Review on Obstructive Sleep Apnea Detection using ECG Signals

Abstract

Obstructive Sleep apnea is a frequent sleep condition marked by pauses in the respiratory breathing cycle or episodes of unusual or shallow respiration while sleeping. It is majorly classified into obstructive, central, and mixed sleep apnea among which Obstructive Sleep Apnea (OSA) is most common. It is seen among people of different age groups, and most common among elderly people. OSA has a major impact on the normal sleep cycle that leads to various heart related problems. The traditional way of detecting the sleep disorder is through Polysomnography (PSG) and various methods have been proposed in the last few decades to replace the traditional way because of its underlying complexity. Nevertheless, the accuracy of these instruments is often insufficient to make medical assessment. As a result, the goal of this study is to examine current techniques which have not yet been deployed on hardware but have had their accuracy validated by at least one study aimed at detecting and predicting obstructive sleep apnea. The current work uses the public Apnea-ECG dataset, that is published at PhysioNet, to conduct an review of state-of-the-art approaches for OSA prediction.