Biomechanics-guided engineering of in vitro cardiac tissues for mechanical function reconstruction.

Engineered in vitro cardiac tissues are being developed and increasingly used for disease modeling and drug evaluation, but reproducing cardiac cell composition and three-dimensional architecture alone does not ensure physiologically meaningful function. Native myocardial performance arises from coordinated electrical activation, active force generation, and passive mechanical resistance governed by stiffness, anisotropy, nonlinear elasticity, and viscoelastic relaxation. Failure to reproduce these coupled properties can alter force transmission, tissue deformation, mechanosensitive cellular responses, and the interpretation of disease phenotypes or drug effects. Yet myocardial mechanics, computational modeling, biomaterial design, and biofabrication are commonly developed as separate domains. This review integrates these areas through a biomechanics-guided property-parameter-function framework. We first examine the multiscale structural determinants of myocardial mechanics and critically compare cellular, continuum, phenomenological, microstructure-informed, viscoelastic, and pathology-informed models with respect to predictive scope, computational demand, parameter identifiability, experimental validation, and engineering relevance. We then show how model-derived descriptors-including active stress, baseline and nonlinear stiffness, anisotropy, relaxation behavior, and remodeling-related parameters-can be translated into controllable biomaterial properties, scaffold architecture, fabrication and conditioning protocols, and mechanical boundary conditions. Emerging approaches involving artificial intelligence, inverse design, uncertainty quantification, and digital twins are discussed as routes toward adaptive and patient-relevant tissue models. By positioning computational biomechanics as a design and validation tool, this review provides a systematic basis for engineering cardiac tissues with more predictable mechanical function and more reproducible disease- and drug-response readouts.