Comparative pairwise analysis of the relationships between physiological rhythms using synchrosqueezed wavelet transform, phase dynamics modelling and recurrence

The work is devoted to the application of various nonlinear dynamics methods to identify interactions in short noisy time series extracted from biological rhythms related to the respiratory, cardiovascular and nervous systems. These interactions are considered as transitions to synchronized states, as the coupling directionality and the delay time in the influence of one system on another. For the analyzed data, the relationship between these rhythms is due to the fact that the neurons of medulla are related to the brain response to changes in blood pressure and respiration during external pain stimulation. Phase synchronization between the variability of blood pressure and respiratory fluctuations in response to these stimuli was established in half of the analyzed time series using methods for assessing phase synchronization based on finding instantaneous phases by the synchrosqueezed wavelet transform and based on calculating the recurrences of phase trajectories. For unsynchronized time series, a predominantly unidirectional influence of respiratory rhythm fluctuations on the blood pressure variability was obtained by the method of phase dynamics modeling and the method of estimating recurrence probabilities. For two third of the data, a unidirectional influence of the blood pressure variability on the neuronal variability was determined, and the remaining data were characterized by a bidirectional relationship between the neuronal and blood pressure variability. Delay time estimation by phase dynamics modeling showed reduced time of influence of the blood pressure variability on the neuronal variability in bidirectional coupling compared to the influence of the neuronal variability on the blood pressure variability.