Can I use PathML for CODEX and multiplexed imaging?

Yes, PathML includes specialized classes and workflows for analyzing multiparametric data from platforms like CODEX, Vectra, and MERFISH, including marker quantification.

How does PathML handle very large pathology datasets?

PathML utilizes HDF5 storage structures to efficiently manage tiles, masks, and metadata, making it suitable for large-scale machine learning workflows.

What slide formats does the PathML skill support?

PathML supports over 160 proprietary formats, including common ones like Aperio SVS, Hamamatsu NDPI, Leica SCN, Zeiss ZVI, DICOM, and OME-TIFF.

Does this skill include pre-trained machine learning models?

Yes, it provides built-in support for model architectures like HoVer-Net for nucleus segmentation and HACTNet for hierarchical classification.

Is PathML better than histolab for simple tasks?

While PathML is a full-featured toolkit, for very simple tile extraction from H&E slides, the documentation suggests that histolab might be a simpler alternative.

PathML Computational Pathology

Name: PathML Computational Pathology
Author: Yezez9

byYezez9

0•

데이터 과학 및 ML

Enables advanced whole-slide image analysis, nucleus segmentation, and machine learning workflows for computational pathology.

PathML is a comprehensive Python toolkit designed for end-to-end processing of whole-slide pathology images within the Claude Code environment. It provides specialized capabilities for loading over 160 proprietary slide formats, performing stain normalization, constructing spatial tissue graphs, and training deep learning models like HoVer-Net and HACTNet. This skill is ideal for researchers and developers working on spatial proteomics (CODEX, Vectra), automated histology pipelines, and large-scale pathology data management using HDF5 storage, bridging the gap between raw medical imaging and actionable machine learning insights.

주요 기능

01Spatial proteomics analysis for CODEX, Vectra, and multiplex IF data

020 GitHub stars

03Unified loading for 160+ proprietary WSI formats including SVS, NDPI, and OME-TIFF

04Modular preprocessing pipelines for stain normalization and tissue detection

05Advanced nucleus segmentation and classification using HoVer-Net and HACTNet

06Efficient large-scale data management using HDF5 for tiles and metadata

사용 사례

01High-plex spatial analysis and marker quantification of CODEX imaging data

02Training and deploying hierarchical cell-type classification models on tissue graphs

03Automated nucleus detection and feature extraction from H&E stained slides

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add yezez9/research-agent pathml

For use in Claude.ai and ChatGPT