Spaces:
Runtime error
Runtime error
File size: 3,862 Bytes
f73e73c b587cc2 f73e73c b587cc2 f73e73c b587cc2 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 | ---
title: RadioFlow - AI Radiology Workflow Agent
emoji: 🩻
colorFrom: blue
colorTo: indigo
sdk: gradio
sdk_version: 4.19.0
app_file: app.py
pinned: true
license: cc-by-4.0
tags:
- medical
- radiology
- chest-x-ray
- medgemma
- hai-def
- healthcare
- agentic
- multi-agent
---
# RadioFlow: AI-Powered Radiology Workflow Agent
> **MedGemma Impact Challenge Submission**
> Targeting: Main Track + Agentic Workflow Prize
## Overview
RadioFlow is a multi-agent AI system that streamlines radiology workflows by processing chest X-rays through an orchestrated pipeline of specialized agents. Built with Google's Health AI Developer Foundations (HAI-DEF) models.
## Architecture
```
┌─────────────────────────────────────────────────────────────────┐
│ RADIOFLOW ORCHESTRATOR │
├─────────────────────────────────────────────────────────────────┤
│ │
│ ┌──────────────┐ ┌──────────────┐ ┌──────────────┐ │
│ │ Agent 1 │───▶│ Agent 2 │───▶│ Agent 3 │ │
│ │ CXR Analyzer │ │ Finding │ │ Report │ │
│ │ │ │ Interpreter │ │ Generator │ │
│ └──────────────┘ └──────────────┘ └──────────────┘ │
│ │ │ │
│ │ ▼ │
│ │ ┌──────────────┐ │
│ │ │ Agent 4 │ │
│ └─────────────────────────────▶│ Priority │ │
│ │ Router │ │
│ └──────────────┘ │
└─────────────────────────────────────────────────────────────────┘
```
## Agents
| Agent | Model | Function |
|-------|-------|----------|
| **CXR Analyzer** | CXR Foundation | Process chest X-ray, extract features, detect abnormalities |
| **Finding Interpreter** | MedGemma | Interpret visual findings into clinical language |
| **Report Generator** | MedGemma | Create structured radiology report |
| **Priority Router** | MedGemma | Assess urgency, route to care pathway |
## HAI-DEF Models Used
- **CXR Foundation**: [google/cxr-foundation](https://huggingface.co/google/cxr-foundation)
- **MedGemma**: [google/medgemma-4b-it](https://huggingface.co/google/medgemma-4b-it)
## Usage
1. Upload a chest X-ray image
2. (Optional) Add clinical context
3. Click "Analyze X-Ray"
4. View the generated report, priority assessment, and visualizations
## License
This project is submitted under CC BY 4.0 as required by the competition.
## Disclaimer
⚠️ **For demonstration purposes only. Not for clinical use.**
This AI system requires radiologist verification before any clinical decisions.
## Acknowledgments
- Google Health AI Developer Foundations team
- MedGemma Impact Challenge organizers
|