cancer life sciences medical imaging nifti radiology segmentation
Open, multi-center dataset of 1,597 whole-/total-body FDG-PET/CT studies with 130 CT-derived, expert-verified anatomical segmentations per scan (~250 GB). Provided as anonymized NIfTI (PET, CT, labels) with spreadsheet metadata. Designed for segmentation benchmarking, multi-organ analysis, radiomics, and PET/CT AI research.
Ad hoc (new releases aligned with additional cohort availability)
Dataset licensing per originating site: - AutoPET Challenge: CC BY-NC 4.0 (non-commercial use) - University Hospital Leipzig: CC BY 4.0 - Azienda Ospedaliero Universitaria Careggi: CC BY 4.0 Software (MOOSE): Apache-2.0.
https://github.com/ENHANCE-PET/MOOSE/blob/main/DATA_CARD.md
ENHANCE.PET initiative (LMU Klinikum & partners)
See all datasets managed by ENHANCE.PET initiative (LMU Klinikum & partners).
Lalith.shiyam@med.uni-muenchen.de
ENHANCE.PET 1.6k - Whole-/Total-Body [18F]FDG-PET/CT with CT-Derived Segmentations was accessed on DATE
from https://registry.opendata.aws/enhance-pet-1-6k. Ferrara D. et al. (2025). Sharing a whole-/total-body [18F]FDG-PET/CT dataset with CT-derived segmentations - an ENHANCE.PET initiative. https://doi.org/10.21203/rs.3.rs-7169062/v2
arn:aws:s3:::enhance-pet-1-6k
us-west-2
aws s3 ls --no-sign-request s3://enhance-pet-1-6k/
arn:aws:sns:us-west-2:602670427264:enhance-pet-1-6k-object_created
us-west-2