Visualization of short-term heart period variability with network tools as a method for quantifying autonomic drive

TL;DR

This paper introduces a network-based visualization method for analyzing short-term heart rate variability, revealing nonlinear autonomic regulation features and differences between healthy individuals and transplant recipients.

Contribution

The study presents a novel network visualization approach for heart rate variability analysis, highlighting nonlinear properties and autonomic influences in heart regulation.

Findings

Large accelerations are more likely antipersistent.

Large decelerations are more likely persistent.

Autonomic regulation is significantly reduced in transplant recipients.

Abstract

Signals from heart transplant recipients can be considered to be a natural source of information for a better understanding of the impact of the autonomic nervous system on the complexity of heart rate variability. Beat-to-beat heart rate variability can be represented as a network of increments between subsequent $RR$-intervals, which makes possible the visualization of short-term heart period fluctuations. A network is constructed of vertices representing increments between subsequent $RR$-intervals, and edges which connect adjacent $RR$-increments. Two modes of visualization of such a network are proposed. The method described is applied to nocturnal Holter signals recorded from healthy young people and from cardiac transplant recipients. Additionally, the analysis is performed on surrogate data: shuffled RR-intervals (to display short-range dependence), and shuffled phases of the Fourier Transform of RR-intervals (to filter out linear dependences). Important nonlinear properties of autonomic nocturnal regulation in short-term variability in healthy young persons are associated with $RR$-increments: accelerations and decelerations of a size greater than about 35 ms. They reveal that large accelerations are more likely antipersistent, while large decelerations are more likely persistent. Changes in $RR$-increments in a heart deprived of autonomic supervision are much lower than in a healthy individual, and appear to be maintained around a homeostatic state, but there are indications that this dynamics is nonlinear. The method is fruitful in the evaluation of the vagal activity - the quantity and quality of the vagal tone - during the nocturnal rest of healthy young people. The method also successfully extracts nonlinear effects related to intrinsic mechanisms of the heart regulation.