The ULS23 Challenge Public Training Dataset

This dataset contains part of the imaging data for the Universal Lesion Segmentation Challenge (ULS23). It contains the lesion volumes-of-interest (VOI's) for the novel annotated data of this challenge. It consists of 750 lesions from the DeepLesion dataset, 744 bone and 124 pancreas lesions from the Radboudumc. Pancreas data was collected by Natália Alves, Megan Schuurmans and Henkjan Huisman. The annotations are made available through the Challenge repository on GitHub. The full training dataset: Part 2: kits21, LIDC-IDRI, LiTS Part 3: MDSC task 6/7/10, NIH-LN, CCC18 Part 4: DeepLesion Part 5: DeepLesion Part 6: DeepLesion The Universal Lesion Segmentation 2023 (ULS23) data is licensed under CC BY-NC-SA 4.0