🧠 CARE-AD+

Computer-Aided Recognition of Alzheimer's Disease

---

An advanced AI-powered clinical decision support system for early Alzheimer's disease detection
Deep Learning • Explainable AI (XAI) • RAG-Enhanced LLM • Clinical Reports • Real-time Analytics

Features • Quick Start • Documentation • Demo

📋 Overview

CARE-AD+ (Computer-Aided Recognition of Alzheimer's Disease Plus) is a comprehensive, multi-modal AI system designed to assist healthcare professionals in early detection and diagnosis of Alzheimer's disease. The system combines state-of-the-art deep learning with explainable AI techniques and RAG-enhanced LLM to provide transparent, clinically-relevant insights.

🎯 Mission

Early detection of Alzheimer's disease is crucial for patient care planning and potential intervention. CARE-AD+ provides clinicians with AI-powered analysis of brain MRI scans, backed by visual explanations, medical knowledge retrieval, and natural language interpretations.

✨ Key Features

🏗️ System Architecture

📦 CARE-AD+ System
│
├── 🖥️ Frontend (React + Vite)
│   ├── 📊 Dashboard - Real-time statistics & charts
│   ├── 🔬 Prediction - MRI upload & analysis
│   ├── 📈 Results - Detailed findings with heatmaps
│   ├── 💬 Chat - RAG-enhanced AI Assistant
│   ├── 📄 Reports - PDF generation & download
│   └── ⚙️ Admin - System management
│
├── ⚡ Backend (FastAPI)
│   ├── 🔐 Authentication - JWT-based security
│   ├── 👤 Patients - Simplified CRUD (ID, Name, Age)
│   ├── 🧠 Predictions - ML inference pipeline
│   ├── 💬 Chat - LLM with RAG integration
│   ├── 📄 Reports - PDF generation
│   └── ⚙️ Admin - Training & metrics
│
├── 🤖 ML Pipeline (PyTorch)
│   ├── 📦 Dataset - Data loading & augmentation
│   ├── 🏗️ Model - EfficientNet/ResNet architecture
│   ├── 🏋️ Training - Complete training pipeline
│   └── 📊 Evaluation - Metrics & visualization
│
├── 🔍 XAI Services
│   └── 🔥 Grad-CAM - Visual explanations
│
├── 📚 RAG Pipeline
│   ├── 🏥 Medical Knowledge Base
│   ├── 🔎 Context Retrieval
│   └── 💡 Prompt Enhancement
│
└── 💬 LLM Service (Ollama)
    ├── 👨‍⚕️ Technical Mode - For clinicians
    └── 👤 Patient Mode - Simplified explanations

🏥 Classification Categories

🚀 Quick Start

Prerequisites

⚡ One-Click Setup (Windows)

# Just double-click:
QUICK_START.bat

This automatically:

• ✅ Creates Python virtual environment
• ✅ Installs all dependencies
• ✅ Pulls Ollama phi3 model
• ✅ Starts backend & frontend servers

🔧 Manual Installation

# 1. Clone repository
git clone https://github.com/abhishekk-y/COMPUTER-AIDED-RECOGNITION-OF-ALZHEIMER-DISEASE.git
cd COMPUTER-AIDED-RECOGNITION-OF-ALZHEIMER-DISEASE

# 2. Setup backend
cd backend
python -m venv venv
venv\Scripts\activate
pip install -r requirements.txt

# 3. Setup frontend
cd ../frontend
npm install

# 4. Setup Ollama
ollama pull phi3
ollama serve

# 5. Start servers
cd ..
start_app.bat

Access: http://localhost:3000

🎨 Tech Stack

Backend

Technology	Purpose
	REST API Framework
	Deep Learning
	Database ORM
	Local LLM

Frontend

Technology	Purpose
	UI Framework
	Build Tool
	Data Visualization

📁 Project Structure

COMPUTER-AIDED-RECOGNITION-OF-ALZHEIMER-DISEASE/
│
├── 📂 backend/
│   ├── 📂 app/
│   │   ├── main.py              # FastAPI application
│   │   ├── config.py            # Configuration
│   │   ├── 📂 routers/          # API endpoints
│   │   ├── 📂 services/         # Business logic
│   │   │   ├── ml_service.py    # ML inference
│   │   │   ├── xai_service.py   # Grad-CAM
│   │   │   ├── llm_service.py   # LLM integration
│   │   │   ├── rag_service.py   # RAG pipeline
│   │   │   └── report_service.py # PDF generation
│   │   └── 📂 models/           # Database models
│   ├── 📂 ml/
│   │   ├── model.py             # CNN architecture
│   │   ├── dataset.py           # Data loading
│   │   ├── train.py             # Training pipeline
│   │   └── evaluate.py          # Evaluation
│   └── requirements.txt
│
├── 📂 frontend/
│   ├── 📂 src/
│   │   ├── 📂 pages/            # React pages
│   │   ├── 📂 components/       # Reusable components
│   │   ├── 📂 services/         # API client
│   │   └── 📂 styles/           # CSS
│   └── package.json
│
├── 📂 assets/                   # Project assets
├── QUICK_START.bat              # One-click setup
├── setup_ollama.bat             # Ollama setup
├── train_model.bat              # Model training
├── INSTALLATION.md              # Installation guide
├── OLLAMA_GUIDE.md              # LLM + RAG guide
└── README.md                    # This file

📚 Documentation

🧠 Model Training

# Quick training
train_model.bat

# Custom training
cd backend
python -m ml.train --dataset ../archive/combined_images --epochs 50

🤖 RAG Pipeline

The system includes a Retrieval-Augmented Generation pipeline that enhances LLM responses with medical knowledge:

• Medical Knowledge Base: CDR staging, biomarkers, treatments
• Context Retrieval: Automatic relevant knowledge extraction
• Prompt Enhancement: Evidence-based medical facts
• Clinical Guidelines: Recommendations per disease stage

See OLLAMA_GUIDE.md for details.

🐳 Docker Deployment

docker-compose up -d

Services:

• Backend: http://localhost:8000
• Frontend: http://localhost:3000
• Ollama: http://localhost:11434

👥 Default Credentials

⚠️ Change in production!

📊 Model Performance

🤝 Contributing

1. Fork the repository
2. Create feature branch: git checkout -b feature/AmazingFeature
3. Commit changes: git commit -m 'Add AmazingFeature'
4. Push: git push origin feature/AmazingFeature
5. Open Pull Request

📝 License

MIT License - see LICENSE file.

⚠️ Medical Disclaimer

IMPORTANT: CARE-AD+ is a clinical decision support tool. It is NOT intended to replace professional medical judgment, diagnosis, or treatment. All predictions should be reviewed by qualified healthcare professionals.

🙏 Acknowledgments

• Academic Guidance: University project supervision
• Open Source: PyTorch, FastAPI, React communities
• Medical Research: Alzheimer's disease research community