Heart Disease Prediction & Patient Profiling System

Advanced ML/DL system for heart disease prediction with patient segmentation

---

Overview

This project combines predictive analytics with patient clustering to provide comprehensive heart disease risk assessment. The system uses both Machine Learning and Deep Learning approaches to predict heart disease likelihood while segmenting patients into distinct risk profiles for personalized healthcare interventions.

Key Features

• Predictive Modeling: ANN, RNN, Random Forest, Logistic Regression
• Patient Profiling: K-Means clustering for risk segmentation
• Interactive Dashboard: Real-time predictions with visualizations
• Model Comparison: Performance analysis across multiple algorithms

---

Dataset Information

Variable	Type	Description	Units	Missing
age	Continuous	Patient age	Years	❌
cholesterol	Continuous	Cholesterol level	mg/dL	❌
blood_pressure	Continuous	Systolic BP	mmHg	❌
chest_pain_type	Categorical	Chest pain type	0-3	❌
max_heart_rate	Continuous	Max heart rate	BPM	❌
target	Binary	Heart disease presence	0/1	❌

Dataset: 1,025 patients | Features: 13 clinical indicators

---

Model Architecture

Neural Networks

ANN: Input(13) → Dense(64,ReLU) → Dense(32,ReLU) → Output(1,Sigmoid) RNN: Input(Sequential) → LSTM(50) → Dropout(0.2) → Dense(25,ReLU) → Output(1,Sigmoid)

Clustering

K-Means: 3 clusters (Low/Medium/High Risk) with Silhouette Score validation

---

Installation & Setup

# Clone repository
git clone https://github.com/Aryanwadhwa14/HEART-DISEASE.git
cd HEART-DISEASE

# Install dependencies
pip install -r requirements.txt

# Run dashboard
streamlit run app/dashboard.py

Required Libraries

pandas>=1.3.0, numpy>=1.21.0, scikit-learn>=1.0.0
tensorflow>=2.6.0, streamlit>=1.2.0, matplotlib>=3.4.0

---

Usage

Training Models

python src/train_models.py --models all    # Train all models
python src/train_models.py --models rnn    # Train specific model
python src/clustering.py                   # Run clustering

Making Predictions

from src.predictor import HeartDiseasePredictor

predictor = HeartDiseasePredictor()
patient_data = {'age': 45, 'cholesterol': 240, 'blood_pressure': 140}
prediction = predictor.predict(patient_data)

---

Performance Results

Model	Accuracy	Precision	Recall	F1-Score	ROC-AUC
Logistic Regression	0.834	0.821	0.847	0.834	0.891
Random Forest	0.852	0.839	0.865	0.852	0.923
ANN	0.847	0.834	0.859	0.847	0.912
RNN	0.901	0.897	0.905	0.901	0.956

Best Performer: RNN model with 90.1% accuracy and 0.956 ROC-AUC

---

Project Structure

HEART-DISEASE/
├── data/                    # Dataset files
├── src/                     # Source code
│   ├── models/             # ML/DL model implementations
│   ├── data_preprocessing.py
│   ├── clustering.py
│   └── predictor.py
├── app/                    # Streamlit dashboard
├── notebooks/              # Jupyter notebooks
├── models/                 # Saved model files
└── requirements.txt

---

Web Dashboard Features

• Real-time Prediction: Interactive patient data input
• Risk Visualization: Probability gauges and charts
• Patient Clustering: Visual cluster analysis
• Model Comparison: Performance metrics comparison
• Data Explorer: Interactive dataset exploration

---

API Endpoints

POST /api/predict
{
    "age": 45,
    "cholesterol": 240,
    "blood_pressure": 140
}

Response:
{
    "prediction": 1,
    "probability": 0.78,
    "risk_level": "High"
}

---

Achievement

PATENT RECOGNITION

This heart disease prediction system has been officially recognized and patented by the Harvard Innovation Labs, demonstrating its novel approach in combining ML/DL models with patient clustering for cardiovascular risk assessment.

---

Contributing

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

Areas for contribution: Model improvements, UI enhancements, documentation, testing

---

License

MIT License - see LICENSE file for details.