Performance Analysis of Machine Learning Algorithms for Sleep Apnea Detection Using ECG

Abstract

Sleep apnea is a sleep disorder in which a sleeping person’s breathing is disturbed. Subjects suffering from sleep apnea undergo periods of no or shallow breathing during their sleep. Sleep apnea may lead to severe issues such as diabetes, cardiovascular problems, hypertension, neurological issues and liver problems. Because of the global prevalence of sleep apnea as well as the direct and indirect long-term problems it brings about, it is important to diagnose and treat this condition. Sleep apnea is detected clinically by the Polysomnography (PSG) test which measures various biomedical parameters such as electrocardiogram (ECG), electroencephalogram (EEG) and oxygen saturation (SpO2) over a full night’s sleep. The application of machine learning to detect sleep apnea from these parameters has gained ground in the recent past because of its ability to learn from the training datasets and generalize well to make predictions on new data. In this paper, we look at the performance of 6 machine learning classifiers—k-nearest neighbors (kNN), Support Vector Machine (SVM), Artificial Neural Networks (ANN), Random Forest and XGBoost in their ability to detect apneic events. The study is based on datasets with ECG signals.