Hey all,
I just visited the guys at XNAT Works during RSNA in Chicago, where we discussed a recent Horizon EU project which just finalized - EuCanImage (
eucanimage.eu). The project built up a repository of 25k subjects in Colorectal, Liver and Breast cancer using CT, MRI and Mammography imaging including pixel/voxel perfect segmentations stored directly in DICOM RTSTRUCT and later converted to Nifti.
The two main platforms used to achieve this was Collective Minds (
collectiveminds.health) and XNAT. The data de-identification, transfer, ingestion and segmentation, annotation and classification was handled on the collective minds research platform, where after an integration tooling was built to continuously move the resulting augmented data to XNAT as the final, long term repository.
Should you be interested in the integration tooling for other projects, you can find it here.
Kudos to Ivan Bocharov at Erasmus MC for these efforts and publishing the tooling.
All the best and safe flights home for those who were just in Chicago!
Pär (CTO here at Collective Minds)