pathml

To analyze whole-slide images for computational pathology, you can load over 160 slide formats, generate tiles, and apply preprocessing techniques like stain normalization and nucleus detection within a single workflow environment.

Yes, you can process multiplex imaging data from CODEX, Vectra, and MERFISH for spatial proteomics and transcriptomics to identify immune cell populations and map their spatial relationships in tissue microenvironments.

You can train and deploy deep learning models like HoVer-Net for nucleus segmentation and classification by utilizing built-in preprocessing pipelines and spatial graph construction directly on your whole-slide imaging datasets.

Yes, the computational pathology toolkit integrates directly with AnnData to support single-cell analysis, allowing you to transition seamlessly from whole-slide image preprocessing and nucleus detection to downstream spatial graph analysis.

The best way to map spatial relationships in a tumor microenvironment is to process multiplexed immunofluorescence imaging data, build spatial graphs, and quantify distinct immune cell populations using integrated machine learning models.

Yes, normalizing staining is a critical preprocessing step before running nucleus detection on whole-slide images, ensuring that color variations across different slides do not negatively impact the accuracy of downstream deep learning models.