AbdomenAtlas/SwissPancreas

Dataset: Swiss-Test is a new dataset, sourced from the University Hospital Basel (Basel, Switzerland). This dataset includes 684 CT-Report pairs, of which 445 include pancreatic tumors. It was used as an external test set in the paper Report Supervision (Medical Image Analysis).

• Reports and metadata (e.g., annonymized patient ids and exam dates) are available inside the file SwissPancreas_metadata.csv.

Report supervision: We present R-Super, a training strategy that transforms radiology reports (text) into direct (per-voxel) supervision for tumor segmentation AI. Before training, we use LLM to extract tumor information from radiology reports. Then, R-Super introduces new loss functions (Volume Loss & Ball Loss), which use this extracted information to teach the AI to segment tumors that are coherent with reports, in terms of tumor count, diameters, and locations. R-Super can train AI on large-scale CT-Report datasets (e.g., Merlin) plus small or large CT-Mask datasets (e.g., AbdomenAtlas, PanTS). In comparison to traditional training with masks only, using R-Super to train with masks and reports improved the performance of tumor segmentation AI by up to +16% in sensitivity, F1, AUC, DSC and NSD.

Report supervision code: https://github.com/MrGiovanni/R-Super

Papers

Report Supervision
Pedro R. A. S. Bassi, Wenxuan Li, Jakob Wasserthal, Jieneng Chen, Zheren Zhu, Chuntung Zhuang, Sergio Decherchi, Andrea Cavalli, Kang Wang, Yang Yang, Alan L. Yuille, Zongwei Zhou*
Johns Hopkins University
Medical Image Analysis 2026, Best Paper Award (runner-up)

Learning Segmentation from Radiology Reports
Pedro R. A. S. Bassi, Wenxuan Li, Jieneng Chen, Zheren Zhu, Tianyu Lin, Sergio Decherchi, Andrea Cavalli, Kang Wang, Yang Yang, Alan Yuille, Zongwei Zhou*
Johns Hopkins University
MICCAI 2025, Best Paper Award (runner-up)

Scaling Artificial Intelligence for Multi-Tumor Early Detection with More Reports, Fewer Masks
Pedro R. A. S. Bassi, Xinze Zhou, Wenxuan Li, Szymon Płotka, Jieneng Chen, Qi Chen, Zheren Zhu, Jakub Prządo, Ibrahim E. Hamacı, Sezgin Er, Yuhan Wang, Ashwin Kumar, Bjoern Menze, Jarosław B. Ćwikła, Yuyin Zhou, Akshay S. Chaudhari, Curtis P. Langlotz, Sergio Decherchi, Andrea Cavalli, Kang Wang, Yang Yang, Alan L. Yuille, Zongwei Zhou*
Johns Hopkins University, Stanford University, University of Zurich, University of Warsaw, Italian Institute of Technology, UCSF
Preprint

Awards

• RSNA 2025 Certificate of Merit Award
• MICCAI 2025 Best Paper Award (Runner-up)
• Elsevier-Medical Image Analysis 2026 Best Paper Award (Runner-up)

Citation

If you use the code, data or methods in this repository, please cite:

@article{BASSI2026104257,
title = {Report Supervision},
journal = {Medical Image Analysis},
volume = {114},
pages = {104257},
year = {2026},
issn = {1361-8415},
doi = {https://doi.org/10.1016/j.media.2026.104257},
url = {https://www.sciencedirect.com/science/article/pii/S1361841526003269},
author = {Pedro R.A.S. Bassi and Wenxuan Li and Jakob Wasserthal and Jieneng Chen and Xinze Zhou and Zheren Zhu and Chuntung Zhuang and Sergio Decherchi and Andrea Cavalli and Kang Wang and Yang Yang and Alan Yuille and Zongwei Zhou},
}

@inproceedings{bassi2025learning,
  title={Learning Segmentation from Radiology Reports},
  author={Bassi, Pedro R. A. S. and Li, Wenxuan and Chen, Jieneng and Zhu, Zheren and Lin, Tianyu and Decherchi, Sergio and Cavalli, Andrea and Wang, Kang and Yang, Yang and Yuille, Alan L. and Zhou, Zongwei},
  booktitle={International Conference on Medical Image Computing and Computer-Assisted Intervention},
  pages={305--315},
  year={2025},
  publisher={Springer},
  doi={10.1007/978-3-032-04971-1_29}
}

@article{bassi2025scaling,
  title={Scaling Artificial Intelligence for Multi-Tumor Early Detection with More Reports, Fewer Masks},
  author={Bassi, Pedro R. A. S. and Zhou, Xinze and Li, Wenxuan and P{\l}otka, Szymon and Chen, Jieneng and Chen, Qi and Zhu, Zheren and Prz{\k{a}}do, Jakub and Hamamci, Ibrahim E. and Er, Sezgin and others},
  journal={arXiv preprint arXiv:2510.14803},
  year={2025}
}

Caution for training/testing: University Hospital Basel is the source of this dataset, CancerVerse, and TotalSegmentator. Therefore, patient overlap is possible across these datasets.