Information-Based Similarity of Binary Symbol Sequence as a New Index for Sleep Apnea Research

Abstract

Sleep apnea (SA) as abreathing disorder during sleep, has an increasing incidence in recent years. As a non-invasive method, heart rate variability (HRV) analysis can be effectively used in sleep apnea research. The information-based similarity of binary symbol sequence (BS_IBS) is a symbolic dynamics analysis method, which can be employed to perform nonlinear HRV analysis of sleep apnea patients. In the study, 60 electrocardiogram recordings were used for analysis, and multiple indices in time domain, frequency domain, and nonlinearity were calculated at the same. The similarity between adjacent RR segments was evaluated by calculating the BS_IBS value of them. Our research shows that BS_IBS is a useful sleep apnea screening method and can improve the effect of sleep apnea screening with traditional time-frequency domain indices (the accuracy is increased from 81.7% to 86.7%). This study proves that nonlinear information-based similarity assessment can be used in sleep apnea research and further expands the nonlinear HRV analysis method of sleep apnea.