Automatic Detection of Obstructive Sleep Apnea Using Wavelet Transform and Entropy-Based Features From Single-Lead ECG Signal

Obstructive sleep apnea (OSA) is a prevalent sleep disorder and highly affects the quality of human life. Currently, gold standard for OSA detection is polysomnogram. Since this method is time consuming and cost inefficient, practical systems focus on the usage of electrocardiogram (ECG) signals for OSA detection. In this paper, a novel automatic OSA detection method using a single-lead ECG signal has been proposed. A nonlinear feature extraction using wavelet transform (WT) coefficients obtained by an ECG signal decomposition is employed. In addition, different classification methods are investigated. ECG signals are decomposed into eight levels using a Symlet function as a mother Wavelet function with third order. Then, the entropy-based features including fuzzy/approximate/sample/correct conditional entropy as well as other nonlinear features including interquartile range, mean absolute deviation, variance, Poincare plot, and recurrence plot are extracted from WT coefficients. The best features are chosen using the automatic sequential forward feature selection algorithm. In order to assess the introduced method, 95 single-lead ECG recordings are used. The support vector machine classifier having a radial basis function kernel leads to an accuracy of 94.63% (sensitivity: 94.43% and specificity: 94.77%) and 95.71% (sensitivity: 95.83% and specificity: 95.66%) for minute-by-minute and subject-by-subject classifications, respectively. The results show that applying entropy-based features for extracting hidden information of the ECG signals outperforms other available automatic OSA detection methods. The results indicate that a highly accurate OSA detection is attained by just exploiting the single-lead ECG signals. Furthermore, due to the low computational load in the proposed method, it can easily be applied to the home monitoring systems.

I. Introduction

Obstructive Sleep Apnea (OSA) is a prevalent sleep disorder (2% of women and 4% of men suffering from) which can be characterized using repetitive respiratory cessation during sleep [1], [2]. Clinically, there are three types of Sleep Apnea (SA): OSA, Central SA and Mixed SA [3], [4]. When there is a significant reduction in the volume of the air entering into the lungs, it is called Hypopnea (HA) [5]. In OSA, a temporal obstruction happens at the upper airway, especially throat, and causes throat collapse. During OSA, the airway is obstructed while there are still respiratory efforts against the obstruction [6]. OSA causes excessive daytime drowsiness, neurocognitive deficits, fatigue, depression, and heart stroke [7]–[9]. Also, undiagnosed and untreated OSA may lead to a high blood pressure, brain stroke, myocardial infarction, arrhythmias, and ischemia [10]–[12]. Even though OSA is detectable, the most cases are still not recognized [13]. Polysomnogram (PSG) is the gold standard for OSA detection, which is based on the comprehensive evaluation of the cardio-respiratory system and sleep signals [14]. In this method, case studies should be asleep for a couple of nights in the exclusive sleep laboratory in order to record the 16 major signals such as Electrocardiogram (ECG), Electroencephalogram (EEG), respiratory effort, airflow, and oxygen saturation (SaO2) [15], [16]. A PSG device needs at least 12 channels to record the data using 22 wire connectors [17]. The large number of the necessary wire connectors in a PSG device would interrupt the sleep, which affects the OSA detection. Moreover, the PSG test is typically performed in a hospital setting and it requires the supervision of a clinical expert, factors that make PSG an uncomfortable and costly procedure [18], [19]. When an OSA takes place, the oxygen saturation level falls while the cardiovascular and the automatic neural systems try to maintain this level [20]. Moreover, abnormal activities of the heart or significant changes in heart rate may indicate an OSA. Thus, among the developing trustworthy and low-cost techniques only single-lead ECG is used which can improve the early detection of OSA. So, the OSA detection would be possible by the friendly-used at home setting. [3], [6], [7], [21]. In 2000, the organizers of Physionet database held a challenge to detect the OSA using a single-lead ECG signal, in order to show the importance of the issue [22]–[24]. Khandoker et al. extracted 28 features from the heart rate variability (HRV) and ECG derived respiratory (EDR) signals [2]. Bsoul et al. proposed a real-time OSA detection system which has 111 features extracted from RR interval time series and EDR signals in time and frequency domains. [17]. Varon et al. used the extracted features from ECG, HRV and EDR signals [6]. Song et al. applied the extracted features from EDR signals and RR interval time series to the combination of the Hidden Markov Model (HMM) and Support Vector Machine (SVM) [15]. In 2017, Hassan et al. employed the features extracted from the ECG signals where, the signals were decomposed to some sub-bands through a tunable-Q factor wavelet transform (TQWT) and focused on the statistical features to detect OSA [15].