README.md

metadata

title: AI Diagnostics Agent - Early Cancer Discovery
emoji: 🏥
colorFrom: purple
colorTo: pink
sdk: gradio
sdk_version: 5.49.1
app_file: app.py
pinned: false

🏥 AI Diagnostics Agent: Early Cancer Discovery (Demo)

⚠️ Research Demo Only — Not for Clinical Use

---

🌍 Overview

AI Diagnostics (POC) is a research project designed to demonstrate early cancer risk screening using AI-driven imaging + lab report synthesis.

It integrates:

• 🖼️ Imaging Agent — Detects lung abnormalities (mass, opacity, nodules) from chest X-rays.
• 🧪 Lab Agent — Parses tumor markers (PSA, CA-125, AFP) and highlights abnormal results.
• 🧠 AI Coordinator — Generates diagnostic summaries powered by Moonshot.ai with a local T5 fallback for offline use.
• 🔥 Heatmap Overlay — Highlights model-identified lung regions for interpretability.

---

⚙️ Features

Agent	Description
Imaging Agent	TorchXRayVision + PyTorch for lung-related cancer indicators
Lab Agent	Extracts structured tumor marker insights
Coordinator	Combines both into concise AI summaries (Moonshot.ai LLM)
Fallback AI	Local text generation using T5-small if network/API fails
Heatmap Overlay	Visual interpretability of X-ray activations
Gradio UI	Easy drag-and-drop web interface for demos

---

🧠 How It Works

1. Upload or select a sample chest X-ray.
2. Paste lab or MRI/CT report text.
3. The AI analyzes the image + text.
4. Results are combined into a natural-language diagnostic summary.

---

🚀 Run Locally

git clone https://github.com/<your-org>/ai_diagnostics.git
cd ai_diagnostics
pip install -r requirements.txt
python app.py