Locating Order-Disorder Phase Transition in a Cardiac System

Hiroshi Ashikaga, Ameneh Asgari-Targhi

Abstract

To prevent sudden cardiac death, predicting where in the cardiac system an order-disorder phase transition into ventricular fibrillation begins is as important as when it begins. We present a computationally efficient, information-theoretic approach to predicting the locations of wavebreaks that initiate fibrillation in a cardiac system where the order-disorder behavior is controlled by a single driving component, mimicking electrical misfiring from the pulmonary veins or the Purkinje fibers. Communication analysis between the driving component and each component of the system reveals that channel capacity, mutual information and transfer entropy can locate the wavebreaks. This approach is applicable to interventional therapies to prevent sudden death, as well as to a wide range of systems to mitigate or prevent imminent phase transitions.

Introduction

Sudden cardiac death, primarily driven by ventricular fibrillation, remains a leading cause of mortality globally, accounting for nearly 20% of all deaths. Fibrillation represents an order-disorder phase transition where the heart's synchronized electrical activity collapses into chaotic wave patterns. While significant research has focused on identifying "early warning signs" to predict when such transitions occur, determining where they originate within the cardiac tissue is equally vital for clinical interventions. Specifically, identifying the spatial initiation of disorder is crucial for targeted therapies like catheter ablation, which aim to neutralize the specific tissue regions responsible for triggering arrhythmias.

Methodology: The Heart as a Communication Network

This study introduces a novel information-theoretic framework that treats the heart as a communication system. In this model, a "driving component"—such as the sinoatrial node or an ectopic trigger—acts as the sender, while individual cardiac cells function as receivers.

Figure 1: Information-theoretic modeling of cardiac dynamics. (A) Binary encoding of action potentials based on excitation states. (B) Block diagram of the cardiac communication system including noise sources. (C) The binary asymmetric channel model with error probabilities ε₀ and ε₁.

The cardiac microstate is represented as a binary time series: a value of '1' indicates an excited state (action potential), while '0' represents a resting state. The communication channel between cells is modeled as a binary asymmetric channel. In this context, intrinsic physiological properties—such as the restitution curve (recovery rate), action potential duration (APD) alternans, and wavefront curvature—are treated as sources of "noise" that can induce transmission errors, such as flipping a 0 to a 1 or vice versa.

Four primary metrics were employed to quantify the information flow:

  1. Shannon Entropy: Measures the uncertainty or information content of a single component's state.
  2. Mutual Information (MI): Quantifies the shared information between the driver and a local tissue component.
  3. Channel Capacity: Defines the theoretical maximum rate of reliable information transmission.
  4. Transfer Entropy (TE): Measures the directed, time-lagged information flow from the driver to the tissue.

The framework was validated using the Fenton-Karma model, a three-variable ionic model capable of reproducing complex cardiac dynamics. Simulations were conducted on 1-D cables and 2-D lattices to observe the emergence of "wavebreaks"—the critical points where wavefronts and wavebacks intersect, initiating disorder.

Key Insights and Results

The study reveals that biological properties like the restitution curve act as fundamental constraints on communication. As the system approaches fibrillation, channel capacity and mutual information function as "order parameters," progressively declining and reaching zero precisely at the boundary of the phase transition.

In 2-D simulations, channel capacity successfully demarcated a circular border that predicted the location of subsequent wavebreaks. Notably, this predicted region was located "upstream" of the actual physical wavebreaks, suggesting that these metrics can identify the "frontline" of transition before chaotic behavior fully manifests. Furthermore, Transfer Entropy was found to peak just prior to the transition, serving as a high-sensitivity locator for imminent disorder.

Even in uniform, isotropic tissue models, the dynamics of wave propagation create "functional heterogeneity." Some regions lose communication with the driver faster than others based on their geometric relationship to the stimulus, highlighting that the spatial initiation of fibrillation is a dynamic property of the system's state.

Conclusion and Future Directions

This information-theoretic approach offers a computationally efficient method for mapping the spatial initiation of cardiac disorder. By focusing on the communication between a single driver and individual components, the method avoids the complexity of all-to-all interaction analysis, making it potentially suitable for real-time clinical applications.

Future research will focus on extending this framework to biophysically detailed 3D cardiac models and evaluating its robustness against the higher noise levels inherent in clinical data, such as electrocardiograms or optical mapping. Beyond cardiology, this methodology holds promise for predicting phase transitions in other complex networked systems, including power grids and financial markets.

Summary updated

Abstract

To prevent sudden cardiac death, predicting where in the cardiac system an order-disorder phase transition into ventricular fibrillation begins is as important as when it begins. We present a computationally efficient, information-theoretic approach to predicting the locations of wavebreaks that initiate fibrillation in a cardiac system where the order-disorder behavior is controlled by a single driving component, mimicking electrical misfiring from the pulmonary veins or the Purkinje fibers. Communication analysis between the driving component and each component of the system reveals that channel capacity, mutual information and transfer entropy can locate the wavebreaks. This approach is applicable to interventional therapies to prevent sudden death, as well as to a wide range of systems to mitigate or prevent imminent phase transitions.
BibTeX
@inproceedings{ashikaga2018locating,
  author = {Hiroshi Ashikaga and Ameneh Asgari-Targhi},
  title = {{Locating Order-Disorder Phase Transition in a Cardiac System}},
  booktitle = {Scientific},
  year = {2018},
  doi = {10.1038/s41598-018-20109-6},
}
Figure