressurizes the air . subsequent downward movement of the piston 40 will force water through the horizontal conduit 51 and upwards through the conduit 48 into the accumulator . stop valves 52 and 54 in the conduits 51 and 47 , respectively , prevent a back flow of water when water is being forced out of the other end of the cylinder . with pressured water flowing into the accumulator , the air pressure buildup results in a constant flow of water through the outlet 55 in the accumulator and upwards through the conduit 56 to a turbine 57 which when rotated turns a shaft 58 . coupled to the shaft is a drive belt 59 leading to a power generator 60 . thus pressured water forced into the accumulator 49 will result in the turning of the turbine to subsequently drive the power generator . by use of the accumulator there is a buffet provided which in essence stores energy in the form of air pressure in the accumulator to supply a constant source of pressured fluid to the turbine . similarly the cylinder - piston combination at the other corner of the barge functions to supply pressure fluid to the accumulator . the cylinder 37a and the connections with the conduits are marked with a similar numbered prefix and the suffix &# 34 ; a &# 34 ; when the function is identical . thus the piston 40a is pulled up and down by the similar diagonal cables 32a and 34a . these cables pass over pulleys 35a and 36a which run through seals 41a and into the cylinder . as the piston 40a is forced up and down , water is drawn into the inlet 44a in the same manner as previously described and another inlet ( not shown ) for passage into the cylinder . the cylinder in turn forces fluid through the horizontal pipes 61 and 62 to pass into the same accumulator 49 in the same manner as previously described . thus there is a parallel flow of fluid into the accumulator as both floats move relative to the barge . similarly there are positioned in parallel to the cylinders 37 and 37a other cylinders connecting with the horizontal cables 30 and 31 , which cylinders function in the same manner to supply pressured water either to the accumulator 49 or to another accumulator ( not shown ). thus as described there is provided a pump and valve system for transmitting the fluid from the pump to the turbine for driving the generator 60 . by use of the closed pressure system , the energy input to the turbine is smoothed out so as not to be as cyclic as might occur otherwise due to the rocking motion of the floats . water is readily available from the surrounding medium and the power can be generated on a more or less constant basis assuming the presence of current and / or wind for moving the float . additionally any number of closed pressure systems can be positioned on the barge so long as there is physical space and flotation and these other systems . all the systems can be caused to turn turbines operating on the same shaft 58 , or turbines independently connected to other generators . in fig4 is shown another embodiment of the invention . shown therein is a barge 64 and float 65 . the barge and float are connected by a pin 66 extending lengthwise through a housing rigidly fixed to the adjacent side of the floating craft . this connecting arrangement allows the craft to pivot under action of the ocean current and wind current but prohibits relatife up and down motion . such an arrangement simplifies the relative motion and makes more rigid the connecting of the craft . cables 66 and 67 are connected between the barge and float in the same manner as described before , with the ends of the cables connected to opposite sides of a piston 40b to pump water to a system in the same manner as previously described ."}
|
Is the patent correctly categorized?
| 0.25 |
c84b2d917423930170182b72df89909a565c510d41712fe2673b0852d7275aab
| 0.020996 | 0.029297 | 0.013245 | 0.002625 | 0.376953 | 0.147461 |
null |
{"category": "Mechanical Engineering; Lightning; Heating; Weapons; Blasting", "patent": "in fig1 is shown a preferred embodiment of a wave powered generator incorporating the subject invention . a central barge 12 is positioned between and in spaced parallel relationship to a pair of elongated floats 14 and 15 . the floats 14 and 15 are each connected to the central positioned barge by a pair of pins 16 , 17 and 18 , 19 which are pivotally connected at opposite ends by ball - type sockets 20 . by such connections the floats are maintained more or less in parallel relationship to the barge but are permitted to rise and fall , sway , rock and pitch relative to each other and within the constraints of the rigid pins 16 , 17 , 18 and 19 . it is this relative movement caused primarily by the wave and wind action acting on the floats and barge that is used to generate power . the relative motions are generally described in the previously identified patent application . each float 14 and 15 includes a mast structure 21 held in position by guy wires 22 and supporting a plurality of wind vanes 24 which are controlled in a manner to enhance the relative motion between the floats and the barge . the wind vanes are supported in a manner ( not shown ) so as to pivot between a position extending in a vertical plane so as to catch the wind and a horizontal position so as not to catch the wind . preferably these wind current vanes are each made of a rigid material which is strong enough to withstand the force of the wind and also will hold up in the sea spray environment . additionally there is provided a mechanism ( not shown ) for actuating the vanes on each float in unison much in the manner that a venetian blind is moved between open and closed positions . a wind pressing on the closed vanes , i . e . vertically positioned vanes , will cause the float to heel over and the opening of the vanes will allow the float to move back to the normal position thereby creating a rocking motion . a plurality of water current vanes 27 ( fig1 and 2 ) are provided under each float 14 and 15 . these vanes are supported on vertically extending supports 28 and are rotatable ( preferably in unison ) about a longitudinal axis . by specific control of the vanes ( in a manner not shown ) the vanes can be manipulated so as to cause a rocking motion of the floats and , by cycling such vane movement with the wind vane movement previously described , the floats can be caused to rock relative to the barge . to make use of this rocking motion in the generation of power , there are provided a plurality of elongated members or cables 30 , 31 , 32 and 34 having one end fixed to the various locations on the floats . the cables 32 and 34 extend diagonally upward and downward , respectively , to upper and lower pulleys 35 and 36 located at the ends of vertically extending watertight cylinders 37 in the barge . similarly the cables 30 and 31 extend substantially horizontally over pulleys 38 and 39 positioned at opposite ends of the cylinders 37 and 38 . the cylinders 37 and 38 each function in a similar manner such that only one will be described . however , pairs of such cylinders preferably are positioned at the corners of the float 12 and , if desires , more cylinders can be spaced along the barge . as shown primarily in fig3 wherein a pair of the cylinders are shown in cross - section , a piston 40 is supported therein by the cables 32 and 34 connected to opposite sides . the cables pass through watertight seals 41 at the end of the cylinder . at the lower end of the cylinder is a port 42 connecting to a water inlet 44 extending beneath the float . a one - way valve 45 allows water to be drawn through the inlet 44 and into the cylinder when the piston 40 is moved upward . a separate such inlet with one - way valve is connected with a port 46 positioned at the top of the cylinder ( but for simplification is not shown ). thus as the piston moves up and down the evacuated side of the cylinder is always refilled and maintained full by water drawn in from the sea . in the embodiment shown , upper movement of the piston 40 will force water out through the port 46 , down through the conduit 47 , up through the conduit 48 and into an accumulator 49 . this accumulator is a closed chamber having an air space 50 so that the pumping of pressured water therein pressurizes the air . subsequent downward movement of the piston 40 will force water through the horizontal conduit 51 and upwards through the conduit 48 into the accumulator . stop valves 52 and 54 in the conduits 51 and 47 , respectively , prevent a back flow of water when water is being forced out of the other end of the cylinder . with pressured water flowing into the accumulator , the air pressure buildup results in a constant flow of water through the outlet 55 in the accumulator and upwards through the conduit 56 to a turbine 57 which when rotated turns a shaft 58 . coupled to the shaft is a drive belt 59 leading to a power generator 60 . thus pressured water forced into the accumulator 49 will result in the turning of the turbine to subsequently drive the power generator . by use of the accumulator there is a buffet provided which in essence stores energy in the form of air pressure in the accumulator to supply a constant source of pressured fluid to the turbine . similarly the cylinder - piston combination at the other corner of the barge functions to supply pressure fluid to the accumulator . the cylinder 37a and the connections with the conduits are marked with a similar numbered prefix and the suffix &# 34 ; a &# 34 ; when the function is identical . thus the piston 40a is pulled up and down by the similar diagonal cables 32a and 34a . these cables pass over pulleys 35a and 36a which run through seals 41a and into the cylinder . as the piston 40a is forced up and down , water is drawn into the inlet 44a in the same manner as previously described and another inlet ( not shown ) for passage into the cylinder . the cylinder in turn forces fluid through the horizontal pipes 61 and 62 to pass into the same accumulator 49 in the same manner as previously described . thus there is a parallel flow of fluid into the accumulator as both floats move relative to the barge . similarly there are positioned in parallel to the cylinders 37 and 37a other cylinders connecting with the horizontal cables 30 and 31 , which cylinders function in the same manner to supply pressured water either to the accumulator 49 or to another accumulator ( not shown ). thus as described there is provided a pump and valve system for transmitting the fluid from the pump to the turbine for driving the generator 60 . by use of the closed pressure system , the energy input to the turbine is smoothed out so as not to be as cyclic as might occur otherwise due to the rocking motion of the floats . water is readily available from the surrounding medium and the power can be generated on a more or less constant basis assuming the presence of current and / or wind for moving the float . additionally any number of closed pressure systems can be positioned on the barge so long as there is physical space and flotation and these other systems . all the systems can be caused to turn turbines operating on the same shaft 58 , or turbines independently connected to other generators . in fig4 is shown another embodiment of the invention . shown therein is a barge 64 and float 65 . the barge and float are connected by a pin 66 extending lengthwise through a housing rigidly fixed to the adjacent side of the floating craft . this connecting arrangement allows the craft to pivot under action of the ocean current and wind current but prohibits relatife up and down motion . such an arrangement simplifies the relative motion and makes more rigid the connecting of the craft . cables 66 and 67 are connected between the barge and float in the same manner as described before , with the ends of the cables connected to opposite sides of a piston 40b to pump water to a system in the same manner as previously described ."}
|
{"patent": "in fig1 is shown a preferred embodiment of a wave powered generator incorporating the subject invention . a central barge 12 is positioned between and in spaced parallel relationship to a pair of elongated floats 14 and 15 . the floats 14 and 15 are each connected to the central positioned barge by a pair of pins 16 , 17 and 18 , 19 which are pivotally connected at opposite ends by ball - type sockets 20 . by such connections the floats are maintained more or less in parallel relationship to the barge but are permitted to rise and fall , sway , rock and pitch relative to each other and within the constraints of the rigid pins 16 , 17 , 18 and 19 . it is this relative movement caused primarily by the wave and wind action acting on the floats and barge that is used to generate power . the relative motions are generally described in the previously identified patent application . each float 14 and 15 includes a mast structure 21 held in position by guy wires 22 and supporting a plurality of wind vanes 24 which are controlled in a manner to enhance the relative motion between the floats and the barge . the wind vanes are supported in a manner ( not shown ) so as to pivot between a position extending in a vertical plane so as to catch the wind and a horizontal position so as not to catch the wind . preferably these wind current vanes are each made of a rigid material which is strong enough to withstand the force of the wind and also will hold up in the sea spray environment . additionally there is provided a mechanism ( not shown ) for actuating the vanes on each float in unison much in the manner that a venetian blind is moved between open and closed positions . a wind pressing on the closed vanes , i . e . vertically positioned vanes , will cause the float to heel over and the opening of the vanes will allow the float to move back to the normal position thereby creating a rocking motion . a plurality of water current vanes 27 ( fig1 and 2 ) are provided under each float 14 and 15 . these vanes are supported on vertically extending supports 28 and are rotatable ( preferably in unison ) about a longitudinal axis . by specific control of the vanes ( in a manner not shown ) the vanes can be manipulated so as to cause a rocking motion of the floats and , by cycling such vane movement with the wind vane movement previously described , the floats can be caused to rock relative to the barge . to make use of this rocking motion in the generation of power , there are provided a plurality of elongated members or cables 30 , 31 , 32 and 34 having one end fixed to the various locations on the floats . the cables 32 and 34 extend diagonally upward and downward , respectively , to upper and lower pulleys 35 and 36 located at the ends of vertically extending watertight cylinders 37 in the barge . similarly the cables 30 and 31 extend substantially horizontally over pulleys 38 and 39 positioned at opposite ends of the cylinders 37 and 38 . the cylinders 37 and 38 each function in a similar manner such that only one will be described . however , pairs of such cylinders preferably are positioned at the corners of the float 12 and , if desires , more cylinders can be spaced along the barge . as shown primarily in fig3 wherein a pair of the cylinders are shown in cross - section , a piston 40 is supported therein by the cables 32 and 34 connected to opposite sides . the cables pass through watertight seals 41 at the end of the cylinder . at the lower end of the cylinder is a port 42 connecting to a water inlet 44 extending beneath the float . a one - way valve 45 allows water to be drawn through the inlet 44 and into the cylinder when the piston 40 is moved upward . a separate such inlet with one - way valve is connected with a port 46 positioned at the top of the cylinder ( but for simplification is not shown ). thus as the piston moves up and down the evacuated side of the cylinder is always refilled and maintained full by water drawn in from the sea . in the embodiment shown , upper movement of the piston 40 will force water out through the port 46 , down through the conduit 47 , up through the conduit 48 and into an accumulator 49 . this accumulator is a closed chamber having an air space 50 so that the pumping of pressured water therein pressurizes the air . subsequent downward movement of the piston 40 will force water through the horizontal conduit 51 and upwards through the conduit 48 into the accumulator . stop valves 52 and 54 in the conduits 51 and 47 , respectively , prevent a back flow of water when water is being forced out of the other end of the cylinder . with pressured water flowing into the accumulator , the air pressure buildup results in a constant flow of water through the outlet 55 in the accumulator and upwards through the conduit 56 to a turbine 57 which when rotated turns a shaft 58 . coupled to the shaft is a drive belt 59 leading to a power generator 60 . thus pressured water forced into the accumulator 49 will result in the turning of the turbine to subsequently drive the power generator . by use of the accumulator there is a buffet provided which in essence stores energy in the form of air pressure in the accumulator to supply a constant source of pressured fluid to the turbine . similarly the cylinder - piston combination at the other corner of the barge functions to supply pressure fluid to the accumulator . the cylinder 37a and the connections with the conduits are marked with a similar numbered prefix and the suffix &# 34 ; a &# 34 ; when the function is identical . thus the piston 40a is pulled up and down by the similar diagonal cables 32a and 34a . these cables pass over pulleys 35a and 36a which run through seals 41a and into the cylinder . as the piston 40a is forced up and down , water is drawn into the inlet 44a in the same manner as previously described and another inlet ( not shown ) for passage into the cylinder . the cylinder in turn forces fluid through the horizontal pipes 61 and 62 to pass into the same accumulator 49 in the same manner as previously described . thus there is a parallel flow of fluid into the accumulator as both floats move relative to the barge . similarly there are positioned in parallel to the cylinders 37 and 37a other cylinders connecting with the horizontal cables 30 and 31 , which cylinders function in the same manner to supply pressured water either to the accumulator 49 or to another accumulator ( not shown ). thus as described there is provided a pump and valve system for transmitting the fluid from the pump to the turbine for driving the generator 60 . by use of the closed pressure system , the energy input to the turbine is smoothed out so as not to be as cyclic as might occur otherwise due to the rocking motion of the floats . water is readily available from the surrounding medium and the power can be generated on a more or less constant basis assuming the presence of current and / or wind for moving the float . additionally any number of closed pressure systems can be positioned on the barge so long as there is physical space and flotation and these other systems . all the systems can be caused to turn turbines operating on the same shaft 58 , or turbines independently connected to other generators . in fig4 is shown another embodiment of the invention . shown therein is a barge 64 and float 65 . the barge and float are connected by a pin 66 extending lengthwise through a housing rigidly fixed to the adjacent side of the floating craft . this connecting arrangement allows the craft to pivot under action of the ocean current and wind current but prohibits relatife up and down motion . such an arrangement simplifies the relative motion and makes more rigid the connecting of the craft . cables 66 and 67 are connected between the barge and float in the same manner as described before , with the ends of the cables connected to opposite sides of a piston 40b to pump water to a system in the same manner as previously described .", "category": "Fixed Constructions"}
|
Is the category the most suitable category for the given patent?
| 0.25 |
c84b2d917423930170182b72df89909a565c510d41712fe2673b0852d7275aab
| 0.005219 | 0.722656 | 0.004608 | 0.419922 | 0.072754 | 0.34375 |
null |
{"category": "Mechanical Engineering; Lightning; Heating; Weapons; Blasting", "patent": "in fig1 is shown a preferred embodiment of a wave powered generator incorporating the subject invention . a central barge 12 is positioned between and in spaced parallel relationship to a pair of elongated floats 14 and 15 . the floats 14 and 15 are each connected to the central positioned barge by a pair of pins 16 , 17 and 18 , 19 which are pivotally connected at opposite ends by ball - type sockets 20 . by such connections the floats are maintained more or less in parallel relationship to the barge but are permitted to rise and fall , sway , rock and pitch relative to each other and within the constraints of the rigid pins 16 , 17 , 18 and 19 . it is this relative movement caused primarily by the wave and wind action acting on the floats and barge that is used to generate power . the relative motions are generally described in the previously identified patent application . each float 14 and 15 includes a mast structure 21 held in position by guy wires 22 and supporting a plurality of wind vanes 24 which are controlled in a manner to enhance the relative motion between the floats and the barge . the wind vanes are supported in a manner ( not shown ) so as to pivot between a position extending in a vertical plane so as to catch the wind and a horizontal position so as not to catch the wind . preferably these wind current vanes are each made of a rigid material which is strong enough to withstand the force of the wind and also will hold up in the sea spray environment . additionally there is provided a mechanism ( not shown ) for actuating the vanes on each float in unison much in the manner that a venetian blind is moved between open and closed positions . a wind pressing on the closed vanes , i . e . vertically positioned vanes , will cause the float to heel over and the opening of the vanes will allow the float to move back to the normal position thereby creating a rocking motion . a plurality of water current vanes 27 ( fig1 and 2 ) are provided under each float 14 and 15 . these vanes are supported on vertically extending supports 28 and are rotatable ( preferably in unison ) about a longitudinal axis . by specific control of the vanes ( in a manner not shown ) the vanes can be manipulated so as to cause a rocking motion of the floats and , by cycling such vane movement with the wind vane movement previously described , the floats can be caused to rock relative to the barge . to make use of this rocking motion in the generation of power , there are provided a plurality of elongated members or cables 30 , 31 , 32 and 34 having one end fixed to the various locations on the floats . the cables 32 and 34 extend diagonally upward and downward , respectively , to upper and lower pulleys 35 and 36 located at the ends of vertically extending watertight cylinders 37 in the barge . similarly the cables 30 and 31 extend substantially horizontally over pulleys 38 and 39 positioned at opposite ends of the cylinders 37 and 38 . the cylinders 37 and 38 each function in a similar manner such that only one will be described . however , pairs of such cylinders preferably are positioned at the corners of the float 12 and , if desires , more cylinders can be spaced along the barge . as shown primarily in fig3 wherein a pair of the cylinders are shown in cross - section , a piston 40 is supported therein by the cables 32 and 34 connected to opposite sides . the cables pass through watertight seals 41 at the end of the cylinder . at the lower end of the cylinder is a port 42 connecting to a water inlet 44 extending beneath the float . a one - way valve 45 allows water to be drawn through the inlet 44 and into the cylinder when the piston 40 is moved upward . a separate such inlet with one - way valve is connected with a port 46 positioned at the top of the cylinder ( but for simplification is not shown ). thus as the piston moves up and down the evacuated side of the cylinder is always refilled and maintained full by water drawn in from the sea . in the embodiment shown , upper movement of the piston 40 will force water out through the port 46 , down through the conduit 47 , up through the conduit 48 and into an accumulator 49 . this accumulator is a closed chamber having an air space 50 so that the pumping of pressured water therein pressurizes the air . subsequent downward movement of the piston 40 will force water through the horizontal conduit 51 and upwards through the conduit 48 into the accumulator . stop valves 52 and 54 in the conduits 51 and 47 , respectively , prevent a back flow of water when water is being forced out of the other end of the cylinder . with pressured water flowing into the accumulator , the air pressure buildup results in a constant flow of water through the outlet 55 in the accumulator and upwards through the conduit 56 to a turbine 57 which when rotated turns a shaft 58 . coupled to the shaft is a drive belt 59 leading to a power generator 60 . thus pressured water forced into the accumulator 49 will result in the turning of the turbine to subsequently drive the power generator . by use of the accumulator there is a buffet provided which in essence stores energy in the form of air pressure in the accumulator to supply a constant source of pressured fluid to the turbine . similarly the cylinder - piston combination at the other corner of the barge functions to supply pressure fluid to the accumulator . the cylinder 37a and the connections with the conduits are marked with a similar numbered prefix and the suffix &# 34 ; a &# 34 ; when the function is identical . thus the piston 40a is pulled up and down by the similar diagonal cables 32a and 34a . these cables pass over pulleys 35a and 36a which run through seals 41a and into the cylinder . as the piston 40a is forced up and down , water is drawn into the inlet 44a in the same manner as previously described and another inlet ( not shown ) for passage into the cylinder . the cylinder in turn forces fluid through the horizontal pipes 61 and 62 to pass into the same accumulator 49 in the same manner as previously described . thus there is a parallel flow of fluid into the accumulator as both floats move relative to the barge . similarly there are positioned in parallel to the cylinders 37 and 37a other cylinders connecting with the horizontal cables 30 and 31 , which cylinders function in the same manner to supply pressured water either to the accumulator 49 or to another accumulator ( not shown ). thus as described there is provided a pump and valve system for transmitting the fluid from the pump to the turbine for driving the generator 60 . by use of the closed pressure system , the energy input to the turbine is smoothed out so as not to be as cyclic as might occur otherwise due to the rocking motion of the floats . water is readily available from the surrounding medium and the power can be generated on a more or less constant basis assuming the presence of current and / or wind for moving the float . additionally any number of closed pressure systems can be positioned on the barge so long as there is physical space and flotation and these other systems . all the systems can be caused to turn turbines operating on the same shaft 58 , or turbines independently connected to other generators . in fig4 is shown another embodiment of the invention . shown therein is a barge 64 and float 65 . the barge and float are connected by a pin 66 extending lengthwise through a housing rigidly fixed to the adjacent side of the floating craft . this connecting arrangement allows the craft to pivot under action of the ocean current and wind current but prohibits relatife up and down motion . such an arrangement simplifies the relative motion and makes more rigid the connecting of the craft . cables 66 and 67 are connected between the barge and float in the same manner as described before , with the ends of the cables connected to opposite sides of a piston 40b to pump water to a system in the same manner as previously described ."}
|
{"category": "Physics", "patent": "in fig1 is shown a preferred embodiment of a wave powered generator incorporating the subject invention . a central barge 12 is positioned between and in spaced parallel relationship to a pair of elongated floats 14 and 15 . the floats 14 and 15 are each connected to the central positioned barge by a pair of pins 16 , 17 and 18 , 19 which are pivotally connected at opposite ends by ball - type sockets 20 . by such connections the floats are maintained more or less in parallel relationship to the barge but are permitted to rise and fall , sway , rock and pitch relative to each other and within the constraints of the rigid pins 16 , 17 , 18 and 19 . it is this relative movement caused primarily by the wave and wind action acting on the floats and barge that is used to generate power . the relative motions are generally described in the previously identified patent application . each float 14 and 15 includes a mast structure 21 held in position by guy wires 22 and supporting a plurality of wind vanes 24 which are controlled in a manner to enhance the relative motion between the floats and the barge . the wind vanes are supported in a manner ( not shown ) so as to pivot between a position extending in a vertical plane so as to catch the wind and a horizontal position so as not to catch the wind . preferably these wind current vanes are each made of a rigid material which is strong enough to withstand the force of the wind and also will hold up in the sea spray environment . additionally there is provided a mechanism ( not shown ) for actuating the vanes on each float in unison much in the manner that a venetian blind is moved between open and closed positions . a wind pressing on the closed vanes , i . e . vertically positioned vanes , will cause the float to heel over and the opening of the vanes will allow the float to move back to the normal position thereby creating a rocking motion . a plurality of water current vanes 27 ( fig1 and 2 ) are provided under each float 14 and 15 . these vanes are supported on vertically extending supports 28 and are rotatable ( preferably in unison ) about a longitudinal axis . by specific control of the vanes ( in a manner not shown ) the vanes can be manipulated so as to cause a rocking motion of the floats and , by cycling such vane movement with the wind vane movement previously described , the floats can be caused to rock relative to the barge . to make use of this rocking motion in the generation of power , there are provided a plurality of elongated members or cables 30 , 31 , 32 and 34 having one end fixed to the various locations on the floats . the cables 32 and 34 extend diagonally upward and downward , respectively , to upper and lower pulleys 35 and 36 located at the ends of vertically extending watertight cylinders 37 in the barge . similarly the cables 30 and 31 extend substantially horizontally over pulleys 38 and 39 positioned at opposite ends of the cylinders 37 and 38 . the cylinders 37 and 38 each function in a similar manner such that only one will be described . however , pairs of such cylinders preferably are positioned at the corners of the float 12 and , if desires , more cylinders can be spaced along the barge . as shown primarily in fig3 wherein a pair of the cylinders are shown in cross - section , a piston 40 is supported therein by the cables 32 and 34 connected to opposite sides . the cables pass through watertight seals 41 at the end of the cylinder . at the lower end of the cylinder is a port 42 connecting to a water inlet 44 extending beneath the float . a one - way valve 45 allows water to be drawn through the inlet 44 and into the cylinder when the piston 40 is moved upward . a separate such inlet with one - way valve is connected with a port 46 positioned at the top of the cylinder ( but for simplification is not shown ). thus as the piston moves up and down the evacuated side of the cylinder is always refilled and maintained full by water drawn in from the sea . in the embodiment shown , upper movement of the piston 40 will force water out through the port 46 , down through the conduit 47 , up through the conduit 48 and into an accumulator 49 . this accumulator is a closed chamber having an air space 50 so that the pumping of pressured water therein pressurizes the air . subsequent downward movement of the piston 40 will force water through the horizontal conduit 51 and upwards through the conduit 48 into the accumulator . stop valves 52 and 54 in the conduits 51 and 47 , respectively , prevent a back flow of water when water is being forced out of the other end of the cylinder . with pressured water flowing into the accumulator , the air pressure buildup results in a constant flow of water through the outlet 55 in the accumulator and upwards through the conduit 56 to a turbine 57 which when rotated turns a shaft 58 . coupled to the shaft is a drive belt 59 leading to a power generator 60 . thus pressured water forced into the accumulator 49 will result in the turning of the turbine to subsequently drive the power generator . by use of the accumulator there is a buffet provided which in essence stores energy in the form of air pressure in the accumulator to supply a constant source of pressured fluid to the turbine . similarly the cylinder - piston combination at the other corner of the barge functions to supply pressure fluid to the accumulator . the cylinder 37a and the connections with the conduits are marked with a similar numbered prefix and the suffix &# 34 ; a &# 34 ; when the function is identical . thus the piston 40a is pulled up and down by the similar diagonal cables 32a and 34a . these cables pass over pulleys 35a and 36a which run through seals 41a and into the cylinder . as the piston 40a is forced up and down , water is drawn into the inlet 44a in the same manner as previously described and another inlet ( not shown ) for passage into the cylinder . the cylinder in turn forces fluid through the horizontal pipes 61 and 62 to pass into the same accumulator 49 in the same manner as previously described . thus there is a parallel flow of fluid into the accumulator as both floats move relative to the barge . similarly there are positioned in parallel to the cylinders 37 and 37a other cylinders connecting with the horizontal cables 30 and 31 , which cylinders function in the same manner to supply pressured water either to the accumulator 49 or to another accumulator ( not shown ). thus as described there is provided a pump and valve system for transmitting the fluid from the pump to the turbine for driving the generator 60 . by use of the closed pressure system , the energy input to the turbine is smoothed out so as not to be as cyclic as might occur otherwise due to the rocking motion of the floats . water is readily available from the surrounding medium and the power can be generated on a more or less constant basis assuming the presence of current and / or wind for moving the float . additionally any number of closed pressure systems can be positioned on the barge so long as there is physical space and flotation and these other systems . all the systems can be caused to turn turbines operating on the same shaft 58 , or turbines independently connected to other generators . in fig4 is shown another embodiment of the invention . shown therein is a barge 64 and float 65 . the barge and float are connected by a pin 66 extending lengthwise through a housing rigidly fixed to the adjacent side of the floating craft . this connecting arrangement allows the craft to pivot under action of the ocean current and wind current but prohibits relatife up and down motion . such an arrangement simplifies the relative motion and makes more rigid the connecting of the craft . cables 66 and 67 are connected between the barge and float in the same manner as described before , with the ends of the cables connected to opposite sides of a piston 40b to pump water to a system in the same manner as previously described ."}
|
Is the patent correctly categorized?
| 0.25 |
c84b2d917423930170182b72df89909a565c510d41712fe2673b0852d7275aab
| 0.020996 | 0.402344 | 0.013245 | 0.875 | 0.376953 | 0.808594 |
null |
{"category": "Mechanical Engineering; Lightning; Heating; Weapons; Blasting", "patent": "in fig1 is shown a preferred embodiment of a wave powered generator incorporating the subject invention . a central barge 12 is positioned between and in spaced parallel relationship to a pair of elongated floats 14 and 15 . the floats 14 and 15 are each connected to the central positioned barge by a pair of pins 16 , 17 and 18 , 19 which are pivotally connected at opposite ends by ball - type sockets 20 . by such connections the floats are maintained more or less in parallel relationship to the barge but are permitted to rise and fall , sway , rock and pitch relative to each other and within the constraints of the rigid pins 16 , 17 , 18 and 19 . it is this relative movement caused primarily by the wave and wind action acting on the floats and barge that is used to generate power . the relative motions are generally described in the previously identified patent application . each float 14 and 15 includes a mast structure 21 held in position by guy wires 22 and supporting a plurality of wind vanes 24 which are controlled in a manner to enhance the relative motion between the floats and the barge . the wind vanes are supported in a manner ( not shown ) so as to pivot between a position extending in a vertical plane so as to catch the wind and a horizontal position so as not to catch the wind . preferably these wind current vanes are each made of a rigid material which is strong enough to withstand the force of the wind and also will hold up in the sea spray environment . additionally there is provided a mechanism ( not shown ) for actuating the vanes on each float in unison much in the manner that a venetian blind is moved between open and closed positions . a wind pressing on the closed vanes , i . e . vertically positioned vanes , will cause the float to heel over and the opening of the vanes will allow the float to move back to the normal position thereby creating a rocking motion . a plurality of water current vanes 27 ( fig1 and 2 ) are provided under each float 14 and 15 . these vanes are supported on vertically extending supports 28 and are rotatable ( preferably in unison ) about a longitudinal axis . by specific control of the vanes ( in a manner not shown ) the vanes can be manipulated so as to cause a rocking motion of the floats and , by cycling such vane movement with the wind vane movement previously described , the floats can be caused to rock relative to the barge . to make use of this rocking motion in the generation of power , there are provided a plurality of elongated members or cables 30 , 31 , 32 and 34 having one end fixed to the various locations on the floats . the cables 32 and 34 extend diagonally upward and downward , respectively , to upper and lower pulleys 35 and 36 located at the ends of vertically extending watertight cylinders 37 in the barge . similarly the cables 30 and 31 extend substantially horizontally over pulleys 38 and 39 positioned at opposite ends of the cylinders 37 and 38 . the cylinders 37 and 38 each function in a similar manner such that only one will be described . however , pairs of such cylinders preferably are positioned at the corners of the float 12 and , if desires , more cylinders can be spaced along the barge . as shown primarily in fig3 wherein a pair of the cylinders are shown in cross - section , a piston 40 is supported therein by the cables 32 and 34 connected to opposite sides . the cables pass through watertight seals 41 at the end of the cylinder . at the lower end of the cylinder is a port 42 connecting to a water inlet 44 extending beneath the float . a one - way valve 45 allows water to be drawn through the inlet 44 and into the cylinder when the piston 40 is moved upward . a separate such inlet with one - way valve is connected with a port 46 positioned at the top of the cylinder ( but for simplification is not shown ). thus as the piston moves up and down the evacuated side of the cylinder is always refilled and maintained full by water drawn in from the sea . in the embodiment shown , upper movement of the piston 40 will force water out through the port 46 , down through the conduit 47 , up through the conduit 48 and into an accumulator 49 . this accumulator is a closed chamber having an air space 50 so that the pumping of pressured water therein pressurizes the air . subsequent downward movement of the piston 40 will force water through the horizontal conduit 51 and upwards through the conduit 48 into the accumulator . stop valves 52 and 54 in the conduits 51 and 47 , respectively , prevent a back flow of water when water is being forced out of the other end of the cylinder . with pressured water flowing into the accumulator , the air pressure buildup results in a constant flow of water through the outlet 55 in the accumulator and upwards through the conduit 56 to a turbine 57 which when rotated turns a shaft 58 . coupled to the shaft is a drive belt 59 leading to a power generator 60 . thus pressured water forced into the accumulator 49 will result in the turning of the turbine to subsequently drive the power generator . by use of the accumulator there is a buffet provided which in essence stores energy in the form of air pressure in the accumulator to supply a constant source of pressured fluid to the turbine . similarly the cylinder - piston combination at the other corner of the barge functions to supply pressure fluid to the accumulator . the cylinder 37a and the connections with the conduits are marked with a similar numbered prefix and the suffix &# 34 ; a &# 34 ; when the function is identical . thus the piston 40a is pulled up and down by the similar diagonal cables 32a and 34a . these cables pass over pulleys 35a and 36a which run through seals 41a and into the cylinder . as the piston 40a is forced up and down , water is drawn into the inlet 44a in the same manner as previously described and another inlet ( not shown ) for passage into the cylinder . the cylinder in turn forces fluid through the horizontal pipes 61 and 62 to pass into the same accumulator 49 in the same manner as previously described . thus there is a parallel flow of fluid into the accumulator as both floats move relative to the barge . similarly there are positioned in parallel to the cylinders 37 and 37a other cylinders connecting with the horizontal cables 30 and 31 , which cylinders function in the same manner to supply pressured water either to the accumulator 49 or to another accumulator ( not shown ). thus as described there is provided a pump and valve system for transmitting the fluid from the pump to the turbine for driving the generator 60 . by use of the closed pressure system , the energy input to the turbine is smoothed out so as not to be as cyclic as might occur otherwise due to the rocking motion of the floats . water is readily available from the surrounding medium and the power can be generated on a more or less constant basis assuming the presence of current and / or wind for moving the float . additionally any number of closed pressure systems can be positioned on the barge so long as there is physical space and flotation and these other systems . all the systems can be caused to turn turbines operating on the same shaft 58 , or turbines independently connected to other generators . in fig4 is shown another embodiment of the invention . shown therein is a barge 64 and float 65 . the barge and float are connected by a pin 66 extending lengthwise through a housing rigidly fixed to the adjacent side of the floating craft . this connecting arrangement allows the craft to pivot under action of the ocean current and wind current but prohibits relatife up and down motion . such an arrangement simplifies the relative motion and makes more rigid the connecting of the craft . cables 66 and 67 are connected between the barge and float in the same manner as described before , with the ends of the cables connected to opposite sides of a piston 40b to pump water to a system in the same manner as previously described ."}
|
{"patent": "in fig1 is shown a preferred embodiment of a wave powered generator incorporating the subject invention . a central barge 12 is positioned between and in spaced parallel relationship to a pair of elongated floats 14 and 15 . the floats 14 and 15 are each connected to the central positioned barge by a pair of pins 16 , 17 and 18 , 19 which are pivotally connected at opposite ends by ball - type sockets 20 . by such connections the floats are maintained more or less in parallel relationship to the barge but are permitted to rise and fall , sway , rock and pitch relative to each other and within the constraints of the rigid pins 16 , 17 , 18 and 19 . it is this relative movement caused primarily by the wave and wind action acting on the floats and barge that is used to generate power . the relative motions are generally described in the previously identified patent application . each float 14 and 15 includes a mast structure 21 held in position by guy wires 22 and supporting a plurality of wind vanes 24 which are controlled in a manner to enhance the relative motion between the floats and the barge . the wind vanes are supported in a manner ( not shown ) so as to pivot between a position extending in a vertical plane so as to catch the wind and a horizontal position so as not to catch the wind . preferably these wind current vanes are each made of a rigid material which is strong enough to withstand the force of the wind and also will hold up in the sea spray environment . additionally there is provided a mechanism ( not shown ) for actuating the vanes on each float in unison much in the manner that a venetian blind is moved between open and closed positions . a wind pressing on the closed vanes , i . e . vertically positioned vanes , will cause the float to heel over and the opening of the vanes will allow the float to move back to the normal position thereby creating a rocking motion . a plurality of water current vanes 27 ( fig1 and 2 ) are provided under each float 14 and 15 . these vanes are supported on vertically extending supports 28 and are rotatable ( preferably in unison ) about a longitudinal axis . by specific control of the vanes ( in a manner not shown ) the vanes can be manipulated so as to cause a rocking motion of the floats and , by cycling such vane movement with the wind vane movement previously described , the floats can be caused to rock relative to the barge . to make use of this rocking motion in the generation of power , there are provided a plurality of elongated members or cables 30 , 31 , 32 and 34 having one end fixed to the various locations on the floats . the cables 32 and 34 extend diagonally upward and downward , respectively , to upper and lower pulleys 35 and 36 located at the ends of vertically extending watertight cylinders 37 in the barge . similarly the cables 30 and 31 extend substantially horizontally over pulleys 38 and 39 positioned at opposite ends of the cylinders 37 and 38 . the cylinders 37 and 38 each function in a similar manner such that only one will be described . however , pairs of such cylinders preferably are positioned at the corners of the float 12 and , if desires , more cylinders can be spaced along the barge . as shown primarily in fig3 wherein a pair of the cylinders are shown in cross - section , a piston 40 is supported therein by the cables 32 and 34 connected to opposite sides . the cables pass through watertight seals 41 at the end of the cylinder . at the lower end of the cylinder is a port 42 connecting to a water inlet 44 extending beneath the float . a one - way valve 45 allows water to be drawn through the inlet 44 and into the cylinder when the piston 40 is moved upward . a separate such inlet with one - way valve is connected with a port 46 positioned at the top of the cylinder ( but for simplification is not shown ). thus as the piston moves up and down the evacuated side of the cylinder is always refilled and maintained full by water drawn in from the sea . in the embodiment shown , upper movement of the piston 40 will force water out through the port 46 , down through the conduit 47 , up through the conduit 48 and into an accumulator 49 . this accumulator is a closed chamber having an air space 50 so that the pumping of pressured water therein pressurizes the air . subsequent downward movement of the piston 40 will force water through the horizontal conduit 51 and upwards through the conduit 48 into the accumulator . stop valves 52 and 54 in the conduits 51 and 47 , respectively , prevent a back flow of water when water is being forced out of the other end of the cylinder . with pressured water flowing into the accumulator , the air pressure buildup results in a constant flow of water through the outlet 55 in the accumulator and upwards through the conduit 56 to a turbine 57 which when rotated turns a shaft 58 . coupled to the shaft is a drive belt 59 leading to a power generator 60 . thus pressured water forced into the accumulator 49 will result in the turning of the turbine to subsequently drive the power generator . by use of the accumulator there is a buffet provided which in essence stores energy in the form of air pressure in the accumulator to supply a constant source of pressured fluid to the turbine . similarly the cylinder - piston combination at the other corner of the barge functions to supply pressure fluid to the accumulator . the cylinder 37a and the connections with the conduits are marked with a similar numbered prefix and the suffix &# 34 ; a &# 34 ; when the function is identical . thus the piston 40a is pulled up and down by the similar diagonal cables 32a and 34a . these cables pass over pulleys 35a and 36a which run through seals 41a and into the cylinder . as the piston 40a is forced up and down , water is drawn into the inlet 44a in the same manner as previously described and another inlet ( not shown ) for passage into the cylinder . the cylinder in turn forces fluid through the horizontal pipes 61 and 62 to pass into the same accumulator 49 in the same manner as previously described . thus there is a parallel flow of fluid into the accumulator as both floats move relative to the barge . similarly there are positioned in parallel to the cylinders 37 and 37a other cylinders connecting with the horizontal cables 30 and 31 , which cylinders function in the same manner to supply pressured water either to the accumulator 49 or to another accumulator ( not shown ). thus as described there is provided a pump and valve system for transmitting the fluid from the pump to the turbine for driving the generator 60 . by use of the closed pressure system , the energy input to the turbine is smoothed out so as not to be as cyclic as might occur otherwise due to the rocking motion of the floats . water is readily available from the surrounding medium and the power can be generated on a more or less constant basis assuming the presence of current and / or wind for moving the float . additionally any number of closed pressure systems can be positioned on the barge so long as there is physical space and flotation and these other systems . all the systems can be caused to turn turbines operating on the same shaft 58 , or turbines independently connected to other generators . in fig4 is shown another embodiment of the invention . shown therein is a barge 64 and float 65 . the barge and float are connected by a pin 66 extending lengthwise through a housing rigidly fixed to the adjacent side of the floating craft . this connecting arrangement allows the craft to pivot under action of the ocean current and wind current but prohibits relatife up and down motion . such an arrangement simplifies the relative motion and makes more rigid the connecting of the craft . cables 66 and 67 are connected between the barge and float in the same manner as described before , with the ends of the cables connected to opposite sides of a piston 40b to pump water to a system in the same manner as previously described .", "category": "Electricity"}
|
Does the category match the content of the patent?
| 0.25 |
c84b2d917423930170182b72df89909a565c510d41712fe2673b0852d7275aab
| 0.012024 | 0.710938 | 0.014038 | 0.28125 | 0.075684 | 0.341797 |
null |
{"patent": "in fig1 is shown a preferred embodiment of a wave powered generator incorporating the subject invention . a central barge 12 is positioned between and in spaced parallel relationship to a pair of elongated floats 14 and 15 . the floats 14 and 15 are each connected to the central positioned barge by a pair of pins 16 , 17 and 18 , 19 which are pivotally connected at opposite ends by ball - type sockets 20 . by such connections the floats are maintained more or less in parallel relationship to the barge but are permitted to rise and fall , sway , rock and pitch relative to each other and within the constraints of the rigid pins 16 , 17 , 18 and 19 . it is this relative movement caused primarily by the wave and wind action acting on the floats and barge that is used to generate power . the relative motions are generally described in the previously identified patent application . each float 14 and 15 includes a mast structure 21 held in position by guy wires 22 and supporting a plurality of wind vanes 24 which are controlled in a manner to enhance the relative motion between the floats and the barge . the wind vanes are supported in a manner ( not shown ) so as to pivot between a position extending in a vertical plane so as to catch the wind and a horizontal position so as not to catch the wind . preferably these wind current vanes are each made of a rigid material which is strong enough to withstand the force of the wind and also will hold up in the sea spray environment . additionally there is provided a mechanism ( not shown ) for actuating the vanes on each float in unison much in the manner that a venetian blind is moved between open and closed positions . a wind pressing on the closed vanes , i . e . vertically positioned vanes , will cause the float to heel over and the opening of the vanes will allow the float to move back to the normal position thereby creating a rocking motion . a plurality of water current vanes 27 ( fig1 and 2 ) are provided under each float 14 and 15 . these vanes are supported on vertically extending supports 28 and are rotatable ( preferably in unison ) about a longitudinal axis . by specific control of the vanes ( in a manner not shown ) the vanes can be manipulated so as to cause a rocking motion of the floats and , by cycling such vane movement with the wind vane movement previously described , the floats can be caused to rock relative to the barge . to make use of this rocking motion in the generation of power , there are provided a plurality of elongated members or cables 30 , 31 , 32 and 34 having one end fixed to the various locations on the floats . the cables 32 and 34 extend diagonally upward and downward , respectively , to upper and lower pulleys 35 and 36 located at the ends of vertically extending watertight cylinders 37 in the barge . similarly the cables 30 and 31 extend substantially horizontally over pulleys 38 and 39 positioned at opposite ends of the cylinders 37 and 38 . the cylinders 37 and 38 each function in a similar manner such that only one will be described . however , pairs of such cylinders preferably are positioned at the corners of the float 12 and , if desires , more cylinders can be spaced along the barge . as shown primarily in fig3 wherein a pair of the cylinders are shown in cross - section , a piston 40 is supported therein by the cables 32 and 34 connected to opposite sides . the cables pass through watertight seals 41 at the end of the cylinder . at the lower end of the cylinder is a port 42 connecting to a water inlet 44 extending beneath the float . a one - way valve 45 allows water to be drawn through the inlet 44 and into the cylinder when the piston 40 is moved upward . a separate such inlet with one - way valve is connected with a port 46 positioned at the top of the cylinder ( but for simplification is not shown ). thus as the piston moves up and down the evacuated side of the cylinder is always refilled and maintained full by water drawn in from the sea . in the embodiment shown , upper movement of the piston 40 will force water out through the port 46 , down through the conduit 47 , up through the conduit 48 and into an accumulator 49 . this accumulator is a closed chamber having an air space 50 so that the pumping of pressured water therein pressurizes the air . subsequent downward movement of the piston 40 will force water through the horizontal conduit 51 and upwards through the conduit 48 into the accumulator . stop valves 52 and 54 in the conduits 51 and 47 , respectively , prevent a back flow of water when water is being forced out of the other end of the cylinder . with pressured water flowing into the accumulator , the air pressure buildup results in a constant flow of water through the outlet 55 in the accumulator and upwards through the conduit 56 to a turbine 57 which when rotated turns a shaft 58 . coupled to the shaft is a drive belt 59 leading to a power generator 60 . thus pressured water forced into the accumulator 49 will result in the turning of the turbine to subsequently drive the power generator . by use of the accumulator there is a buffet provided which in essence stores energy in the form of air pressure in the accumulator to supply a constant source of pressured fluid to the turbine . similarly the cylinder - piston combination at the other corner of the barge functions to supply pressure fluid to the accumulator . the cylinder 37a and the connections with the conduits are marked with a similar numbered prefix and the suffix &# 34 ; a &# 34 ; when the function is identical . thus the piston 40a is pulled up and down by the similar diagonal cables 32a and 34a . these cables pass over pulleys 35a and 36a which run through seals 41a and into the cylinder . as the piston 40a is forced up and down , water is drawn into the inlet 44a in the same manner as previously described and another inlet ( not shown ) for passage into the cylinder . the cylinder in turn forces fluid through the horizontal pipes 61 and 62 to pass into the same accumulator 49 in the same manner as previously described . thus there is a parallel flow of fluid into the accumulator as both floats move relative to the barge . similarly there are positioned in parallel to the cylinders 37 and 37a other cylinders connecting with the horizontal cables 30 and 31 , which cylinders function in the same manner to supply pressured water either to the accumulator 49 or to another accumulator ( not shown ). thus as described there is provided a pump and valve system for transmitting the fluid from the pump to the turbine for driving the generator 60 . by use of the closed pressure system , the energy input to the turbine is smoothed out so as not to be as cyclic as might occur otherwise due to the rocking motion of the floats . water is readily available from the surrounding medium and the power can be generated on a more or less constant basis assuming the presence of current and / or wind for moving the float . additionally any number of closed pressure systems can be positioned on the barge so long as there is physical space and flotation and these other systems . all the systems can be caused to turn turbines operating on the same shaft 58 , or turbines independently connected to other generators . in fig4 is shown another embodiment of the invention . shown therein is a barge 64 and float 65 . the barge and float are connected by a pin 66 extending lengthwise through a housing rigidly fixed to the adjacent side of the floating craft . this connecting arrangement allows the craft to pivot under action of the ocean current and wind current but prohibits relatife up and down motion . such an arrangement simplifies the relative motion and makes more rigid the connecting of the craft . cables 66 and 67 are connected between the barge and float in the same manner as described before , with the ends of the cables connected to opposite sides of a piston 40b to pump water to a system in the same manner as previously described .", "category": "Mechanical Engineering; Lightning; Heating; Weapons; Blasting"}
|
{"patent": "in fig1 is shown a preferred embodiment of a wave powered generator incorporating the subject invention . a central barge 12 is positioned between and in spaced parallel relationship to a pair of elongated floats 14 and 15 . the floats 14 and 15 are each connected to the central positioned barge by a pair of pins 16 , 17 and 18 , 19 which are pivotally connected at opposite ends by ball - type sockets 20 . by such connections the floats are maintained more or less in parallel relationship to the barge but are permitted to rise and fall , sway , rock and pitch relative to each other and within the constraints of the rigid pins 16 , 17 , 18 and 19 . it is this relative movement caused primarily by the wave and wind action acting on the floats and barge that is used to generate power . the relative motions are generally described in the previously identified patent application . each float 14 and 15 includes a mast structure 21 held in position by guy wires 22 and supporting a plurality of wind vanes 24 which are controlled in a manner to enhance the relative motion between the floats and the barge . the wind vanes are supported in a manner ( not shown ) so as to pivot between a position extending in a vertical plane so as to catch the wind and a horizontal position so as not to catch the wind . preferably these wind current vanes are each made of a rigid material which is strong enough to withstand the force of the wind and also will hold up in the sea spray environment . additionally there is provided a mechanism ( not shown ) for actuating the vanes on each float in unison much in the manner that a venetian blind is moved between open and closed positions . a wind pressing on the closed vanes , i . e . vertically positioned vanes , will cause the float to heel over and the opening of the vanes will allow the float to move back to the normal position thereby creating a rocking motion . a plurality of water current vanes 27 ( fig1 and 2 ) are provided under each float 14 and 15 . these vanes are supported on vertically extending supports 28 and are rotatable ( preferably in unison ) about a longitudinal axis . by specific control of the vanes ( in a manner not shown ) the vanes can be manipulated so as to cause a rocking motion of the floats and , by cycling such vane movement with the wind vane movement previously described , the floats can be caused to rock relative to the barge . to make use of this rocking motion in the generation of power , there are provided a plurality of elongated members or cables 30 , 31 , 32 and 34 having one end fixed to the various locations on the floats . the cables 32 and 34 extend diagonally upward and downward , respectively , to upper and lower pulleys 35 and 36 located at the ends of vertically extending watertight cylinders 37 in the barge . similarly the cables 30 and 31 extend substantially horizontally over pulleys 38 and 39 positioned at opposite ends of the cylinders 37 and 38 . the cylinders 37 and 38 each function in a similar manner such that only one will be described . however , pairs of such cylinders preferably are positioned at the corners of the float 12 and , if desires , more cylinders can be spaced along the barge . as shown primarily in fig3 wherein a pair of the cylinders are shown in cross - section , a piston 40 is supported therein by the cables 32 and 34 connected to opposite sides . the cables pass through watertight seals 41 at the end of the cylinder . at the lower end of the cylinder is a port 42 connecting to a water inlet 44 extending beneath the float . a one - way valve 45 allows water to be drawn through the inlet 44 and into the cylinder when the piston 40 is moved upward . a separate such inlet with one - way valve is connected with a port 46 positioned at the top of the cylinder ( but for simplification is not shown ). thus as the piston moves up and down the evacuated side of the cylinder is always refilled and maintained full by water drawn in from the sea . in the embodiment shown , upper movement of the piston 40 will force water out through the port 46 , down through the conduit 47 , up through the conduit 48 and into an accumulator 49 . this accumulator is a closed chamber having an air space 50 so that the pumping of pressured water therein pressurizes the air . subsequent downward movement of the piston 40 will force water through the horizontal conduit 51 and upwards through the conduit 48 into the accumulator . stop valves 52 and 54 in the conduits 51 and 47 , respectively , prevent a back flow of water when water is being forced out of the other end of the cylinder . with pressured water flowing into the accumulator , the air pressure buildup results in a constant flow of water through the outlet 55 in the accumulator and upwards through the conduit 56 to a turbine 57 which when rotated turns a shaft 58 . coupled to the shaft is a drive belt 59 leading to a power generator 60 . thus pressured water forced into the accumulator 49 will result in the turning of the turbine to subsequently drive the power generator . by use of the accumulator there is a buffet provided which in essence stores energy in the form of air pressure in the accumulator to supply a constant source of pressured fluid to the turbine . similarly the cylinder - piston combination at the other corner of the barge functions to supply pressure fluid to the accumulator . the cylinder 37a and the connections with the conduits are marked with a similar numbered prefix and the suffix &# 34 ; a &# 34 ; when the function is identical . thus the piston 40a is pulled up and down by the similar diagonal cables 32a and 34a . these cables pass over pulleys 35a and 36a which run through seals 41a and into the cylinder . as the piston 40a is forced up and down , water is drawn into the inlet 44a in the same manner as previously described and another inlet ( not shown ) for passage into the cylinder . the cylinder in turn forces fluid through the horizontal pipes 61 and 62 to pass into the same accumulator 49 in the same manner as previously described . thus there is a parallel flow of fluid into the accumulator as both floats move relative to the barge . similarly there are positioned in parallel to the cylinders 37 and 37a other cylinders connecting with the horizontal cables 30 and 31 , which cylinders function in the same manner to supply pressured water either to the accumulator 49 or to another accumulator ( not shown ). thus as described there is provided a pump and valve system for transmitting the fluid from the pump to the turbine for driving the generator 60 . by use of the closed pressure system , the energy input to the turbine is smoothed out so as not to be as cyclic as might occur otherwise due to the rocking motion of the floats . water is readily available from the surrounding medium and the power can be generated on a more or less constant basis assuming the presence of current and / or wind for moving the float . additionally any number of closed pressure systems can be positioned on the barge so long as there is physical space and flotation and these other systems . all the systems can be caused to turn turbines operating on the same shaft 58 , or turbines independently connected to other generators . in fig4 is shown another embodiment of the invention . shown therein is a barge 64 and float 65 . the barge and float are connected by a pin 66 extending lengthwise through a housing rigidly fixed to the adjacent side of the floating craft . this connecting arrangement allows the craft to pivot under action of the ocean current and wind current but prohibits relatife up and down motion . such an arrangement simplifies the relative motion and makes more rigid the connecting of the craft . cables 66 and 67 are connected between the barge and float in the same manner as described before , with the ends of the cables connected to opposite sides of a piston 40b to pump water to a system in the same manner as previously described .", "category": "General tagging of new or cross-sectional technology"}
|
Is the category the most suitable category for the given patent?
| 0.25 |
c84b2d917423930170182b72df89909a565c510d41712fe2673b0852d7275aab
| 0.016357 | 0.326172 | 0.006287 | 0.161133 | 0.092773 | 0.125977 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.