README.md

---
title: MedGemma Radiology Report Generator
emoji: 🧠
colorFrom: purple
colorTo: indigo
sdk: gradio
python_version: '3.10'
app_file: app.py
sdk_version: 5.39.0
---

# MedGemma Radiology Report Generator

This Hugging Face Space demonstrates the capabilities of Google's **MedGemma 4b-it** model for generating radiology reports from medical images.

Simply upload a medical image (such as a chest X-ray), and the model will return a detailed natural language report describing relevant findings.

---

### 🔬 How to Use

1. Upload a medical image (e.g., PNG, JPEG, X-ray) using the image input box.
2. The model will automatically generate a text-based radiology report.

> ⚠️ **Note:** The MedGemma model requires that the prompt includes an `<image>` token to correctly associate the image with the generated response. This is handled automatically by the app.

---

### 🧠 Model Information

- **Model:** `google/medgemma-4b-it`
- **Hugging Face Link:** [https://huggingface.co/google/medgemma-4b-it](https://huggingface.co/google/medgemma-4b-it)
- **Requirements:** You must request access to the model and provide a valid `HF_TOKEN` secret in your Space for it to work.

---

### ⚠️ Disclaimer

This application is intended for **research and demonstration purposes only**. It should **not** be used for clinical decision-making. All generated reports must be reviewed and validated by a qualified medical professional.

---

### ℹ️ Known Limitations

- DICOM files are not yet supported — please convert to PNG or JPEG before uploading.
- Model output can vary — results may include artifacts or hallucinated content.
- Performance is limited by available hardware in public Spaces. Consider running locally or on a private GPU Space for heavy usage.

---

Feel free to fork this Space or modify it to suit your own clinical AI research workflows!