Reduced Cellular Model for Cardiac Electromechanical Simulations

Helena Carvalho Lannes Corrêa Salles1, Joao Gabriel Rocha Silva2, Carolina Ribeiro Xavier3, Bernardo Martins Rocha4, Joventino de Oliveira Campos5, Rodrigo Weber dos Santos5
1UFJF, 2Federal Institute of Brasília, 3UFSJ, 4Universidade Federal de Juiz de Fora, 5Federal University of Juiz de Fora


Abstract

Introduction: Despite significant advances in cardiac modeling, the computational demands of electromechanical simulations continue to limit their clinical translation. We introduce the MMS model, a biophysically-constrained reduced-order formulation that achieves substantial computational efficiency without compromising physiological accuracy.

Objective: To develop and validate a minimal cardiac electromechanical model that faithfully reproduces action potential morphology, calcium handling dynamics, and active tension generation while demonstrating superior computational performance relative to established biophysical models.

Methods: The proposed model integrates: (1) a modified Bueno-Orovio electrophysiology framework, (2) phenomenological calcium dynamics, and (3) simplified active tension generation. Parameter optimization employed a multi-objective genetic algorithm to match ToR-Ord-Land outputs. Validation included 3D electromechanical simulations on a hexahedral mesh (6,859 nodes) using GPU-accelerated finite element methods.

Results:
MMS demonstrated excellent agreement with the reference model (mean error <4% across all outputs) while achieving a 6.5x overall speedup. The ODE solver showed particular efficiency 290x faster), enabling larger time steps (0.05 ms vs 0.001 ms) without loss of numerical stability.

Conclusion: This work establishes MMS as a computationally efficient alternative for cardiac electromechanical simulations, successfully balancing model complexity with physiological fidelity. The demonstrated performance gains enable practical applications in whole-organ modeling and clinical translation.