What problem does it solve?
This Skill automates the complex and time-consuming process of extracting and preprocessing tiles from large whole slide images (WSIs) in digital pathology, making them ready for analysis or deep learning.
Core Features & Use Cases
- Tissue Detection & Masking: Automatically identify and segment tissue regions, filtering out background and artifacts.
- Tile Extraction: Extract uniform tiles from WSIs using various strategies (random, grid, score-based).
- Image Preprocessing: Apply filters for stain normalization, artifact removal, and quality control.
- Use Case: Prepare a dataset of 512x512 pixel tiles from hundreds of H&E stained prostate cancer slides, ensuring each tile contains at least 80% tissue, for training a deep learning model to detect tumor regions.
Quick Start
Use the histolab skill to extract 100 random 512x512 tiles from the slide file 'slide.svs' and save them to the 'output/tiles/' directory.