Analysis of Obstructive Sleep Apnea using ECG Signals

Abstract

Sleep is an important component in one's daily life and comprises about one-third of one's day. Sleep loss and disorders effect one's productivity thereby causing a significant impact on the economy. Polysomnography (PSG), which is considered the gold standard for sleep diagnosis, contains recording of multiple physiological signals Electroencephalogram (EEG), Electrooculogram (EOG), Electrocardiogram (ECG), Electromyogram (EMG), blood oxygen levels (oximetry). It has been observed that the PSG recordings obtained from the patients suffering from Obstructive Sleep Apnea (OSA) contain consistent, often repetitive, episodes of breathing pauses. However, PSG recordings are very expensive thus limiting its accessibility by the financially weaker section of the society. They are also at a greater risk of human error as over 7–8 hours recording is visually evaluated by a neurologist. The aim of this paper is to simplify the tools used for the analysis of sleep apnea. The signals procured via ECG for the analysis of OSA were explored and the accuracy for the same was analyzed. Three parameters of the signals namely Power Spectral Density, Correlation and R-R peak interval were analyzed.