AI-Driven Cardiac Monitoring System

CardioAI is an AI-powered cardiac monitoring platform designed to analyze ECG signals, monitor patient vitals in real time, and detect potential heart abnormalities using machine learning and signal processing techniques.

The system integrates medical signal processing, machine learning models, a real-time monitoring dashboard, and a backend patient management system to assist healthcare professionals in cardiac monitoring and early diagnosis.

📌 Table of Contents

Project Overview

Problem Statement

Solution

System Architecture

Features

Technology Stack

Project Structure

Installation Guide

Usage Guide

API Endpoints

ECG Processing Pipeline

Machine Learning Models

Database Schema

Future Improvements

Author

🧠 Project Overview

CardioAI is a smart healthcare monitoring system that processes ECG signals and extracts important cardiac health metrics.

The platform provides:

Real-time ECG visualization

Heart rate and HRV analysis

AI-based anomaly detection

Patient monitoring dashboard

Medical data storage and analysis

The goal is to assist doctors and healthcare professionals in detecting cardiac abnormalities earlier and monitoring patient health continuously.

⚠️ Problem Statement

Heart diseases are one of the leading causes of death worldwide.

Traditional cardiac monitoring systems have several limitations:

Manual interpretation of ECG signals

Delay in identifying abnormalities

Lack of automated risk prediction

Limited real-time patient monitoring

Doctors often need to manually analyze ECG waveforms, which can be time-consuming and error-prone.

💡 Proposed Solution

CardioAI addresses these issues by providing an AI-driven monitoring platform that:

Automatically processes ECG signals

Extracts cardiac features

Uses machine learning to detect abnormalities

Displays patient vitals on a real-time dashboard

Stores patient health data for long-term analysis

This system helps healthcare professionals make faster and more informed medical decisions.

🏗️ System Architecture ECG Signal Input │ ▼ Signal Preprocessing (Bandpass Filter + Noise Removal) │ ▼ Feature Extraction (HR, HRV, R-Peaks) │ ▼ Machine Learning Models (Random Forest / SVM / LSTM) │ ▼ Risk Prediction │ ▼ Flask Backend API │ ▼ Dashboard Interface │ ▼ SQLite Database 🚀 Features ❤️ Real-time ECG Monitoring

Displays ECG waveform and heart activity in real time.

📊 Heart Rate & HRV Analysis

Calculates important metrics such as:

Heart Rate (BPM)

HRV RMSSD

SDNN

🧠 AI-based Risk Prediction

Predicts arrhythmia probability and cardiac risk score using machine learning models.

👤 Patient Management System

Allows doctors to:

Add new patients

Store medical history

Monitor vitals

📈 Vitals Monitoring

Tracks multiple health parameters:

Heart Rate

Blood Pressure

SpO2

Respiratory Rate

Temperature

🚨 Alert System

Triggers alerts when abnormal heart rhythms or vitals are detected.

📂 ECG Upload & Analysis

Users can upload ECG CSV files for automated analysis.

📋 Clinical Reports

Stores historical health logs and generates medical insights.

🧰 Technology Stack Backend

Python

Flask

Flask-CORS

Signal Processing

NeuroKit2

SciPy

NumPy

Pandas

Machine Learning

Scikit-learn

TensorFlow

Frontend

HTML

CSS

JavaScript

Chart.js

Database

SQLite

📁 Project Structure CardioAI/ │ ├── app.py │ Flask backend server and API routes │ ├── database.py │ Database initialization and schema │ ├── preprocessing.py │ ECG signal cleaning and filtering │ ├── feature_extraction.py │ Extract physiological features from ECG │ ├── cardiac_monitor.db │ SQLite database │ ├── index.html │ Main monitoring dashboard │ ├── app.js │ Frontend logic and real-time updates │ ├── styles.css │ Custom UI styling │ ├── metrics_comparison.json │ ML model evaluation results │ ├── requirements.txt │ Python dependencies │ └── README.md