ECG Based Heart Disease Diagnosis Using Machine Learning Approaches

Abstract

The electrocardiogram (ECG) is crucial to monitor cardiac health, especially since its signal is the most important diagnostic tools to help in the detection of heart disease. ECG interpretation has largely been confined to manual analysis, which suffers from constraints such as expert availability in underserved areas, and diagnostic errors. Addressing these issues underwent research via machine learning through the fusion of ECG data for improved heartbeat classification. This study presents a novel approach that incorporates a Support Vector Machine (SVM) with Random Forest (RF), Logistic Regression (LR), Decision Tree (DT) models in a comprehensive method designed to classify heartbeats into normal, abnormal and COVID-19 affected. The individual performance of Decision Tree, Logistic Regression, Random Forest and Support Vector Machine models are evaluated on ECG image dataset. The respective accuracy rates were 77%, 82%, 78%, and 83%. The SVM model produced a superior accuracy of 84%. This comparative analysis thus identifies the potential for SVM model to the empower ECG signal interpretation and take the clinical depart towards remote diagnostics while ensuring early detection of cardiac anomalies.