Sub-frame based apnea detection exploiting delta band power ratio extracted from EEG signals

Abstract

Sleep apnea syndrome (SAS) is one of the most common sleep disorder diseases. In this paper, we have proposed a new apnea detection method exploiting delta band power ratio of EEG signal. Each EEG signal frame of 30 seconds is divided into 10 seconds sub frames. Delta band power ratio is extracted as a feature from these sub frames. The feature thus formed is fed to KNN and SVM classifier, respectively to detect apnea and non apnea frames in patients with apnea, which is a challenging task. From MIT-BIH sleep apnea database, the proposed method is tested with 14 overnight polysomnographic (PSG) records. The proposed method outperforms one state-of-the-art method in terms of accuracy, sensitivity and specificity.