Linear and non-linear brain-heart and brain-brain interactions during sleep.

摘要: In this study, the physiological networks underlying joint modulation of parasympathetic component heart rate variability (HRV) and different electroencephalographic (EEG) rhythms during sleep were assessed using two popular measures directed interaction in multivariate time series, namely Granger causality (GC) transfer entropy (TE). Time series representative cardiac brain activities obtained 10 young healthy subjects as normalized high frequency (HF) HRV EEG power δ, θ, α, σ, β bands, measured whole duration sleep. The magnitude statistical significance GC TE evaluated between each pair conditional on remaining respectively a linear model-based approach exploiting regression models, nonlinear model-free combining nearest-neighbor estimation with procedure for dimensionality reduction. contribution dynamics to was also surrogate data. consistently detected structured interactions, links predominantly from waves brain–heart network, σ α brain–brain network. While these common patterns supported suitability analysis, we found significant dynamics, particularly involving information transferred out δ node networks. This suggested importance nonparametric evidencing fine structure autonomic regulation functions