Exploring ECG Signal Analysis Techniques for Arrhythmia Detection: A Review

Abstract

The heart holds paramount importance in the human body as it serves the crucial function of supplying blood and nutrients to various organs. Thus, maintaining its health is imperative. Arrhythmia, a heart disorder, arises when the heart's rhythm becomes irregular. Electrocardiogram (ECG) signals are commonly utilized for analyzing arrhythmia due to their simplicity and cost-effectiveness. The peaks observed in ECG graphs, particularly the R peak, are indicative of heart conditions, facilitating arrhythmia diagnosis. Arrhythmia is broadly categorized into Tachycardia and Bradycardia for identification purposes. This paper explores diverse techniques such as Deep Convolutional Neural Networks (CNNs), Long Short-Term Memory (LSTM), Support Vector Machines (SVM), Neural Network (NN) classifiers, as well as Wavelet and Time–Frequency Transform (TQWT), which have been employed over the past decade for arrhythmia detection using various datasets. The study delves into the analysis of arrhythmia classification on ECG datasets, highlighting the effectiveness of data preprocessing, feature extraction, and classification techniques in achieving superior performance in classifying ECG signals for arrhythmia detection.