OHIF viewer integration
OHIF is an open-source, zero-footprint web viewer for DICOM. RadioGx now hosts a build of it at /assets/ohif/ and embeds it on the Imaging page, where AI analysis with your own key can read its viewport.
When to use OHIF instead of the Workbench
- Compressed DICOM (JPEG, JPEG 2000, JPEG-LS, RLE)
- Whole studies with several series, hanging protocols and multiplanar reformatting
- Segmentations, structured reports and RT structures
- Fused PET/CT display
Set it up
- Build or download OHIF from the OHIF repository (opens in a new tab) and host the static files on your own HTTPS domain.
- Point it at a DICOMweb server you control, such as Orthanc or a cloud healthcare API, holding only data you are permitted to use.
- To use a different OHIF deployment, set
ohifUrlinassets/js/config.jsto its HTTPS address. Leave it empty to use the bundled viewer. - On the Imaging page you can hand the bundled viewer a DICOMweb QIDO/WADO root and an optional Authorization header; they stay in the tab.
Rules that still apply
- Never place identifiable patient data on a public server. Use de-identified or synthetic data for teaching.
- The operator of the viewer is responsible for its access control, logging and privacy terms.
- AI analysis is optional, uses your own key and is for research and teaching, not diagnosis.
Public data without hosting anything
The NCI Imaging Data Commons offers public, de-identified collections with built-in viewers. Explore IDC (opens in a new tab).
About this tool
What is OHIF viewer integration?
Runs in your browser
The OHIF integration notes explain how RadioGx pairs with a self-hosted OHIF DICOM viewer: hosting, DICOMweb data sources, configuration, content security policy and privacy rules for imaging data.
Key features
- Hosting and DICOMweb configuration notes
- Content security policy guidance
- Privacy rules for imaging data