An Experimental Review on Obstructive Sleep Apnea Detection Based on Heart Rate Variability and Machine Learning Techniques

摘要: Obstructive sleep apnea (OSA) is a respiratory syndrome of high incidence in the general population and correlated with some cardiovascular diseases. Several techniques have been proposed last decades to find surrogate method polysomnography (PSG), gold standard for diagnosis OSA. The present study comprises an experimental review on state-of-the-art methods OSA detection through public Apnea-ECG database, which available at PhysioNet. Precisely, traditional time-frequency domain features were extracted from heart rate variability (HRV) signal, together common complexity measures. Given their ability deal real-world time series, two additional entropy-based measures also tested, i.e., Renyi Tsallis entropies. Moreover, univariate multivariate classifiers applied, including diagnostic test, support vectors machine, k-nearest neighbors. Ultimately, sequential feature selection (SFS) algorithms employed reduce computational cost resulting discriminant models. major findings reported that reached similar results those found literature. classification suggested frequency provided best detection, although well-known entropy index obtained good performance.