Health Wellness App

A comprehensive health and wellness application that combines AI/ML technologies with modern mobile development to provide personalized health insights and medical assistance.

🎯 Project Overview

This application is an innovative health platform that leverages artificial intelligence and machine learning to provide users with:

• AI-Powered Health Chatbot: Intelligent medical assistance using BioBERT model
• Disease Prediction: Machine learning model for diabetes prediction
• Secure Authentication: User registration and login system
• Modern UI/UX: Clean, intuitive Flutter-based mobile interface

🏗️ Architecture

The application follows a hybrid architecture:

• Frontend: Flutter mobile application with cross-platform support
• Backend: Python Flask API with AI/ML capabilities
• Database: Supabase for user authentication and data storage
• AI Models: BioBERT for medical Q&A, Random Forest for disease prediction

✨ Features

1. User Authentication

• Secure user registration and login
• Password-based authentication via Supabase
• Session management and auto-logout

2. AI Medical Chatbot

• BioBERT Integration: Uses the specialized biomedical BERT model (dmis-lab/biobert-v1.1)
• Real-time Medical Q&A: Provides accurate answers to health-related queries
• Context-Aware Responses: Fetches relevant medical information from web sources
• Reference Citations: Includes links to authoritative medical sources

3. Disease Prediction Model

• Diabetes Risk Assessment: ML model trained on medical parameters
• Input Parameters:
  • Number of pregnancies
  • Glucose levels
  • Blood pressure
  • Skin thickness
  • Insulin levels
  • BMI (Body Mass Index)
  • Diabetes pedigree function
  • Age
• Prediction Output: Risk assessment with probability scores

4. Modern Mobile Interface

• Clean, responsive Flutter UI
• Cross-platform compatibility (Android, iOS, Web, Desktop)
• Intuitive navigation and user experience
• Material Design principles

🛠️ Technology Stack

Frontend

• Flutter: Cross-platform mobile development
• Dart: Programming language for Flutter
• Material Design: UI/UX framework
• HTTP: API communication

Backend

• Python: Core backend language
• Flask: Web framework
• Flask-CORS: Cross-origin resource sharing
• Transformers: Hugging Face library for BioBERT
• Scikit-learn: Machine learning framework
• Pickle: Model serialization
• Requests: HTTP client for external APIs

Database & Authentication

• Supabase: Backend-as-a-Service platform
• PostgreSQL: Database (via Supabase)
• JWT: Token-based authentication

AI/ML Models

• BioBERT: Biomedical language model for Q&A
• Random Forest: Disease prediction model
• SerpAPI: Web search integration for medical context

📱 App Structure

lib/
├── main.dart              # App entry point and routing
├── home_page.dart         # Dashboard with navigation
├── login_page.dart        # User authentication
├── signup_page.dart       # User registration
├── chatbot_page.dart      # AI medical chatbot interface
├── ml_model_page.dart     # Disease prediction form
├── profile_page.dart      # User profile management
├── app.py                 # Flask backend API
├── BIO_GPT.ipynb         # Model training notebook
└── random_forest_model.pkl # Trained ML model

🚀 Getting Started

Prerequisites

• Flutter SDK (>=3.4.3)
• Python 3.8+
• Supabase account
• SerpAPI account (for web search functionality)

Installation

1. Clone the repository

   git clone <repository-url>
   cd health

2. Install Flutter dependencies

   flutter pub get

3. Install Python dependencies

   pip install -r requirements.txt

4. Environment Setup Create a .env file in the root directory:

   SUPABASE_URL=your_supabase_url
   SUPABASE_ANON_KEY=your_supabase_anon_key
   SERPAPI_KEY=your_serpapi_key

5. Run the Backend API

   python app.py

6. Run the Flutter App

   flutter run

🔧 Configuration

Supabase Setup

1. Create a new Supabase project
2. Set up authentication with email/password
3. Add your Supabase URL and anon key to the .env file

SerpAPI Setup

1. Sign up for a SerpAPI account at https://serpapi.com
2. Get your API key from the dashboard
3. Add the SERPAPI_KEY to your .env file

📊 Machine Learning Models

BioBERT Chatbot

• Model: dmis-lab/biobert-v1.1
• Purpose: Medical question answering
• Features: Context-aware responses, reference citations
• Max Length: 512 tokens

Diabetes Prediction Model

• Algorithm: Random Forest Classifier
• Features: 8 medical parameters
• Output: Binary classification (diabetic/non-diabetic) with probabilities
• Accuracy: Trained on standard diabetes dataset

🔐 Security Features

• Secure user authentication via Supabase
• Environment variable protection for sensitive keys
• CORS configuration for API security
• Input validation and error handling
• Session management with automatic logout

📚 API Endpoints

/chat (POST)

• Purpose: Medical chatbot interaction
• Input: {"message": "user_query"}
• Output: AI response with references

/predict (POST)

• Purpose: Disease prediction
• Input: Medical parameters array
• Output: Prediction result with probability

🎨 UI/UX Features

• Responsive Design: Adapts to different screen sizes
• Material Design: Follows Google's design principles
• Intuitive Navigation: Easy-to-use interface
• Loading States: Visual feedback during API calls
• Error Handling: User-friendly error messages

🔄 Development Workflow

1. Frontend Development: Flutter for cross-platform mobile app
2. Backend Development: Python Flask for AI/ML APIs
3. Model Training: Jupyter notebooks for ML model development
4. Testing: Comprehensive testing for both frontend and backend
5. Deployment: Cloud deployment with CI/CD pipeline

🌟 Future Enhancements

• Additional ML models for different diseases
• Integration with wearable devices
• Telemedicine consultation features
• Personalized health recommendations
• Health data visualization
• Multi-language support
• Offline mode capabilities

📄 License

This project is part of BITS CSA (Computer Science and Applications) coursework and is intended for educational purposes.

🤝 Contributing

This is an academic project. For any suggestions or improvements, please reach out to the development team.

📞 Support

For technical support or queries related to this project, please contact the development team through the appropriate academic channels.

---

This application represents the intersection of healthcare technology and artificial intelligence, demonstrating how modern mobile development can be combined with advanced AI/ML models to create meaningful health solutions.