Tracking instantaneous entropy in heartbeat dynamics through inhomogeneous point-process nonlinear models

Area of Science:

• Nonlinear Dynamics
• Physiological Signal Processing
• Information Theory

Background:

• Entropy measures quantify complex nonlinear systems, especially in heartbeat dynamics.
• Existing entropy measures lack time-varying definitions for physiological dynamics.
• A need exists for instantaneous entropy quantification of autonomic nervous system (ANS) activity.

Purpose of the Study:

• To introduce two novel, time-varying entropy measures for heartbeat dynamics.
• To enable instantaneous tracking of ANS complexity using RR interval series.
• To develop new indices for complexity variability analysis.

Main Methods:

• Utilizing inhomogeneous point-process theory to model RR interval series.
• Employing probability density functions (pdfs) to predict event timing based on past history.
• Incorporating Laguerre expansions of Wiener-Volterra terms for nonlinear information.
• Calculating phase-space vector distances using Kolmogorov-Smirnov distance between pdfs.

Main Results:

• The proposed entropy indices provide instantaneous tracking of heartbeat complexity.
• Demonstrated ability to capture time-varying physiological dynamics.
• Enabled the definition of novel complexity variability indices.

Conclusions:

• The novel entropy measures effectively track instantaneous heartbeat complexity.
• These indices offer a powerful tool for analyzing ANS dynamics in real-time.
• The approach advances the quantification of nonlinear physiological systems.