Datasets:

mrmrx
/

CADS-dataset

@inproceedings{landman2015miccai,
  title={Miccai multi-atlas labeling beyond the cranial vault--workshop and challenge},
  author={Landman, Bennett and Xu, Zhoubing and Igelsias, Juan and Styner, Martin and Langerak, Thomas and Klein, Arno},
  booktitle={Proc. MICCAI multi-atlas labeling beyond cranial vault—workshop challenge},
  volume={5},
  pages={12},
  year={2015},
  organization={Munich, Germany}
}

Dataset description

This dataset extends multi-atlas labeling to clinically acquired abdominal CT, focusing on cervical cancer patients eligible for radiotherapy. Scans were obtained under a full-bladder protocol, with most patients in prone position and high-resolution axial slices. Standard registration data was generated using ElastiX.

Challenge homepage: https://www.synapse.org/Synapse:syn3193805/wiki/89480

Number of CT volumes: 50

Contrast: Contrast-enhanced (portal venous phase)

CT body coverage: Abdomen

Does the dataset include any ground truth annotations?: Yes

Original GT annotation targets: Bladder, uterus, rectum, small bowel

Number of annotated CT volumes: 30

Annotator: Human

Acquisition centers: Vanderbilt University Medical Center (VUMC)

Pathology/Disease: Cervical cancer patients at various disease stages, all eligible for radiotherapy

Original dataset download link: https://www.synapse.org/Synapse:syn3193805/files/

Original dataset format: nifti