QuPath
Open-source software for bioimage analysis, with a focus on digital pathology and whole-slide imaging.
Overview
QuPath is built around efficient viewing and annotation of large 2D images (whole-slide scans, tiled images). It combines classical detection (cell/nucleus detection, pixel classifiers) with deep-learning extensions and is scriptable in Groovy for reproducible pipelines.
Key features
- Fast viewing of multi-gigapixel whole-slide images
- Annotation, cell detection, and pixel classification built in
- Groovy scripting with a clear API
- Extension system covering deep-learning segmentation and external data sources
Installation
Download the latest version from qupath.github.io. Detailed install notes (including GPU setup for deep-learning extensions) are in the official installation guide.
Extensions
Useful extensions commonly used alongside QuPath:
- InstanSeg — generalist cell/nucleus segmentation built into recent QuPath versions.
- StarDist — star-convex nucleus segmentation.
- Cellpose — calls a local Cellpose install from QuPath.
- OMERO extension — load images directly from an OMERO server into a QuPath project.
Official documentation
Learning resources
- The official documentation has very clear tutorials
Written
- The official documentation has very clear Tutorials
- Scripting guide — automating analysis with Groovy.
Video
- QuPath tutorials by Pete Bankhead — the maintainer's own walkthroughs; the best starting point.