WSI Tissue Tiler

This project processes whole slide images (WSI) to extract tissue tiles using parallel processing. It utilizes the SlideProcessor class to handle the image processing efficiently.

Getting Started

These instructions will guide you through the setup and execution of the WSI Tissue Tiler, including the use of the Docker container.

Prerequisites

• Docker
• Python 3.8 or later
• Git

Method 1: Using Docker

If you want to start quickly, you can pull the ready-made Docker image and run it.

Using Pre-Built Docker

Pull the Docker image from Docker Hub with the following command:

docker pull vatsalpatel18/wsi_tissue_tiler:latest

Method 2: Manual Setup

Clone the Repository

First, clone this repository to your local machine:

git clone git@github.com:VatsalPatel18/wsi_tissue_tiler.git
cd wsi_tissue_tiler

Build and Run Docker Container

Build the Docker Image

docker build -t wsi_tissue_tiler:latest .

Run the Docker container: For Jupyter lab:

docker run -p 7878:7878 -v /path/to/WSI:/app/WSI -v /path/to/outputs:/app/outputs wsi_tissue_tiler

For Unix Systems:

docker run -v /path/to/WSI:/app/WSI -v /path/to/outputs:/app/outputs wsi_tissue_tiler process_wsi -d /app/WSI -o /app/outputs -w 60

Replace /path/to/WSI and /path/to/outputs with the actual paths to your data and output directories on your host machine.

Parameters

-d, --directory: Directory containing whole slide image files. -o, --output_dir: Directory to save the processed tiles. -t, --tile_size: Size of the tile (default: 1024). -v, --overlap: Overlap of tiles (default: 0). -th, --tissue_threshold: Threshold for tissue detection (default: 0.65). -w, --max_workers: Maximum number of worker threads/processes (default: 30).