Automated recognition of obstructive sleep apnoea syndrome from ECG recordings

Abstract

Obstructive sleep apnoea syndrome (OSAS) is a highly prevalent sleep disorder. The traditional diagnosis methods of the disorder are cumbersome and expensive. The ability to automatically identify OSAS from ECG recordings is important for clinical diagnosis and treatment. In this study, we presented a system for the automatic recognition of patients with OSA from nocturnal electrocardiogram (ECG) recordings. The presented OSA recognition system comprises of three stages. In the first stage, an algorithm based on DWT was used to analyze ECG recordings for detection ECG-derived respiration (EDR) changes. In the second stage, a FFT based Power spectral density method was used for feature extraction from EDR changes. In the third stage, using a least squares support vector machine (LS-SVM) classifier; normal subjects were separated from subjects with OSA based on obtained features. Using 10 fold cross validation method, the accuracy of proposed system was found 96.7%. The results confirmed that the presented system can aid sleep specialists in the initial assessment of patients with suspected OSA.