Conditional entropy approach for the evaluation of the coupling strength

Area of Science:

• Physiological systems analysis
• Information theory applications
• Biomedical signal processing

Background:

• Quantifying signal coupling is crucial for understanding physiological system interactions.
• Existing methods may struggle with short data segments and non-linear dynamics.
• Beat-to-beat variability analysis is key in cardiovascular and respiratory research.

Purpose of the Study:

• To present a novel method for measuring the degree of coupling between two signals.
• To introduce an uncoupling function based on entropy rates for quantifying independent information.
• To validate a derived synchronisation index for assessing information exchange in physiological signals.

Main Methods:

• Developed an uncoupling function using entropy rates to estimate independent information.
• Created an estimator for the uncoupling function suitable for short data segments.
• Derived a synchronisation index via minimization, quantifying maximum information exchange.

Main Results:

• Simulations confirmed the index's ability to measure synchronisation under induced coupling changes.
• Sympathetic discharge and ventilation in cats showed strong coupling, unaffected by sympathetic activity modulation.
• Heart period and ventricular repolarisation interval synchronisation decreased significantly in myocardial infarction patients.

Conclusions:

• The proposed synchronisation analysis method accurately measures signal coupling in physiological systems.
• Sympathetic discharge and ventilation exhibit robust coupling in decerebrate cats.
• Myocardial infarction leads to a loss of synchronisation between heart period and ventricular repolarisation interval.