Description
pathml is an agent skill for local computational pathology research with PathML. It covers loading and tiling slides, preprocessing and quality control, storage in h5path files, multiplex quantification and spatial analysis. The skill is clear about its boundaries. PathML is beta research software. It is not a validated medical device, diagnostic system or clinical decision support tool, and it is not a substitute for a pathologist, so the outputs must not be used to diagnose, grade, stage or treat a patient. Pathology files can contain identifying information, such as faces, labels, accession numbers, patient identifiers, DICOM tags, file names and linked clinical data. The skill asks you to confirm authorisation, consent or waiver, data-use terms and institutional policy before you process anything. It asks you to de-identify the pixels and the metadata, keep the re-identification key outside the analysis workspace, and use pseudonymous identifiers for patients, slides and specimens. Inputs, intermediate files and outputs should stay on approved, encrypted local storage, and you should split the data by patient, then by slide, before tiling or fitting any preprocessing step.…
Highlights
- Slide loading and tiling
- Preprocessing and quality control
- h5path storage
- Multiplex quantification and spatial analysis
- Research use with privacy safeguards
Ratings & Reviews
No ratings yet
Be the first to rate pathml.
People also visit
pathml FAQ
What is PathML?
Research software for computational pathology, with slide loading, tiling, preprocessing, storage, quantification and spatial analysis.
Can I use the outputs for diagnosis?
No. The skill says PathML is beta research software, not a validated diagnostic tool, and its outputs must not be used to diagnose, grade, stage or treat a patient.
What must I do before processing slides?
Confirm authorisation, consent or waiver, data-use terms and institutional policy, and de-identify the images and metadata.
How should identifiers be handled?
Use pseudonymous patient, slide and specimen identifiers, and keep the re-identification key outside the analysis workspace.
How do I install it?
Run npx skills add https://github.com/k-dense-ai/scientific-agent-skills --skill pathml.
Is it checked for security?
Its security audits show passes from Gen Agent Trust Hub, Socket and Snyk.
What are the best pathml alternatives?
Popular alternatives to pathml include content-research-writer, pdf and sherpa-onnx-tts. See all pathml alternatives