Małgorzata Andrzejewska
supervisor: Jan Żebrowski
The human body is a complex system that consists of many subsystems working together. This results in non-linear interactions between the individual components. Aging and disease can be associated with a loss of complexity, which can be estimated using heart rate variability signals. In this presentation, two approaches will be presented to determine irreversibility in time series. The first is an asymmetry analysis of time series using, for example, the Porta (P%) and Guzik (G%) indices. More advanced methods for estimating irreversibility are based on directed horizontal visibility graphs (dHVG). They consider the direction of the arrow of time by comparing the degree distributions of the vertices entering and leaving each vertex of the graph. Measures that can be used to compare these distributions are the Kullback-Leibler divergence (KLD) and the Jensen-Shannon divergence (JSD). It appears that measures based on the concept of irreversibility over time can be used to create better methods for stratifying the risk of sudden cardiac death (SCD). Results from the analysis of the THEW database, which includes healthy subjects and patients with long QT syndrome (LQTS), will be presented. Analysis of univariate and multivariate series shows, in general, a higher complexity in case of healthy people.