🧠 ClarityNet | Brain Tumor Detection with CNN

This project is a deep learning-based web application for brain tumor detection from MRI scans. It uses a ResNet50 model and provides an interactive Streamlit interface for users to upload MRI images, get predictions, visualize model attention (Grad-CAM), and find nearby hospitals using OpenStreetMap and OpenRouteService.

✨ Features

• 📤 Upload MRI scans and get instant predictions (tumor/no tumor)
• 📊 Probability distribution for model confidence
• 🔥 Grad-CAM visualization to highlight regions influencing the model's decision
• 📁 Batch analysis: Upload multiple scans and view trends
• 💡 Recommendations based on results
• 📄 Downloadable PDF/Text reports
• 🏥 Find nearby hospitals (using OSM/Overpass/ORS)
• 🗺️ Route map to the nearest hospital (in-app and Google Maps link)
• 📚 Resource links for further reading/support

📂 Project Structure

Brain Tumor Detection with CNN/
├── app/
│   ├── streamlit_app.py      # Main Streamlit app
│   ├── ui.py                # UI components
│   ├── gradcam.py           # Grad-CAM utility
│   ├── main.py              # Model loading and prediction
├── data/
│   ├── yes/                 # MRI images with tumor
│   └── no/                  # MRI images without tumor
├── model/
│   ├── brain_tumor_model_resnet50_final.h5  # Trained model
│   └── training_history.png # Training history plot
├── utils/
│   ├── preprocess.py        # Image preprocessing utilities
├── requirements.txt         # Python dependencies
├── README.md                # Project documentation

🚀 Setup & Usage

1. Install dependencies:

   pip install -r requirements.txt

2. Run the app:

   streamlit run app/streamlit_app.py

3. Upload MRI images and explore the results in your browser.

📝 Notes

• 🌍 The app uses OpenStreetMap (OSM) and OpenRouteService (ORS) for hospital search and routing. No API key is needed for OSM; a free key is used for ORS.
• 🧹 Temporary files and compiled Python files are cleaned up for production.
• 🗑️ The test.py and tempCodeRunnerFile.py files have been removed as they are not needed.

Screenshots

📄 License

See LICENSE for details.