Time-Resolved Brain-to-Heart Probabilistic Information Transfer Estimation Using Inhomogeneous Point-Process Models.

Area of Science:

• Neuroscience
• Computational Biology
• Physiology

Background:

• Quantifying brain-heart interplay is crucial for understanding cognitive, emotional, and autonomic states.
• Existing computational models lack directional, time-resolved, and probabilistic estimations of this interaction.

Purpose of the Study:

• To develop a novel computational framework for directional, time-resolved, and probabilistic quantification of brain-heart interplay.
• To establish subject-specific, dynamic functional estimates of brain-to-heart interactions.

Main Methods:

• Utilized a multivariate inhomogeneous point-process model for heartbeat dynamics.
• Employed electroencephalography (EEG) and R-peak intervals to represent brain and heart activity, respectively.
• Modeled neural dynamics as an exogenous input to autoregressive cardiac dynamics using an inverse-Gaussian probability density function.

Main Results:

• Demonstrated directional interplay from cortical dynamics to heartbeat series with time delays of 30-60s and 90-120s post-stimulus.
• Validated model performance using heart rate variability and EEG data from healthy volunteers during a cold-pressor test.
• Assessed model goodness-of-fit using the time-rescaling theorem.

Conclusions:

• The proposed framework offers a fully probabilistic definition of continuous functional brain-heart interplay.
• Provides novel insights into human neurophysiology and the dynamic coupling between neural and cardiac systems.
• Suggests potential for developing advanced biomarkers for physiological and psychological states.