A Bayesian constitutive model selection framework for biaxial mechanical testing of planar soft tissues: Application to porcine aortic valves

Planar soft tissue mechanics were evaluated using biaxial experimental data collected from porcine aortic valve cusp samples. Mechanical loading was performed under multiple stress ratios, with deformation measured using integrated imaging mechanical testing to capture in-plane tissue response during heart valve biaxial testing. The resulting datasets were analyzed using principal component analysis and Bayesian probability methods to compare the relative likelihood of several hyperelastic constitutive formulations. Model selection was based on how well each formulation represented variability across samples and loading conditions, rather than on deterministic curve fitting alone. Across the set of candidate models, the May–Newman formulation showed the highest likelihood for describing the observed biaxial response. The framework provides a structured approach for comparing constitutive descriptions of valve tissue mechanics using probabilistic criteria.