An Approach to the Improvement of Electrocardiogram-based Sleep Breathing Pauses Detection by means of Permutation Entropy of the Heart Rate Variability

Permutation entropy obtained from heart rate variability (HRV) is analyzed in a statistical model integrating electrocardiogram derived respiratory (EDR) features and cepstrum coefficients in order to detect obstructive sleep apnea (OSA) events. 70 ECG recordings from Physionet database are divided into a learning set and a test set of equal size. Each set consists of 35 recordings, containing a single ECG signal. Each recording includes a set of reference annotations, one for each minute, which indicates the presence or absence of apnea during that minute. Statistical classification methods based on Logistic Regression (LR) is applied to the classification of sleep apnea epochs. EDR presents a sensitivity of 64.3% and specificity of 86.5% (auc=83.9). Cepstrum presents a sensitivity of 63.8% and specificity of 89.2% (auc=86). Contribution of the permutation entropy increases the performance of the LR model, playing an important role in the OSA quantification task. In particular, when all features are analyzed, classifier reaches a sensitivity of 70.2% and specificity of 91.8% (auc=89.8).