🎯 NeuroSight - Brain Tumor Classification System

📖 Overview

NeuroSight is a web-based application that uses deep learning to classify brain tumors from MRI scans. It can identify three types of brain tumors:

• Glioma
• Meningioma
• Pituitary Tumor

✨ Features

• User-friendly web interface
• Real-time image processing
• High-accuracy tumor classification
• Confidence rate display for each tumor type
• Responsive design for both desktop and mobile devices

🛠️ Technical Stack

Component	Technologies
Frontend	HTML, TailwindCSS, JavaScript
Backend	Flask (Python)
Machine Learning	TensorFlow/Keras
Image Processing	PIL, NumPy

🚀 Installation

1. Clone the repository:

   git clone https://github.com/penzulo/neurosight.git
   cd neurosight

2. Install required packages:

   pip install flask tensorflow pillow numpy werkzeug

3. Download the pre-trained model: Place brain_tumor_classifier.keras in the root directory.

💡 Usage

1. Start the server:

   python server.py

2. Open a web browser and navigate to:

   http://localhost:5000
   

3. Upload an MRI scan on the home page and view classification results.

🧠 Model Information

This system leverages a Convolutional Neural Network (CNN) trained on brain MRI scans. Key details:

• Image Format: 512x512 RGB
• Output: Classification probabilities for tumor types
• Automatic Preprocessing: Resize and preprocess uploaded images

🌐 API Endpoints

• GET / - Home page
• GET /upload - Upload page
• POST /predict - Processes uploaded images and returns predictions
• GET /results/<filename> - Displays the results page

⚠️ Error Handling

The application includes error handling for:

• Invalid file uploads
• Missing files
• Server processing errors
• Model prediction errors

🤝 Contributing

Contributions are welcome! Feel free to fork the project and submit a pull request.

---