In Vivo cardiac biomechanical model parameter estimation from ultrafast shear wave elastography

Quantifying myocardial stiffness (MS) over the full cardiac cycle offers a direct window into tissue structure and mechanics, yet translating time-resolved shear wave elastography (SWE) into intrinsic material properties remains a complex inverse problem. While SWE provides detailed stiffness quantification, these measurements combine passive structural properties with active contractility under subject-specific loading conditions. In this work, we present a computational framework to decompose these contributions using a closed-loop reduced-order electromechanical model. We formulate the personalization process as a sequential inverse problem: first, a passive hyperelastic baseline is identified from the late-diastolic window; second, active contractile parameters are estimated by fitting the full-cycle waveform using a regularized covariance matrix adaptation evolution strategy (CMA-ES). We validated this framework on a cohort of healthy pediatric volunteers (n=20). The personalized models reproduced subject-specific stiffness trajectories with high fidelity (median cross-correlation 0.997, nRMSE 1.02). Uncertainty analysis confirmed the practical identifiability of the active parameters, with stable convergence across random initializations. This study establishes a computationally efficient, robust pipeline for converting non-invasive SWE assessments into interpretable mechanical signatures, paving the way for physics-informed phenotyping in clinical cohorts.