CADS: A Comprehensive Anatomical Dataset and Segmentation for Whole-Body Anatomy in Computed Tomography
Organizations: Department of Quantitative Biomedicine, University of Zurich, Zurich, Switzerland · ETH AI Center, ETH Zurich, Zurich, Switzerland · Department of Radiation Oncology, University Hospital and University of Zurich, Zurich, Switzerland · Faculty of Medicine, University of Zurich, Zurich, Switzerland ·
Abstract
Volumetric computed tomography (CT) segmentation supports quantitative image analysis, but most AI-based methods remain organ-specific, with fragmented models, inconsistent label definitions, and heterogeneous training and evaluation protocols. Here we present CADS, a unified data-model framework for scalable and generalizable whole-body CT segmentation. We construct a SNOMED-CT-referenced anatomical label space harmonizing heterogeneous cohorts from over 100 imaging centers, yielding complete annotations of 167 structures across 22,022 CT volumes and markedly expanding anatomical coverage and scale beyond existing resources. To complete labels on sparsely annotated CTs, we combine self-training with shape-informed quality control based on neural implicit functions, curating high-quality pseudo-labels. Trained on this resource, CADS shows robust whole-body segmentation across distribution shifts and improved performance relative to state-of-the-art approaches across 18 public datasets. We further validate it in an independent, expert-reviewed real-world oncology cohort of 2,864 CTs. The curated dataset, pretrained models, and user-friendly software together provide an accessible, reproducible foundation for large-scale anatomical segmentation and computational anatomy research.
Figures & tables
| Data Source | #Vols | # Ann. Vols | # Ann. Struct. | Body Reg. | Ann. Method | #Cent. | Countries | Contrast | Pathology | License |
| VISCERAL Gold Corpus [ 40 ] | 40 | 40 | 20 | WB, VR | Human | 3 | AT, DE, ES | Mix | Pathologic: Bone Marrow Neoplasm, Lymphoma | Non-commercial (by agreement) |
| VISCERAL Gold Corpus-Extra | - | 40 | 22 | WB, VR | Human | 3 | AT, DE, ES | Mix | Pathologic: Bone Marrow Neoplasm, Lymphoma | Non-commercial (by agreement) |
| VISCERAL Silver Corpus | 127 | 127 | 20 | WB, VR | Automated | 3 | AT, DE, ES | Mix | Pathologic: Bone Marrow Neoplasm, Lymphoma | Non-commercial (by agreement) |
| KiTS [ 68 ] | 300 | 300 | 1 | AB | Human | 1 | US | Contrast-enhanced | Pathologic: Kidney Tumor | CC BY-NC-SA 4.0 |
| LiTS [ 69 ] | 201 | 201 | 1 | AB | Human | 7 | DE, NL, CA, IL, FR | Contrast-enhanced | Pathologic: Liver Tumor (Primary & Metastasis) | CC BY-NC-SA 4.0 |
| BTCV-Abdomen [ 70 ] | 50 | 30 | 13 | AB | Human | 1 | US | Contrast-enhanced | Pathologic: Colorectal Cancer, Hernia | CC BY 4.0 |
| Structure | Cases | Median size | Structure | Cases | Median size | Structure | Cases | Median size | Structure | Cases | Median size |
| Spleen | Kidney R | Kidney L | Gallbladder | ||||||||
| Liver | Stomach | Aorta | Inferior vena cava | ||||||||
| Portal and splenic vein | Pancreas | Adrenal gland R | Adrenal gland L | ||||||||
| Upper lobe of lung L | Lower lobe of lung L | Upper lobe of lung R | Middle lobe lung R | ||||||||
| Lower lobe of lung R | Vertebra L5 | Vertebra L4 | Vertebra L3 | ||||||||
| Vertebra L2 | Vertebra L1 | Vertebra T12 | Vertebra T11 |
Medical Image Segmentation
Medical image segmentation delineates organs, lesions, and other structures in scans to support measurement, diagnosis, treatment planning, and research.
See all 839 papers