Can I use this for CODEX data?

Yes, PathML includes specialized capabilities for multiparametric imaging data like CODEX, Vectra, and MERFISH for spatial proteomics analysis.

How does PathML handle large image files?

It uses tile-based processing and efficient HDF5 storage structures to manage the massive data sizes typical of whole-slide images.

What formats does PathML support?

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

Does it include pre-trained models?

Yes, it provides integration for specialized pathology models like HoVer-Net for nucleus segmentation and HACTNet for cell-type classification.

PathML Pathology Toolkit

Name: PathML Pathology Toolkit
Author: Zehong-Wang

byZehong-Wang

0•

Data Science & ML

Analyzes whole-slide pathology images and multiparametric imaging data using a comprehensive computational pathology toolkit.

PathML is a modular Python framework designed for end-to-end computational pathology workflows, enabling researchers and developers to process over 160 whole-slide image formats. It facilitates advanced tasks such as H&E stain normalization, nucleus segmentation using HoVer-Net, spatial graph construction for tissue analysis, and the processing of multiparametric data from platforms like CODEX and Vectra. By providing a unified interface for complex image pyramids and spatial proteomics, PathML streamlines the transition from raw histopathology slides to actionable machine learning models and spatial insights.

Key Features

01Deep learning integration for nucleus segmentation and classification (HoVer-Net, HACTNet)

02Spatial graph construction for cellular relationship and tissue analysis

03Support for 160+ whole-slide image formats including SVS, NDPI, and DICOM

04Modular preprocessing pipelines for stain normalization and tissue detection

050 GitHub stars

06Multiparametric imaging support for CODEX, Vectra, and spatial proteomics

Use Cases

01Analyzing spatial proteomics and marker expression from multiplex immunofluorescence data

02Building large-scale machine learning datasets for digital pathology research

03Automating nucleus detection and classification in H&E stained tissue slides

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add zehong-wang/kosmos pathml

For use in Claude.ai and ChatGPT

Download Skill