Dataset

Part of the dataset is adapted from MSD ( Liver, Spleen, Pancreas ), NIH Pancreas, and KiTS under their license permission. We also include a hidden testing set with 50 abdomen CT cases from Nanjing University. Annotations include four organs: liver ( label=1 ), kidney ( label=2 ), spleen ( label=3 ), and pancreas ( label=4 ). More details are in the following paper:

---

@article{AbdomenCT-1K,

  title={AbdomenCT-1K: Is Abdominal Organ Segmentation A Solved Problem?},

  author={Ma, Jun and Zhang, Yao and Gu, Song and Zhu, Cheng and Ge, Cheng and Zhang, Yichi and An, Xingle and Wang, Congcong and Wang, Qiyuan and Liu, Xin and Cao, Shucheng and Zhang, Qi and Liu, Shangqing and Wang, Yunpeng and Li, Yuhui and He, Jian and Yang, Xiaoping},

  journal={IEEE Transactions on Pattern Analysis and Machine Intelligence},

  year={2021},

  doi={10.1109/TPAMI.2021.3100536}

}

The testing set is held by the organizers with the following folder structure:

 inputs/
├── test_001_0000.nii.gz
├── test_002_0000.nii.gz
 ...
├── test_100_0000.nii.gz
The segmentation results should be 
 outputs/
├── test_001.nii.gz
├── test_002.nii.gz
 ...
├── test_100.nii.gz

---

Download Training and Validation Set 

(License: CC-BY-NC-SA)

Zenodo 

BaiduNetDisk (PW:2021)

Baseline

FLARE21 nnUNet Baseline

Reference