The ANS Sympathovagal Balance Using a Hybrid Method Based on the Wavelet Packet and the KS-Segmentation Algorithm

摘要

In this work, we have studied the Autonomous Nervous System (ANS) sympathovagal balance during a test period using the LF signal variance to the HF signal variance ratio. To compute, more accurately, these variances of the LF and HF temporal signals, we have extracted the latter from the heart rate variability signal (HRV) using wavelet packet db4, and then we have segmented them into stationary segments by applying the KS-segmentation algorithm which is an approach based on Kolmogorov-Smirnov (KS) statistic to identify the adequate stationary patches in the LF and HF temporal signals. As a result, the estimated variances were obtained, accurately, for each resulting stationary segment. The LF signal variance to the HF signal variance ratio determined using wavelet packet db4 and the KS-segmentation algorithm is more accurate and thus can be used to localize as well as, more importantly, to estimate accurately the duration dominance of either the sympathetic or the parasympathetic activities for each individual during the test period. To clarify this hybrid method, we have applied it to heart beat time series of five young and five old individuals in rest state watching a fantasy film. The results obtained in this work show a significant discrepancy between the autonomic nervous system behaviors of individuals.