Datasets:
metadata
license: cc-by-nc-nd-4.0
task_categories:
- image-classification
tags:
- lung cancer
- CT
- dicom
- chest ct
- medical
- Pathology recognition
size_categories:
- 100K<n<1M
Lung Cancer CT Images - 150,000+ studies
Dataset comprises 50,000+ studies with protocol and 100,000+ studies without protocol, featuring 24 pathologies for comprehensive medical imaging analysis. It provides detailed cross-sectional images of the chest, enabling advanced research in lung diseases, pulmonary nodules, and heart disease. - Get the data
Dataset characteristics:
| Characteristic | Data |
|---|---|
| Description | Chest CT scans with or without protocols |
| Data Types | DICOM |
| Markup | Segmentation of pathologies |
| Tasks | Pathology recognition, computer vision |
| Number of Studies | 150,000+ (50,000+ with protocol, 100,000+ without protocol) |
| Labeling | Binary labels per study (1 = pathology present, 0 = absent) |
| Pathologies | Developmental anomalies of the lower airways, Destruction/abscess of the lung, Chest soft tissue changes (breast/mammary gland masses), Pulmonary embolism, Pulmonary airflow disorders, Lung cancer, Osteoporosis, Hydrothorax, Coronary calcium, Aortic aneurysm, Pulmonary trunk dilation, Lymph node abnormalities, Pulmonary emphysema, Tuberculosis, Sarcoidosis |
📊 Sample dataset available! For full access, contact us to discuss purchase terms.
Dataset structure
- files — studies
- labels —file contains metadata and labels.