ECG-based Analysis of Stress Induced Cardiac Arrhythmias using Machine Learning

Abstract

An extensively used signal for identifying and predicting Cardiovascular Diseases (CVDs) is the Electrocardiogram (ECG), which is especially useful for detecting arrhythmias. Accurate diagnosis of acute and chronic heart diseases depends on the analysis of ECG signals. This study investigates Machine Learning (ML) techniques to categorize and identify cardiac arrhythmias brought on by stress in ECG signals. We used real-world ECG data for our analysis and evaluated previous research. We assessed the efficacy of several machine learning methods by concentrating on balancing different categories of heart conditions within the data. According to our research, the Random Forest method, in particular, improves overall accuracy and attains a more evenly distributed detection rate for various cardiac diseases. These findings highlight the significance of data balance in creating dependable medical diagnostic systems. Future research will assess algorithm performance in dynamic environments and integrate real-time data to enhance patient outcomes and diagnostic accuracy.