Brain Tumour MRI Detection using Deep Learning: Real-Time AI Classification System

🔍 Project Overview

Our Brain Tumour MRI Detection System is a cutting-edge medical AI platform designed to assist clinicians in the rapid and accurate classification of brain tumours from MRI scans. Leveraging advanced deep learning models, this project demonstrates the transformative potential of AI in neuro-oncology diagnostics, streamlining workflows and supporting expert decision-making.

🏆 Key Achievements

• 44 - class fine-grained tumour classification
• Real-world validation accuracy: DenseNet: 55.0% Xception: 61.9% MobileNet: 61.9%
• Fast inference : Results in seconds per scan
• Multi-model ensemble : Robust, comparative analysis
• Clinically relevant dataset : 641 MRI images, 44 tumour types

🎥 Project Explanation Video

📌 Video Highlights

• Live MRI scan classification demo
• Deep learning architecture walkthrough
• Performance metrics and clinical insights

🛠️ Technical Stack

Frontend

• Django Web Framework
• HTML5 / CSS3 / JavaScript
• Responsive UI Design
• Medical-grade interface

Backend

• Python 3.8+
• PyTorch Deep Learning
• SQLite Databasee
• RESTful API endpoints

AI Models

• DenseNet (customized for MRI)
• Xception
• MobileNet
• Real-time inference pipeline

📂 Project Structure: Breast Cancer Segmentation System

📦 BRAIN TUMOUR MRI DETECTION
├── 📁 userapp/
│   ├── views.py          # User dashboard & detection logic
│   ├── models.py         # User data models
│   └── templates/        # User-facing HTML templates
├── 📁 adminapp/
│   ├── views.py          # Admin dashboard & analytics
│   └── templates/        # Admin HTML templates
├── 📁 mainapp/
│   ├── settings.py       # Django settings
│   ├── urls.py           # URL routing
│   └── wsgi.py           # WSGI config
├── 📁 Brain_Tumour_MRI/
│   ├── models/           # Deep learning model scripts
│   │   ├── densenet.py
│   │   ├── xception.py
│   │   └── mobilenet.py
│   ├── utils/            # Preprocessing & evaluation
│   │   ├── preprocessing.py
│   │   └── evaluation.py
│   └── data/             # MRI images & labels
│       ├── images/
│       └── labels/
├── 📁 Database
│   ├── db.sqlite3        # SQLite database
│   └── media/uploads/    # User-uploaded MRI scans
└── 📁 Assets & Environment
    ├── assets/images/    # Static images
    ├── assets/css/       # Stylesheets
    ├── assets/js/        # JavaScript files
    └── env/              # Python virtual environment

📊 System Performance

Metric	DenseNet	Xception	MobileNet
Accuracy	55.0%	61.9%	61.9%
Processing Time	2s	2s	2s
Input Size	224x224x3	224x224x3	224x224x3
Model Size	30MB	25MB	18MB

✨ Core Features

• Real-Time MRI Analysis : Instant tumour classification from uploaded scans
• Multi-Model Comparison: DenseNet, Xception, and MobileNet for robust results
• Medical-Grade Interface: Designed for clinical usability and trust
• Analytics Dashboard: Visualizes predictions, model performance, and class distribution
• Comprehensive Class Coverage: 44 distinct brain tumour typesg

💡 Why This Matters

• Brain tumour detection can save lives. Our system empowers clinicians with AI-driven insights, supporting early diagnosis and personalized treatment planning.

🚀 Ready to Build Something Similar?

Need a custom AI solution? Let’s build a cutting-edge system together.