Heart Rate Variability (HRV) Analysis for the Identification of Cardiac Arrhythmias

Abstract

Sudden Cardiac death is the major cause of death and it causes more than 300,000 deaths annually in the United States. Spontaneous Ventricular Tachyarrhythmia (VTA), Myocardial Infarction (MI), Cardiomyopathy (CM) and Atrial Fibrillation (AF) are the main cause of SCD, contributing to about 82% of SCD. VTA consists of ventricular tachycardia (VT) and ventricular fibrillation (VF). To prevent from above mentioned diseases, we developed an early prediction model that can predict Cardiac Arrhythmias before its onset using Artificial Neural Network (ANN). De-identified raw data of patients were collected from MIT-BIH Physionet database. The dataset consists of 98 recordings obtained 10 seconds prior to the occurrence. ANN model is generated using 17 parameters obtained from Heart Rate Variability (HRV) analysis. Heart Rate (HR) variation analysis has become a popular noninvasive tool for assessing the activities of autonomic nervous system (ANS). Two third of the extracted parameters are used to train the neural network while one third is used to evaluate the performance of learned neural network as well as used for testing purposes. The developed Cardiac arrhythmias prediction model proved its performance by achieving an accuracy of 81 %.