0000000000162733
AUTHOR
Chiara Bara
A new framework for the time- and frequency-domain assessment of high-order interactions in networks of random processes
While the standard network description of complex systems is based on quantifying the link between pairs of system units, higher-order interactions (HOIs) involving three or more units often play a major role in governing the collective network behavior. This work introduces a new approach to quantify pairwise and HOIs for multivariate rhythmic processes interacting across multiple time scales. We define the so-called O-information rate (OIR) as a new metric to assess HOIs for multivariate time series, and present a framework to decompose the OIR into measures quantifying Granger-causal and instantaneous influences, as well as to expand all measures in the frequency domain. The framework ex…
Analysis of Cardiac Pulse Arrival Time Series at Rest and during Physiological Stress
The study of cardiovascular dynamics is pivotal in the prevention and monitoring of cardiovascular diseases. Pulse Arrival Time (PAT) series contain information concerning not only the dynamics of the Autonomic Nervous System (ANS), but of all the systems involved in the regulation of cardiovascular homeostasis. This study aims to highlight how indexes extracted from PAT series in time-, frequency- and information-domain allow to discriminate among different physiological conditions. Analyses were carried out on 76 young healthy subjects, at rest and during orthostatic or mental stress. Our results show that PAT indexes vary according to the ANS condition, and may thus be useful parameters …
Assessment of Cardiorespiratory Interactions During Spontaneous and Controlled Breathing: Linear Parametric Analysis
In this work, we perform a linear parametric analysis of cardiorespiratory interactions in bivariate time series of heart period (HP) and respiration (RESP) measured in 19 healthy subjects during spontaneous breathing and controlled breathing at varying breathing frequency. The analysis is carried out computing measures of the total and causal interaction between HP and RESP variability in both time and frequency domains (low- and high-frequency, LF and HF). Results highlight strong cardiorespiratory interactions in the time domain and within the HF band that are not affected by the paced breathing condition. Interactions in the LF band are weaker and prevalent along the direction from HP t…
Assessment Of Driving Stress Through SVM And KNN Classifiers On Multi-Domain Physiological Data
We propose an objective stress assessment method based on the extraction of features from physiological time series and their classification using Support Vector Machine and K-Nearest Neighbors algorithms. For this purpose, we used an open dataset consisting of multiparametric physiological signals (electrocardiogram, electromyogram, galvanic skin response and breath signal) obtained during the execution of a driving route within the city of Boston with restful, highway and city driving periods indicative of three different stress states. To predict the driver stress level, 21 features were extracted from 122 chunks of raw signals and were subsequently managed by classification algorithms. …
Assessment of Cardiorespiratory Interactions During Spontaneous and Controlled Breathing: Non-linear Model-free Analysis
In this work, nonlinear model-free methods for bivariate time series analysis have been applied to study cardiorespiratory interactions. Specifically, entropy-based (i.e. Transfer Entropy and Cross Entropy) and Convergent Cross Mapping asymmetric coupling measures have been computed on heart rate and breathing time series extracted from electrocardiographic (ECG) and respiratory signals acquired on 19 young healthy subjects during an experimental protocol including spontaneous and controlled breathing conditions. Results evidence a bidirectional nature of cardiorespiratory interactions, and highlight clear similarities and differences among the three considered measures.
Transfer Entropy Analysis of Pulse Arrival Time - Heart Period Interactions during Physiological Stress
Although Heart Period (HP) variability is the most widely used measure to assess cardiovascular oscillations, its evaluation combined with that of Pulse Arrival Time (PAT) variability may provide additional information about cardiac dynamics and cardiovascular interactions. In this study, we computed the transfer entropy from PAT to HP in 76 subjects monitored at rest and during orthostatic and mental stress using both a model-free (k- Nearest Neighbors) and a linear parametric estimator. Our results show how the information flow between these two variables depends on the physiological condition and how the nonlinear measure captures more information than the linear one during orthostatic s…
Spectral analysis of the beat-to-beat variability of arterial compliance
Arterial compliance is an important parameter influencing ventricular-arterial coupling, depending on structural and functional mechanics of arteries. In this study, the spontaneous beat-to-beat variability of arterial compliance was investigated in time and frequency domains in thirty-nine young and healthy subjects monitored in the supine resting state and during head-up tilt. Spectral decomposition was applied to retrieve the spectral content of the time series associated to low (LF) and high frequency (HF) oscillatory components. Our results highlight: (i) a decrease of arterial compliance with tilt, in agreement with previous studies; (ii) an increase of the LF power content concurrent…
Feasibility of Ultra-short Term Complexity Analysis of Heart Rate Variability in Resting State and During Orthostatic Stress
In this work, we study ultra-short term (UST) complexity of Heart Rate Variability (HRV) and its agreement with analysis of standard short-term (ST) HRV recordings obtained at rest and during orthostatic stress. Conditional Entropy (CE) measures have been computed using both a linear Gaussian approximation and a more accurate model-free approach based on nearest neighbors. The agreement between UST and ST indices has been compared via statistical tests and correlation analysis, suggesting the feasibility of exploiting faster algorithms and shorter time series for detecting changes in cardiovascular control during various states.