MedGemma Pathology

Analyze whole slide digital pathology images with the MedGemma 1.5 model via a Streamlit web app. Runs on Apple Silicon using MLX.

Disclaimer: Educational tool only. Not intended for medical diagnosis or treatment.

Requirements

• Python 3.12+
• Apple Silicon Mac (M1 or later)
• Hugging Face account with access to google/medgemma-1.5-4b-it

Setup

Install dependencies:

uv sync

Create a .env file with your Hugging Face token:

HF_TOKEN=hf_your_token_here

Usage

Start the app:

uv run streamlit run streamlit_app.py

1. Upload a pathology image (PNG, JPEG, or TIFF).
2. Adjust patch size and max patches in the sidebar.
3. Click Analyze with MedGemma to run inference.

How It Works

1. The uploaded image is split into non-overlapping patches (default 896px).
2. Background patches are filtered out.
3. Tissue patches are encoded and sent to MedGemma for analysis.
4. The model returns a descriptive summary of the tissue.

Tests

Run all tests:

uv run pytest