🩺 Lifestyle Disease Risk Prediction & Health Insights Dashboard

🔴 Live Dashboard

👉 Click here to view the live dashboard

---

📌 Project Overview

This project analyzes health-related data to predict the risk of lifestyle diseases based on an individual's health parameters. Users input basic health information such as blood pressure, heart rate, smoking status, BMI, and other lifestyle indicators, and the system analyzes this data to estimate potential disease risks.

The project also generates visual insights through a dashboard to help users better understand their health patterns and risk factors.

The goal of this project is to demonstrate how data analysis and predictive modeling can assist in early detection of lifestyle diseases and promote preventive healthcare.

---

🎯 Objectives

• Analyze healthcare datasets to identify patterns related to lifestyle diseases.
• Predict disease risk based on user-provided health parameters.
• Provide data-driven health insights through visualization dashboards.
• Help users understand how their lifestyle factors affect their health.

---

🧠 Key Features

✔ Health data preprocessing and cleaning ✔ Lifestyle disease risk prediction using data analysis techniques ✔ User input system for personal health parameters ✔ Risk evaluation based on dataset insights ✔ Interactive health insights dashboard ✔ Data visualization for better understanding of health patterns

---

🤖 ML Model Results

The backend trains and compares 6 machine learning models on the NHANES dataset (5,735 participants):

Model	Accuracy	F1 Score	AUC-ROC
Logistic Regression	95.12%	0.9712	0.9777
Decision Tree	99.91%	0.9995	0.9973
Random Forest	99.74%	0.9984	1.0000
Gradient Boosting ✅	99.91%	0.9995	1.0000
SVM	97.04%	0.9825	0.9943
KNN	96.34%	0.9783	0.9881

Best Model: Gradient Boosting (AUC = 1.0000)

Training Details

• Dataset: NHANES (National Health and Nutrition Examination Survey)
• Samples: 5,735 participants
• Features: 17 (including 5 engineered features)
• Train/Test Split: 80% / 20% stratified
• Cross-Validation: 5-Fold Stratified KFold

---

📊 Input Parameters

The system analyzes the following user inputs:

• Blood Pressure (Systolic & Diastolic)
• Heart Rate
• BMI
• Smoking Status
• Age
• Waist Circumference
• Physical Activity Level
• Other lifestyle-related indicators (based on dataset features)

Using these parameters, the system estimates the risk probability of lifestyle-related diseases.

---

📈 Output

The system provides:

• Disease risk estimation
• Health insights based on dataset trends
• Visual dashboard displaying:
  • Health indicators
  • Risk distribution
  • Lifestyle impact on disease probability

---

🛠 Technologies Used

Backend / ML

• Python 3
• Pandas – Data preprocessing and analysis
• NumPy – Numerical operations
• Matplotlib / Seaborn – Data visualization
• Scikit-learn – Machine learning models
• SciPy – Statistical analysis

Frontend / Dashboard

• React – Frontend framework
• Vite – Build tool
• Recharts – Interactive charts
• Vercel – Deployment

---

🚀 Run the ML Backend Locally

1. Clone the repo

git clone https://github.com/sam-coolshrestha/Health-Monitoring-System.git
cd Health-Monitoring-System

2. Install dependencies

pip install pandas numpy matplotlib seaborn scikit-learn scipy

3. Run the pipeline

python lifestyle_risk_model.py

This will print all model results and generate 4 plots in an outputs/ folder.

---

🖥 Run the Dashboard Locally

cd dashboard
npm install
npm run dev

Open http://localhost:5173 in your browser.

---

📂 Project Workflow

1️⃣ Data Collection 2️⃣ Data Preprocessing 3️⃣ Exploratory Data Analysis (EDA) 4️⃣ Feature Engineering 5️⃣ Model Training & Evaluation 6️⃣ Dashboard Visualization 7️⃣ User Health Input System (Risk Calculator)

---

📊 Dataset Analysis

The NHANES dataset contains multiple health and lifestyle parameters analyzed to identify correlations between lifestyle habits and disease risks.

Key findings:

• 59.4% of participants are smokers — largest single risk factor
• 71.2% are overweight or obese (BMI ≥ 25)
• Hypertension rate jumps from 13% (age 18–30) to 59% (age 60+)
• BMI and waist circumference are nearly perfectly correlated (r = 0.91)

---

🔮 Future Scope

1️⃣ Computer Vision Based Rehabilitation Assistant

A future version aims to integrate OpenCV-based rehabilitation exercise monitoring, where:

• Patients perform rehabilitation exercises
• The system detects body posture using computer vision
• It provides feedback on whether exercises are performed correctly

This feature is currently proposed as a future enhancement and has not yet been implemented.

2️⃣ Machine Learning Model Improvements

• Train deep learning models (Neural Networks)
• Improve accuracy of disease risk prediction

3️⃣ Real-Time Health Monitoring

Integration with wearable devices for real-time health data.

4️⃣ Backend API

Build a REST API so the dashboard can call the ML model directly in real time.

---

📜 License

This project is open source and available under the MIT License.