// the find
ImagingDataCommons/libdicom
C library for reading DICOM files
A C library for reading DICOM Whole Slide Image (WSI) files, returning compressed frame data rather than decoded pixel arrays. Backed by NCI/NIH and used as the DICOM backend in OpenSlide 4.0, so it has real production users. Aimed at people building medical imaging pipelines who need a lightweight, dependency-free way to pull frames and metadata out of DICOM files.
Zero runtime dependencies — the entire thing builds with meson and ships its own dict tables, which means no libxml or libtiff surprises in your build. Returns compressed frames rather than decoded RGB, which is exactly right for a tiling library; you pick your decoder. Python bindings (pylibdicom) are a first-class citizen with a clean API, not a thin afterthought. OpenSlide 4.0 using this as its DICOM implementation is a strong endorsement — that project has high correctness standards.
Scoped strictly to WSI — if you need to read general DICOM (CT, MRI, RT structs), this is not the library. Only 36 stars despite being NCI-funded and OpenSlide-integrated suggests it hasn't caught on outside that specific ecosystem. No write support and no DICOM tag construction, so it's read-only and you can't round-trip or generate files. The test suite is a single file (check_dicom.c) covering a handful of sample files — adequate for regression but thin for the edge cases DICOM is famous for.