Optimal Classification of Atrial Fibrillation and Congestive Heart Failure Using Machine Learning

Abstract

Cardiovascular disorders, including atrial fibrillation (AF) and congestive heart failure (CHF), are the major causes of mortality worldwide. The diagnosis of cardiovascular disorders is heavily reliant on electrocardiogram (ECG) signals. Therefore, extracting significant features from ECG signals is the most challenging aspect to represent each condition of the ECG signals. Earlier studies have claimed that the Hjorth descriptor is assigned as a simple feature extraction algorithm that has the capability of class separation among AF, CHF, and normal sinus rhythm (NSR) conditions. However, owing to noise interference, certain features do not represent the characteristics of the ECG signals. This study addressed this important gap by applying a discrete wavelet transform (DWT) prior to applying the Hjorth descriptor as a feature extraction method. Furthermore, the feature selection process and optimization of various classifier algorithms, including k-nearest neighbor (k-NN), support vector machine (SVM), and artificial neural network (ANN), were investigated to provide the best system performance. This study obtained accuracies of 95 %, 92 %, and 95 % for the k-NN, SVM, and ANN classifiers, respectively. The results demonstrated that the optimization of the classifier algorithm could improve the classification accuracy of AF, CHF, and NSR conditions, compared to earlier studies.