Obstructive Sleep Apnea (OSA) Classification Based on Heart Rate Variability (HRV) on Electrocardiogram (ECG) Signal Using Support Vector Machine (SVM)

Abstract

Obstructive sleep apnea (OSA) is a disorder that occurs in a person's sleep pattern, and due to its ability to cause other health-related problems, various monitoring methods have been developed. One of the proposed methods is the ECG signal known as ECG-derived respiration (EDR), which uses changes in rhythm and patterns regularity to detect OSA occurrence. Therefore, this study aims to determine the OSA classification based on heart-rate variability (HRV) on electrocardiogram (ECG) signal using a support vector machine (SVM). The HRV parameter displays the rhythmic changes in the ECG signal under normal and OSA conditions. Eleven HRV characteristics were used to produce the highest accuracy of 89.5% with a support vector machine (SVM) as a classifier. The results were tested on a 1 minute long ECG signal annotated by an expert, which indicated that OSA can be detected by observing the dynamics of the distance of the R-R wave on the ECG signal.