NA-MIC Project Weeks
Back to Projects List
DICOM management platforms
Key Investigators
- Andrey Fedorov (BWH)
- James Petts (Institute of Cancer Research, UK)
- Erik Ziegler (Radical Imaging)
- Marco Nolden (DKFZ)
- Jonas Scherer (DKFZ)
Project Description
Evaluate existing/emerging platforms for managing DICOM data.
Objective
We need a platform to support management of the DICOM data within a lab. Required features include:
- user authentication
- storage of the variety of DICOM objects (should not be limited to DICOM images)
- integrated viewer for visualization of the images, and ideally annotations
- API for data access (read and write)
- integration of batch processing tools (e.g., automatically calculate radiomics features for an image when segmentation of that is submitted)
- should be possible to install on an “on-premises” server to allow handling of PHI data (i.e., cannot be a cloud-only solution)
Approach and Plan
Install and test the following solutions:
- XNAT (with James Petts)
- CCC? (with Erik Ziegler)
- Girder? (with JC)
- …
Progress and Next Steps
- Discussed the capabilities and planned release timeline of the platform being developed at DKFZ (Marco and Jonas). Looking forward to try it!
- Installed XNAT with the OHIF plugin (very easy and “just works”!):
- xnat-docker-compose
- ohif-viewer-XNAT-plugi (the more recent packages are in non-master branches, e.g., see https://github.com/JamesAPetts/ohif-viewer-XNAT-plugin/tree/xnatRoi-beta-1.10)
- learned that xnatpy exists, and is preferred over pyxnat (but neither is officially supported by the XNAT team :-( …)
- Did not get to try an improved version of the CCC infrastructure (OHIF + dcm4chee + keyckloak) - Jonas is working on it.
- Did not get to Girder …
Illustrations
