Diagnostic Copilot

AI-powered chest X-ray analysis for diagnostic decision support

Medical Disclaimer

This system is for educational and research purposes only. It is NOT a substitute for professional medical diagnosis or treatment. All AI-generated analyses require review by qualified healthcare professionals.

---

What It Does

Diagnostic AI Copilot analyzes chest X-rays using deep learning and provides:

• 14 Pathology Detection: Pneumonia, Pneumothorax, Cardiomegaly, Effusion, and more
• Visual Explainability: Grad-CAM heatmaps showing exactly where AI focuses
• Auto-Generated Reports: Structured radiology report drafts
• Uncertainty Quantification: Confidence scores with uncertainty estimates
• DICOM Support: Works with medical imaging formats

---

🚀 Quick Start

Prerequisites

• Python 3.11+
• Node.js 18+
• Docker (optional)

Backend Setup

cd backend

# Create virtual environment
python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

# Install dependencies
pip install -r requirements.txt

# Run the server
uvicorn app.main:app --reload --host 0.0.0.0 --port 8000

Frontend Setup

cd frontend

# Install dependencies
npm install

# Run development server
npm run dev

With Docker

# Build and run all services
docker-compose up --build

# Access the app at http://localhost:3000
# API docs at http://localhost:8000/docs

---

🏗️ Architecture

diagnostic-copilot/
├── backend/                 # FastAPI backend
│   ├── app/
│   │   ├── ml/             # ML models (DenseNet121, Grad-CAM)
│   │   ├── routers/        # API endpoints
│   │   ├── schemas/        # Pydantic models
│   │   └── utils/          # Report generation, DICOM
│   └── data/
│       └── models/         # Model weights
│
├── frontend/                # Next.js frontend
│   ├── src/
│   │   ├── app/           # Pages
│   │   ├── components/    # React components
│   │   └── lib/           # API client
│
└── docker-compose.yml       # Container orchestration

---

Model Details

Architecture

• Base Model: DenseNet121 pre-trained on ImageNet
• Fine-tuning: ChestX-ray14 dataset (112,120 images)
• Output: Multi-label classification (14 pathologies)
• Explainability: Grad-CAM attention maps

Detectable Conditions

Condition	Severity
Pneumothorax	🔴 Severe
Mass	🔴 Severe
Pneumonia	🟡 Moderate
Cardiomegaly	🟡 Moderate
Effusion	🟡 Moderate
Edema	🟡 Moderate
Consolidation	🟡 Moderate
Nodule	🟡 Moderate
Atelectasis	🟢 Mild
Infiltration	🟢 Mild
Emphysema	🟢 Mild
Fibrosis	🟢 Mild
Pleural Thickening	🟢 Mild
Hernia	🟢 Mild

---

📡 API Reference

Analyze Image

POST /api/v1/analysis
Content-Type: multipart/form-data

# Parameters
- file: Image file (JPEG, PNG, DICOM)
- patient_age: Optional patient age
- patient_sex: Optional patient sex (M/F)
- clinical_history: Optional clinical notes
- generate_heatmaps: Generate Grad-CAM (default: true)

Get Report

GET /api/v1/analysis/{analysis_id}/report

Health Check

GET /health

Full API documentation available at /docs (Swagger UI).

---

Safety Features

1. Medical Disclaimers: Every output includes clear disclaimers
2. Uncertainty Quantification: Monte Carlo Dropout for confidence intervals
3. Physician Review Required: Explicit flags for human oversight
4. Audit Logging: Track all predictions for compliance
5. Input Validation: Check image quality before analysis

---

Performance Metrics

Metric	Value
Inference Time	~2-3 seconds (CPU)
Model Size	~30MB
Supported Formats	JPEG, PNG, DICOM

---

Contributing

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing)
3. Commit changes (git commit -m 'Add amazing feature')
4. Push to branch (git push origin feature/amazing)
5. Open a Pull Request

---

📄 License

MIT License - see LICENSE for details.

---

🙏 Acknowledgments

• NIH ChestX-ray14 Dataset
• Stanford CheXpert Dataset
• PyTorch and torchvision teams
• FastAPI and Next.js communities

---

Built with ❤️ for improving healthcare through AI

diagnostic-copilot