// the find
OHIF/Viewers
OHIF zero-footprint DICOM viewer and oncology specific Lesion Tracker, plus shared extension packages
OHIF Viewer is a production-grade, zero-footprint DICOM medical image viewer built on React and Cornerstone3D. It handles 2D/3D rendering, MPR, segmentation, RT structs, whole-slide microscopy, and 4D volumes — all in the browser without requiring software installs. Target audience is radiology/oncology researchers, hospital IT teams building imaging portals, and developers integrating DICOM viewing into clinical applications.
Extension and mode system is genuinely well-designed — core features are built on the same plugin API exposed to consumers, so you're not fighting a two-tier abstraction. Cornerstone3D integration testing setup (branch-pinned CI, Netlify preview deploys) is serious engineering that solves a real cross-repo coordination problem. The DICOM coverage is unusually broad: RT structs, parametric maps, dynamic 4D volumes, whole-slide microscopy, ECG waveforms — most viewers drop off at basic CT/MR. 8+ years of NIH-funded production use means it has handled edge cases in real DICOM data that proprietary viewers charge $50k/year to deal with.
Yarn 1.x (not Berry) in 2026 is a red flag — Yarn Classic is effectively abandoned and the monorepo dependency graph is complex enough that this will bite you during upgrades. The webpack-first build with an experimental rsbuild dev mode suggests the build story is in flux; don't plan a tight CI pipeline until that stabilizes. Configuring a custom datasource or non-standard PACS integration requires reading a lot of source code — the docs describe the happy path but the edge cases are underdocumented. Fork ratio (4255 forks vs 4201 stars) implies many private forks that never feed back upstream, which is a sign that customization in practice means forking, not extending.