Oral Cancer Detection Demo

A production-ready Flask web application for oral cancer detection using deep learning.

🚀 Quick Start

1. Install Dependencies

pip install -r requirements.txt

2. Model Included ✅

Trained CNN model is included:

• File: model/oral_cancer_model.keras (43MB)
• Architecture: CNN with Transfer Learning
• Accuracy: ~94% on test data

The app will load this model automatically for real AI predictions.

3. Run the App

python app.py

4. Open Browser

Navigate to: http://localhost:5000

📁 File Structure

oral-cancer-detection-demo/
├── app.py                 # Main Flask application
├── requirements.txt       # Python dependencies
├── README.md             # This file
├── model/                # Model directory
│   └── oral_cancer_model.keras  # Trained CNN model (43MB)
├── uploads/              # Temporary upload directory
├── static/
│   ├── css/             # Stylesheets
│   └── js/              # JavaScript files
└── templates/           # HTML templates
    ├── index.html       # Upload page
    ├── result.html      # Results page
    └── about.html       # About page

🔧 Features

• ✅ Drag & drop image upload
• ✅ Real-time AI analysis
• ✅ Confidence score display
• ✅ Risk level assessment
• ✅ Medical recommendations
• ✅ Mobile responsive design
• ✅ REST API endpoint

🌐 API Usage

Endpoint: /api/predict

Method: POST

Request:

curl -X POST -F "image=@oral_image.jpg" http://localhost:5000/api/predict

Response:

{
  "success": true,
  "predicted_class": "Normal (Non-Cancerous)",
  "confidence": 92.45,
  "risk_level": "Very Low Risk",
  "recommendation": "✅ No significant indicators detected...",
  "model_loaded": true
}

🚦 Demo Mode vs Production

Demo Mode:

• No model file required
• Simulates realistic predictions
• Perfect for demonstrations
• Random confidence scores

Production Mode:

• Trained model included (43MB .keras file)
• Real CNN predictions
• No additional setup required

🏥 Medical Disclaimer

⚠️ This tool is for research and educational purposes only.

• Not a medical device
• Should not replace professional diagnosis
• Always consult healthcare professionals
• Results require expert verification

👨‍💻 Developer

Umar Ranginwala

• M.Sc AI/ML, Gujarat University
• Product Manager
• GitHub: @umarranginwala
• LinkedIn: linkedin.com/in/umarranginwala

📄 License

Academic research project - Gujarat University