Non-linear Analysis of Cardiac Autonomic Nervous Activity using Principal Dynamic Fluctuation Analysis

Heart Rate Variability (HRV) is a marker of sympathovagal balance of Autonomous Nervous System (ANS). The homeostasis of the cardiovascular system is properly maintained by the efficient interaction of parasympathetic and sympathetic nervous systems. The efficiency of such interactions is understood from the computations of the ratio of the sympathetic to parasympathetic spectral powers. This paper introduces a modified Principal Dynamic fluctuation analysis (PDFA) using a non-linear kernel trick to estimate the dynamics of cardiac autonomic nervous systems activity from the heart rate variability. It is found that the wide range of beating ratio has better sympathovagal balance. A large LF/HF ratio suggests predominantly vagal conrol. Comparison of this method to the conventional linear methods shows PDFA provide more accurate assessment of the sympathovagal balance. We obtained consistent results in all of our signals, comprising of seven sudden death cardiac signals and five normal sinus rhythm signals.