Dynamical Landscape of Heart Rhythm in Long-Term Heart Transplant Recipients: A Way to Discern Erratic Rhythms

Wdowczyk, Joanna; Makowiec, Danuta; Gruchała, Marcin; Wejer, Dorota; Struzik, Zbigniew R.

doi:10.3389/fphys.2018.00274

Cited by 9 publications

(9 citation statements)

References 59 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, in certain cases (de Bruyne et al, 1999 ; Stein et al, 2008 ; Almeida-Santos et al, 2016 ; Raman et al, 2017 ; Wdowczyk et al, 2018 ) a paradoxical increase in short-term HR fluctuations was observed in contexts where reduced vagal tone would have been expected based on age and/or advanced heart disease. In the present study, we observed a U-shaped relationship between traditional short-term HRV measures and cross-sectional age (Figure 2 ).…”

Section: Discussionmentioning

confidence: 99%

Heart Rate Fragmentation as a Novel Biomarker of Adverse Cardiovascular Events: The Multi-Ethnic Study of Atherosclerosis

et al. 2018

View full text Add to dashboard Cite

Background: A major objective of precision medicine is the elucidation of non-invasive biomarkers of cardiovascular (CV) risk. Recently, we introduced a new dynamical marker of sino-atrial instability, termed heart rate fragmentation (HRF), which outperformed traditional and nonlinear heart rate variability metrics in separating ostensibly healthy subjects from patients with coronary artery disease. Accordingly, we hypothesized that HRF may be a dynamical biomarker of adverse cardiovascular events (CVEs).Methods: This study employed data from a cohort of participants in the Multi-Ethnic Study of Atherosclerosis (MESA), a prospective study of sub-clinical heart disease. Interbeat interval time series (n = 1963), derived from the electrocardiographic channel of the polysomnogram study, were analyzed using the newly introduced metrics of fragmentation, as well as traditional heart rate variability (HRV) indices and the short-term detrended fluctuation analysis exponent. Cox regression analysis was used to assess the association between HR dynamic indices and CV outcomes in unadjusted and adjusted models.Results: The mean (± SD) follow-up time was 2.97 ± 0.63 years. In adjusted models, higher fragmentation was significantly associated with incident CVEs (number of events; hazard ratio [95% confidence interval]: n = 72, 1.43 [1.16–1.76]) and CV death (n = 21; 1.65 [1.15–2.36]). The traditional HRV and the fractal indices were not associated with CVEs or CV death. The most discriminatory fragmentation indices added significant value to Framingham and MESA CV risk indices in all analyses.Conclusion: Our findings show that HRF has promise as a non-invasive, automatable biomarker of CV risk. The basic mechanisms underlying fragmentation remain to be delineated. Its association with incident outcomes raises the possibility of connections to degenerative changes in the multisystem network controlling SAN function.

show abstract

Section: Discussionmentioning

confidence: 99%

Heart Rate Fragmentation as a Novel Biomarker of Adverse Cardiovascular Events: The Multi-Ethnic Study of Atherosclerosis

et al. 2018

View full text Add to dashboard Cite

show abstract

“…This could be correlated to the fact that has been found in [48] that suggests that any decrease in the fluctuations of these features are hardly to be detected in a recorded ECG data. In addition, a paradoxical increase in these fluctuations comes as a results of a reduced vagal tone that would suggest the occurrence of heart diseases [49].…”

Section: Clinical Relevancementioning

confidence: 99%

Predicting Hypertensive Patients With Higher Risk of Developing Vascular Events Using Heart Rate Variability and Machine Learning

et al. 2020

View full text Add to dashboard Cite

The prognosis of cardiovascular and cerebrovascular events for patients suffering from hypertension is considered of a high importance in preventing any further development of cardiac diseases. Despite of the ability of current gold standard techniques in predicting vascular events risks, they still lack the required clinical efficiency. In this vein, the study proposed herein provides an investigation on the feasibility of using heart rate variability (HRV) utilized through a machine learning approach to predict hypertensive patients at higher risk of developing vascular events. Initially, HRV features were extracted from all patient's data using time-domain, frequency-domain, non-linear, and fragmentation metrics. The extraction of features was based on a 24-hour cycle analysis segmented into four time periods; namely late-night, early-morning, afternoon, and evening. Analysis of all features was performed using a one-way analysis of variance (ANOVA) test on period by period basis. Furthermore, the selection of best features was performed following a Chi-squared test for demographic and HRV features. Then, a model based on decision trees and random under-sampling boosting (RUSBOOST) was trained using demographic features, HRV features, and a combination of both features. The performance of the trained model achieved a maximum accuracy of 97.08% using the combined set of features during the afternoon time period. In addition, the precision and F1-score in predicting high risk patients reached 81.25% and 86.67%, respectively. The overall area under the curve for the model was at 0.98, suggesting a high performance in the sensitivity and specificity measures. This study paves the way towards utilizing machine learning models and heart rate variability for the prognosis of vascular events in hypertensive patient. Furthermore, it assists clinicians in decision making by providing a simple, yet effective, and continuous prediction approach when compared to other available techniques. INDEX TERMS hypertension, cardiovascular events, cerebrovascular events, heart rate variability, ANOVA, machine learning, feature selection, RUSBOOST

show abstract

“…The stochastic features of the short term dynamics can be evaluated by [ 29 ]: entropy of transition rates

self-transfer entropy

…”

Section: Methodsmentioning

confidence: 99%

Patterns of Heart Rate Dynamics in Healthy Aging Population: Insights from Machine Learning Methods

Makowiec

Wdowczyk

2019

Entropy

Self Cite

View full text Add to dashboard Cite

Costa et. al (Frontiers in Physiology (2017) 8255) proved that abnormal features of heart rate variability (HRV) can be discerned by the presence of particular patterns in a signal of time intervals between subsequent heart contractions, called RR intervals. In the following, the statistics of these patterns, quantified using entropic tools, are explored in order to uncover the specifics of the dynamics of heart contraction based on RR intervals. The 33 measures of HRV (standard and new ones) were estimated from four hour nocturnal recordings obtained from 181 healthy people of different ages and analyzed with the machine learning methods. The validation of the methods was based on the results obtained from shuffled data. The exploratory factor analysis provided five factors driving the HRV. We hypothesize that these factors could be related to the commonly assumed physiological sources of HRV: (i) activity of the vagal nervous system; (ii) dynamical balance in the autonomic nervous system; (iii) sympathetic activity; (iv) homeostatic stability; and (v) humoral effects. In particular, the indices describing patterns: their total volume, as well as their distribution, showed important aspects of the organization of the ANS control: the presence or absence of a strong correlation between the patterns’ indices, which distinguished the original rhythms of people from their shuffled representatives. Supposing that the dynamic organization of RR intervals is age dependent, classification with the support vector machines was performed. The classification results proved to be strongly dependent on the parameters of the methods used, therefore determining that the age group was not obvious.

show abstract

Dynamical Landscape of Heart Rhythm in Long-Term Heart Transplant Recipients: A Way to Discern Erratic Rhythms

Cited by 9 publications

References 59 publications

Heart Rate Fragmentation as a Novel Biomarker of Adverse Cardiovascular Events: The Multi-Ethnic Study of Atherosclerosis

Heart Rate Fragmentation as a Novel Biomarker of Adverse Cardiovascular Events: The Multi-Ethnic Study of Atherosclerosis

Predicting Hypertensive Patients With Higher Risk of Developing Vascular Events Using Heart Rate Variability and Machine Learning

Patterns of Heart Rate Dynamics in Healthy Aging Population: Insights from Machine Learning Methods

Contact Info

Product

Resources

About