CardioIntel AI is a state-of-the-art, full-stack medical intelligence platform designed for high-accuracy heart disease risk assessment. By combining an ensemble of advanced machine learning models with the generative power of Google Gemini, CardioIntel transforms raw clinical data into actionable, patient-centric health insights.

• Ensemble Engine: Combines Random Forest, XGBoost, and LightGBM using soft-voting for superior predictive reliability.
• Explainable AI (XAI): Native integration of SHAP and LIME to provide transparency into why the model assigned a specific risk level.
• Clinical Gauge: Real-time visualization of risk probability with calibrated confidence intervals.

• Gemini-Powered Narratives: Automatically translates complex biometric patterns into natural, medically-grounded clinical summaries.
• Risk Multiplication Logic: Explains the interaction between different biomarkers (e.g., how elevated BP multiplies the risk of high cholesterol).
• Urgency Assessment: Context-aware urgency levels (Immediate, Moderate, Routine) based on the severity of identified clinical flags.

• Contextual Chat: A dedicated AI assistant that knows your specific clinical history and answers questions about diet, routine, and risk factors.
• Actionable Advice: Provides personalized lifestyle directives, including salt reduction strategies, exercise approval checks, and dietary improvements.

• Auto-Generated Reports: Professional, print-ready PDF-style clinical reports featuring patient profiles, biometric trends, and diagnostic narratives.
• Patient History Tracking: Persistent session management using MongoDB to track health progress over time.

graph TD
    subgraph "Frontend (React + Vite)"
        UI[User Console] --> Predict[Diagnostic Form]
        UI --> Assistant[AI Health Partner]
        UI --> Report[Clinical Documentation]
    end

    subgraph "Cloud Intelligence"
        Gemini[Google Gemini API] --> Narratives[Clinical Summaries]
        Gemini --> Chat[Interactive Support]
    end

    subgraph "Backend (Flask)"
        API[Inference API] --> ML[Ensemble Model]
        API --> DB[(MongoDB Atlas)]
        ML --> XAI[SHAP / LIME]
    end

    Predict -->|Clinical Features| API
    Assistant -->|Natural Language| Gemini
    API -->|Risk Data| UI
    Gemini -->|Interpretations| UI

• Models: Scikit-Learn (RF), XGBoost, LightGBM (Ensemble)
• GenAI: Google Gemini (via google-generativeai)
• Interpretability: SHAP, LIME
• Preprocessing: Pandas, NumPy, Scikit-Learn Scalers

• Framework: Flask 3.0
• Database: MongoDB (Atlas)
• Serialization: Joblib
• Environment: Python-dotenv

• Framework: React 18 (Vite)
• Styling: Tailwind CSS (Premium Dark Mode support)
• Data Visuals: Framer Motion (Animations), SVG-based gauges
• Communication: Axios

• Python 3.10+
• Node.js 18+
• MongoDB Atlas Cluster
• Google Gemini API Key

cd BackEnd
python -m venv venv
venv\Scripts\activate  # Windows
pip install -r requirements.txt

# Configure .env
# MONGO_URI=your_mongodb_uri
# GEMINI_API_KEY=your_gemini_key

python app.py

cd FrontEnd
npm install
npm run dev

CardioIntel features a custom-built premium design system:

• Logo: Pure SVG implementation with CSS pulse animations.
• Color Palette: Deep Indigo, Emerald Success, and Rose Danger for clear clinical signaling.
• Glassmorphism: Modern UI layers with backdrop blurs and subtle drop shadows.

This project is licensed under the MIT License.

Disclaimer: CardioIntel AI is intended for educational and clinical support research only. It is NOT a substitute for professional medical advice, diagnosis, or treatment. Always consult with a licensed healthcare professional.