Development of Decision Support System for Automatic Sleep Stage Scoring

Abstract

In the diagnosis of sleep disorders, analyzing the patient's sleep stage scores is the basic clinical procedure. The signals obtained from the polysomnography device are used in the sleep stage scoring studies. The polysomnography device records the physiological signals from different channels through electrodes during patients' night sleep. Aim of the study is developing a decision support system which determines the state of sleep and wakefullness by using only one channel electroencephalography signal. The amplitude and frequency values of the sub-bands of the electroencephalogram signal change in the sleep stages. Accordingly, five attributes were determined and classified using support vector machines As a result of the classification study, accuracy, sensitivity and specificity values were calculated for the performance evaluation of the algorithm.