Brain Tumor Detection System Using Machine Learning

• Secure admin login system with separate admin session
• Admin dashboard with total users, MRI uploads, predictions, and tumor vs no-tumor statistics
• User management system with edit, activate, deactivate, reset password, and delete options
• View and manage all uploaded brain MRI images
• Access all prediction records with filters and delete functionality
• Dataset management to upload training images into tumor and non-tumor folders
• Train and retrain the brain tumor detection machine learning model from the admin panel
• View model accuracy, classification report, and confusion matrix
• Generate downloadable HTML reports such as user-wise, date-wise, and overall system reports
• Manage user feedback and contact messages
• Update system settings like allowed file types, upload size, model path, and admin credentials

• User registration with name, email, username, password, and security question
• Secure login and logout using email or username
• Forgot password option using security answer verification
• User profile management with edit profile and change password functionality
• Upload brain MRI scan images in JPG, JPEG, and PNG formats
• Preview MRI images before prediction
• Run brain tumor prediction using the trained machine learning model
• View prediction result with class label, confidence score, timestamp, and image details
• Download prediction reports in PDF and TXT format
• Access prediction history with filters by date, tumor status, and label
• Delete personal prediction records
• User dashboard with total scans and recent prediction history

• Built with Flask web framework for a lightweight and scalable backend
• Uses SQLite database for storing users, predictions, feedback, and system settings
• Implements secure password hashing with Werkzeug
• Supports machine learning model training and inference locally
• Feature extraction using OpenCV, including HOG, histogram, image statistics, and edge detection
• Uses RandomForestClassifier from scikit-learn for classification
• Supports both binary classification and multiclass brain tumor detection
• Generates downloadable reports and visual model performance analytics
• Responsive frontend built with HTML, CSS, and Bootstrap 5
• Ideal for final year projects, portfolio websites, and machine learning project showcases

Step 1: Install Dependencies

Open the project folder and run:

pip install -r requirements.txt

Step 2: Seed Demo Data

To create the database, sample users, sample predictions, and demo model, run:

python seed_data.py

Step 3: Start the Flask Server

Run the application using:

python run.py

Step 4: Open in Browser

Visit:

http://127.0.0.1:5000

Default Admin Login

Username: admin
Password: Admin@123

Default User Login

Username: alice_c
Password: User@123