In Vivo Cardiac Biomechanical Model Parameter Estimation from Ultrafast Shear Wave Elastography
Abstract
Shear wave elastography (SWE) enables non-invasive acquisition of cycle-resolved myocardial stiffness dynamics, but translating the resulting stiffness waveforms into interpretable mechanical parameters remains a challenging inverse problem. In vivo SWE stiffness reflects the combined effects of passive tissue properties, active contraction, and subject-specific loading conditions, making separation of passive and active contributions essential for distinguishing alterations in myocardial material properties from changes in contractile function. We developed a computational framework based on a closed-loop reduced-order electromechanical model to address this problem. Personalisation was formulated as a sequential inverse problem. An effective passive stiffness contribution was first identified from the diastolic portion of the SWE waveform. The active parameters were then estimated from the residual full-cycle stiffness dynamics. The parameter k 0 defines the maximum active stiffness, while k ATP and k SR govern contraction and relaxation rates, respectively. Parameter estimation used a cohort-regularised objective optimized with the covariance matrix adaptation evolution strategy (CMA-ES). The framework was evaluated in 20 healthy paediatric volunteers, achieving a median active-target mean absolute error of 1.19 kPa and a median correlation of 0.995 for the retained subject-specific solutions. Repeated seeded optimisations yielded highly reproducible objective values, while k ATP showed the greatest across-seed variability, indicating lower identifiability than the other active parameters. These results demonstrate the feasibility of converting non-invasive SWE measurements into model-consistent mechanical signatures and provide a basis for physics-informed myocardial phenotyping in clinical cohorts.
Related
- Uncertainty Quantification in Cardiac Model Personalisation from Ultrafast Ultrasound
- Surrogate Gradients for Gradient-Based Parameter Estimation in Simplified Neuron Models
- EG-BINN: Parameter-Efficient Ensembling of Gated Biologically Informed Neural Networks for Epigenetic Age Estimation
- PyFLI: A Python Library for Simulation, Parameter Estimation, and Benchmarking in Fluorescence Lifetime Imaging
- 2D reaction-diffusion model-based biopsy simulation for dynamic tumor growth parameter estimation
BibTeX
@article{ferrario2026vivo,
title = {{In Vivo Cardiac Biomechanical Model Parameter Estimation from Ultrafast Shear Wave Elastography}},
author = {Camilla Ferrario and Maelys Venet and Olivier Villemain and Maxime Sermesant},
journal = {Biomechanics and Modeling in Mechanobiology},
year = {2026},
month = sep,
url = {https://hal.science/hal-05836609v1},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - In Vivo Cardiac Biomechanical Model Parameter Estimation from Ultrafast Shear Wave Elastography
AU - Camilla Ferrario
AU - Maelys Venet
AU - Olivier Villemain
AU - Maxime Sermesant
PY - 2026
DA - 2026/09/19
JO - Biomechanics and Modeling in Mechanobiology
UR - https://hal.science/hal-05836609v1
AB - Shear wave elastography (SWE) enables non-invasive acquisition of cycle-resolved myocardial stiffness dynamics, but translating the resulting stiffness waveforms into interpretable mechanical parameters remains a challenging inverse problem. In vivo SWE stiffness reflects the combined effects of passive tissue properties, active contraction, and subject-specific loading conditions, making separation of passive and active contributions essential for distinguishing alterations in myocardial material properties from changes in contractile function. We developed a computational framework based on a closed-loop reduced-order electromechanical model to address this problem. Personalisation was formulated as a sequential inverse problem. An effective passive stiffness contribution was first identified from the diastolic portion of the SWE waveform. The active parameters were then estimated from the residual full-cycle stiffness dynamics. The parameter k 0 defines the maximum active stiffness, while k ATP and k SR govern contraction and relaxation rates, respectively. Parameter estimation used a cohort-regularised objective optimized with the covariance matrix adaptation evolution strategy (CMA-ES). The framework was evaluated in 20 healthy paediatric volunteers, achieving a median active-target mean absolute error of 1.19 kPa and a median correlation of 0.995 for the retained subject-specific solutions. Repeated seeded optimisations yielded highly reproducible objective values, while k ATP showed the greatest across-seed variability, indicating lower identifiability than the other active parameters. These results demonstrate the feasibility of converting non-invasive SWE measurements into model-consistent mechanical signatures and provide a basis for physics-informed myocardial phenotyping in clinical cohorts.
SN - 1617-7959
SN - 1617-7940
ER -