Decomposition Of The Dynamic Information Shared In Complex Networks Of Physiological Variables

Laura Sparacino1, Chiara Bara'1, yuri antonacci1, Michal Javorka2, Luca Faes1
1University of Palermo, 2Comenius University


Abstract

In the field of Network Physiology, the human organism is described as an integrated network constituted by multiple organ systems interacting to coordinate their functions. For instance, the spontaneous oscillations and interactions between cardiac and vascular or respiratory variables have been studied through noninvasive data-driven methods based on time series analysis, whilst recent studies have been focused on characterizing the short-term-variability of arterial compliance, expected to be mostly affected by sympathetic activity, as well as by heart rate, blood pressure and respiration. Classical approaches defined in the time, frequency and information-theoretic domains have been used to describe the dynamic interplay between the heart and vascular or respiratory systems, being however limited by their intrinsic pairwise formulation. Recent efforts have been oriented to capture the complex dynamics involving three or more processes. In this context, the spectral partial information rate decomposition (PIRD) framework has been recently designed to understand how information is distributed in multivariate systems of stationary random processes with oscillatory components, with the goal to decompose the information rate that a target process shares with a set of source processes into components highlighting the unique information exclusively available from each source, the redundant information obtained from at least two different sources, and the synergistic information revealed only when multiple sources are considered simultaneously. In this study, we exploit the PIRD to elicit the physiological mechanisms underlying complex cardiovascular regulation from the joint analysis of the spontaneous beat-to-beat variability of heart period, arterial pressure, respiration and arterial compliance, in the supine resting state and during head-up tilt. The relevance and meaning of the observed behaviors in different spectral bands with physiological meaning is discussed in terms of novel insights on the complex mechanisms governing the cardiovascular, cardiorespiratory and vascular-respiratory interplays.