Extract nuclei and cellular features from whole-slide pathology images.
npx skills add https://github.com/ovachiever/droid-tings --skill pathml-ovachiever
Please help me install this Agent Skill. Skill: pathml Source: https://github.com/ovachiever/droid-tings/tree/main/skills/pathml Command: npx skills add https://github.com/ovachiever/droid-tings --skill pathml-ovachiever
PathML provides a comprehensive toolkit for computational pathology, enabling automated loading, preprocessing, graph construction, and model deployment on whole-slide pathology images and multiparametric imaging data. It reduces manual, error-prone steps in digital pathology analyses and accelerates throughput from raw slides to ML-ready insights.
Use cases include large-scale slide analysis, spatial omics workflows, and training pathology ML pipelines on diverse datasets.
Install PathML and follow the quick start to load a WSI, apply a preprocessing pipeline, and explore downstream tasks like nucleus detection and graph construction. Example task: load a WSIs, run a tissue detection transform, and stitch tile outputs into a graph for downstream GNN analysis.
High-intent search queries and answers about installing and using this skill.
PathML automates nucleus detection and segmentation across 160+ WSI formats, applying stain normalization and feature extraction pipelines to generate analysis-ready cellular data from H&E slides and multiplex imaging modalities like CODEX and Vectra.
Yes, PathML loads and preprocesses WSIs across 160+ formats and supports multiparametric imaging including H&E, CODEX, Vectra, and MERFISH, enabling unified workflows across diverse histopathology and spatial proteomics datasets.
PathML constructs cell and tissue graphs from detected nuclei, capturing spatial relationships for downstream graph neural network analysis and enabling quantification of cellular neighborhoods and tissue architecture.
PathML integrates PyTorch-based models including HoVer-Net and HACTNet for nucleus detection, plus ONNX inference for deployment, enabling both model training on pathology datasets and scalable inference on large slide collections.
PathML uses HDF5 storage for efficient management of preprocessed pathology data, tile-based loading for memory-scalable WSI processing, and structured output formats supporting large-scale slide analysis and downstream ML pipelines.
PathML applies stain normalization, tissue mask generation, tile-based loading, and feature extraction as preprocessing transforms, standardizing diverse slide images into ML-ready formats for model training and analysis.
K-Dense-AIAutomate tissue detection and tile extraction from whole-slide images.
jackspaceAnalyze whole-slide pathology images and multiparametric imaging data for computational pathology workflows.
rubenslivLoad, preprocess, and analyze whole-slide images with modular pipelines.
jimmc414Processes computational pathology workflows with WSI loading, preprocessing, ML models, and spatial graphs.
