An Algorithm Based On ECG Signal For Sleep Apnea Detection

Sleep apnea syndrome is a common sleep disorders affecting the quality of sleep, it is an independent risk factor of hypertension, coronary heart disease, arrhythmia disease. In recent years, detection, prevention, and treatment of sleep apnea syndrome caused a great deal of concern. Traditional sleep apnea syndrome detection method is polysomnography. This approach to measure multi-channel physiological signals would inevitably interfere with normal sleep, expensive and complicated. Therefore proposing a simple and effective method to detect sleep apnea syndrome have far-reaching implications for patients and physicians.This paper proposed a single-channel ECG sleep apnea syndrome detection algorithm,. The algorithm used a notch filter and a median filter to preprocess ECG frequency interference and baseline noise:QRS detection algorithm based on wavelet decomposition on RR interval correction algorithm was used to reduce the R wave undetected issues, improve the degree of concentration in the scatter plot of the RR interval; Some of the characteristics associated with the sleep apnea syndrome was proposed based on Heart rate variability; Support vector machines classification method was used to solve the traditional machine learning over learning problems;,F-value method was used to solve the contradiction between the and specificity and sensitivity.The detection of the MIT-BIH database (Apnea-ECG database) showed that the accuracy of this algorithm on a training set and a test set of94.44%and87.82%, respectively, reached the international advanced level. Compared with PSG method, the paper algorithm was more simple, accurate, efficient, and automatic. The algorithm can be applied to dynamic electrocardiogram device to make the detection of sleep apnea syndrome become simple and affordable.