pathml

Build computational pathology workflows by loading whole-slide images, applying modular preprocessing pipelines, constructing spatial tissue graphs, and training machine learning models. PathML consolidates these steps from image loading through to model deployment.

Analyze CODEX and Vectra multiplex imaging data using modular preprocessing pipelines, spatial graph construction, and built-in machine learning models. PathML supports these specific multiparametric formats alongside MERFISH for cellular and tissue interaction analysis.

Perform nucleus segmentation and tissue graph construction within a single pipeline using PathML. It provides built-in machine learning models for nucleus detection and spatial graph analysis tools to map cellular interactions directly from preprocessed image tiles.

Staining normalization is supported across 160+ image formats for whole-slide and multiplex imaging. PathML enforces reproducible workflows through a modular transform catalog, ensuring consistent preprocessing before downstream graph analytics and model training.

Manage large multiparametric imaging datasets efficiently using HDF5-based data management and batch processing. PathML utilizes this approach to handle high-volume whole-slide images and multiplex data, enabling smooth machine learning model training and deployment.