How does PathML handle large-scale pathology datasets?

It utilizes HDF5 storage to efficiently manage tiles, masks, and metadata, providing unified structures optimized for high-performance machine learning workflows.

Does PathML include pre-trained models?

Yes, it includes support for popular architectures like HoVer-Net for simultaneous nucleus segmentation and classification, and HACTNet for hierarchical classification.

Can I use this skill for multiplexed imaging data?

Yes, PathML provides specialized tools for analyzing multiparametric imaging from platforms like CODEX, Vectra, and MERFISH, including cell segmentation with Mesmer.

Is stain normalization included in the preprocessing?

Yes, PathML features modular preprocessing pipelines that include Macenko and Vahadane stain normalization methods for H&E stained tissue images.

What imaging formats does PathML support?

PathML supports over 160 proprietary and open slide formats, including Aperio SVS, Hamamatsu NDPI, Leica SCN, DICOM, and OME-TIFF.

PathML Computational Pathology

Name: PathML Computational Pathology
Author: jimmc414

byjimmc414

•

324

•

数据科学与机器学习

Analyzes whole-slide images and multiparametric imaging data for advanced computational pathology and machine learning workflows.

PathML is a modular Python toolkit designed to streamline computational pathology by providing robust tools for processing large-scale whole-slide images (WSI) and multiplexed imaging data. It simplifies complex tasks such as stain normalization, nucleus segmentation, and tissue graph construction, allowing researchers and developers to build scalable pipelines for cancer research and spatial proteomics. Whether you are training deep learning models like HoVer-Net or analyzing high-dimensional CODEX data, PathML offers the specialized infrastructure needed to transform raw pathology slides into actionable biological insights within the Claude Code environment.

主要功能

01Support for 160+ slide formats including SVS, NDPI, and OME-TIFF

02Multiparametric support for CODEX and Vectra spatial proteomics

03324 GitHub stars

04Modular preprocessing for stain normalization and tissue detection

05Spatial graph construction for cell and tissue-level analysis

06Integrated ML models for nucleus segmentation and classification

使用场景

01Analyzing spatial proteomics and marker expression in multiplexed tissue data

02Building large-scale datasets for training and deploying pathology ML models

03Developing AI-driven diagnostic tools using whole-slide histopathology images

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add jimmc414/kosmos pathml

For use in Claude.ai and ChatGPT

Download Skill