PathML Computational Pathology

Acerca de

PathML offers a comprehensive toolkit for digital pathology workflows, enabling researchers and developers to process, analyze, and model high-resolution whole-slide images (WSI). It streamlines complex tasks such as stain normalization, nucleus segmentation, and tissue graph construction across over 160 proprietary slide formats. By integrating specialized deep learning models like HoVer-Net and supporting multiparametric data from platforms like CODEX and Vectra, it facilitates sophisticated spatial proteomics and histopathology analysis within a modular, reproducible framework.

Características Principales

• Pre-built deep learning model integration for HoVer-Net and HACTNet
• Supports 160+ slide formats including Aperio SVS, Hamamatsu NDPI, and OME-TIFF
• Multiparametric imaging workflows for CODEX and multiplex immunofluorescence quantification
• Spatial graph construction for cell-type relationship and tissue-level analysis
• Modular preprocessing pipelines for Macenko/Vahadane stain normalization and tissue detection
• 2,066 GitHub stars

Casos de Uso

• Training and deploying machine learning models for automated nucleus segmentation
• Preprocessing and standardizing large-scale H&E stained tissue datasets
• Quantifying marker expression and cellular spatial relationships in multiplexed imaging