Anisotropic hyperelastic properties of porcine pericardium under equibiaxial loading: implications for aortic valve design.

Background and objective: Bioprosthetic aortic valves with leaflets made from porcine-derived soft tissue are used to treat aortic valve disease. To improve the durability and hemodynamic performance of bioprosthetic aortic valves, it is important to understand the biomechanics of these tissues and to assess the hemodynamic performance by quantifying orifice area, downstream jet metrics, and flow parameters. In this study, we aim to investigate the mechanical properties of porcine pericardium, derive an appropriate constitutive model, and apply it in fluid-structure interaction (FSI) simulations to evaluate the hemodynamic performance of a bioprosthetic aortic valve device. Methods: Porcine pericardium was characterized using equibiaxial tensile testing. Immediately after the slaughter of 51-week-old pigs, nine hearts with intact pericardium were carefully dissected. After performing equibiaxial tensile tests, the anisotropic hyperelastic behavior of porcine pericardium was characterized; ten anisotropic constitutive models were then used to identify the material parameters. Subsequently, these parameters were used in FSI computations of a bioprosthetic valve. Results: Using an evaluation index that considered only the coefficient of determination, it was found that the Fung model, four-fiber-family model, and HGO model fitted the data better than the other models. For the FSI simulations, the Holzapfel (2005) model and the four-fiber-family models were selected because their explicitly fiber-reinforced formulations allow the constitutive description to be directly linked to the prescribed leaflet fiber architecture. The selected models were then used for FSI modeling of aortic valve hemodynamics, and the computed results were consistent with literature-based numerical data. The peak velocity was 1.77 m/s at 323 ms for the Holzapfel (2005) model and 1.94 m/s at 315 ms for the four-fiber-family model. The difference was 0.17 m/s, corresponding to 9%, while the waveform shapes remained qualitatively consistent across models, with a rapid systolic increase followed by a gradual decay during ejection. The obtained values fall within the range expected for normal aortic valve function. Conclusion: To fully characterize the mechanical response of porcine pericardium and identify the most predictive constitutive model, it is recommended to combine uniaxial tension, shear, and biaxial tests rather than relying on a single loading mode.