Exploring the Predictability of EEG Signals Timed with the Heartbeat: A Model-Based Approach for the Temporal and Spatial Characterization of the Brain Dynamics

This study aims to provide a temporal and spatial characterization of the human brain activity related to the cardiac cycle in terms of regularity of the brain wave amplitudes measured from electroencephalographic (EEG) signals. To achieve this objective, linear autoregressive models are employed to characterize time-series of the spectral power extracted from EEG signals, timed with the heartbeat, by using a measure of predictability. The analysis is performed on four different time-series acquired on healthy subjects in a resting state and describing the EEG spectral content over the whole frequency spectrum and within the \(\theta \) , \(\alpha \) and \(\beta \) bands. Our results indicate predictability values with targeted activations in the frontal and parieto-occipital brain regions, which reflect regular amplitude modulations of the brain waves at rest, and could be linked to the cortical processing of the heartbeat.