9,262 fully-annotated abdominal CT volumes containing segmentations for 25 distinct anatomical structures. The data is curated by the CCVL research group at Johns Hopkins University for the Touchstone Benchmarking Project to evaluate semantic segmentation and pre-training algorithms.
Use Cases
- Train semantic segmentation models to identify 25 specific anatomical structures within abdominal CT volumes
- Evaluate the performance of pre-training algorithms against the Touchstone Benchmarking Project standards
- Develop multi-organ segmentation pipelines using the 9,262 annotated CT volumes
Strengths
- 9,262 fully-annotated abdominal CT volumes
- Semantic segmentation masks for 25 different anatomical structures
- Integrated into the Touchstone Benchmarking Project for algorithm comparison