Characterization of Exercise-Induced Myocardium Growth Using Finite Element Modeling and Bayesian Optimization

Area of Science:

• Cardiovascular Physiology
• Computational Biology
• Biomedical Engineering

Background:

• Cardiomyocyte growth underlies physiological and pathological left ventricular (LV) hypertrophy.
• Understanding the link between organ-level and cellular-level cardiac growth is crucial but poorly understood.
• Computational models offer a way to integrate biomechanics and growth processes.

Purpose of the Study:

• To develop and validate a computational framework for characterizing myocardial growth using in vivo imaging data.
• To quantify exercise-induced cardiomyocyte growth in swine, differentiating between transverse and longitudinal directions.
• To provide a tool for studying myocardial growth in various LV hypertrophy conditions.

Main Methods:

• Development of a framework integrating cardiac magnetic resonance (CMR) imaging, finite element (FE) analysis, and Bayesian optimization.
• Utilized subject-specific LV geometries from exercised porcine models.
• Validated the framework using synthetic LV masks and applied it to characterize growth in swine subjects.

Main Results:

• The framework successfully predicted growth parameters for synthetic LV geometries.
• Exercise-induced growth in swine demonstrated a strong preference for longitudinal cardiomyocyte growth (58.0-79.3%) over transverse growth (4.0-7.8%).
• Quantified longitudinal and transverse growth percentages at 6 and 12 weeks post-exercise initiation.

Conclusions:

• The developed computational framework effectively characterizes myocardial growth in response to exercise.
• Exercise promotes predominantly longitudinal cardiomyocyte growth in swine, offering insights into physiological cardiac remodeling.
• This framework can be applied to diverse LV hypertrophy phenotypes and integrated with other models to explore growth mechanisms.