Automated Recognition of Obstructive Sleep Apnea Syndrome Using Support Vector Machine Classifier

Abstract

Obstructive sleep apnea (OSA) is a common sleep disorder that causes pauses of breathing due to repetitive obstruction of the upper airways of the respiratory system. The effect of this phenomenon can be observed in other physiological signals like the heart rate variability, oxygen saturation, and the respiratory effort signals. In this study, features from these signals were extracted from 50 control and 50 OSA patients from the Sleep Heart Health Study database and implemented for minute and subject classifications. A support vector machine (SVM) classifier was used with linear and second-order polynomial kernels. For the minute classification, the respiratory features had the highest sensitivity while the oxygen saturation gave the highest specificity. The polynomial kernel always had better performance and the highest accuracy of 82.4% (Sen: 69.9%, Spec: 91.4%) was achieved using the combined-feature classifier. For subject classification, the polynomial kernel had a clear improvement in the oxygen saturation accuracy as the highest accuracy of 95% was achieved by both the oxygen saturation (Sen: 100%, Spec: 90.2%) and the combined-feature (Sen: 91.8%, Spec: 98.0%). Further analysis of the SVM with other kernel types might be useful for optimizing the classifier with the appropriate features for an OSA automated detection algorithm.

Assessment of Time and Frequency Domain Entropies to Detect Sleep Apnoea in Heart Rate Variability Recordings from Men and Women

“…Data from studies involving SpO2, airflow, snoring, respiratory effort, and HRV were included. In the case of the SpO2 signal, Acc and AROC range from 84.1% to 95% and 0.822 to 0.967, respectively [49][50][51][52]. A database composed of 187 recordings was used to model a multi-layer perceptron (MLP) classifier, which was obtained from three non-linear features [49].…”

“…The best diagnostic ability for SpO2 in terms of AROC (0.967) was achieved by a LR model obtained from four automatically-selected features extracted from the frequency and time domain of 147 recordings [51]. The best Acc (95.0%) was reported in the case of a support vector machine (SVM) classifier evaluated for a 5 e/h AHI threshold [52].…”

“…A LR model obtained from three spectral features reached 82.4% Acc and 0.904 AROC after loo-cv (AHI threshold = 10 e/h). Moreover, five time and phase domain features from the abdominal and thoracic respiratory effort signals were used to feed a SVM classifier which reached 89.0% Acc, evaluated for AHI = 5 e/h [52]. Snoring sounds have been also assessed in the context of SAHS classification.…”

“…Eight features from wavelet analysis were used to build a decision tree, reaching 69.3% Acc [56]. A SVM model was obtained from five time and frequency domain features [52]. Authors reported 79.0% Acc when evaluating the classifier for AHI = 5 e/h.…”

Real-time detection, classification, and quantification of apneic episodes using miniature surface motion sensors in rats

“…While thoraco-abdominal asynchrony is expected during obstructive episodes in adults (25), neonates and infants may normally exhibit this type of breathing (10). In parallel to the reported modest success of polysomnography in detecting obstructive sleep apnea, some published works have demonstrated inaccurate detection in infants, with low specificity of 10.9% (ref.…”