Detection of Obstructive Sleep Apnea

Electronics and communication engineering

Sleep disorders are the most common health condition that can influence various aspects of life. Obstructive sleep apnea (OSA) is one of the serious sleep disorder, which causes the breathing to repeatedly start and stop during sleep. In many countries these kind of disorder is generally analyzed in sleep laboratories by the traditional detection process called Polysomnography. Most of the apnea disease are currently not analysed properly because of high cost of the test and the limitations of overnight sleep in the laboratories, where an expert human observer is needed to work over night. Multiple methods have been proposed to detect the physiological signals that are automatically analysed by different algorithms. In the proposed methodology different techniques are used for detecting the minute based analysis of OSA by Electrocardiogram (ECG) signal processing. Using the Physionet apnea ECG database, QRS complex is detected by pan Tompkins algorithm. Feature like Mean, Standard deviation and covariance is extracted from the output of the QRS complex. The classification algorithm is based on Support Vector Machines (SVM) and has been used to classify the apnea and nonapnea events from the features extracted. The software tool used for the detection of OSA is MATLAB platform. The main objective of the methodology is to detecting the OSA in a more accurate way to compute the sleep apnea score.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.