Development of a Visco-Hyperelastic Constitutive Law for Brain Tissue Based on Finite Element Simulation and Optimization Methodology

The objective of this study was to develop a visco-hyperelastic constitutive law for the FE modeling of brain tissue. Material properties of a visco-hyperelastic constitutive model were determined via the application of FE simulations and optimization methodology. Basic FE models were developed according to referred dynamic uniaxial tension tests of brain tissue with uniform strain rates of 30 s-1 and 90 s-1 up to 30% strain. The optimization objective was to minimize fitting errors between strain-stress curves predicted by FE simulations and corresponding curves measured from referred experiments. Material constants of a visco-hyperelastic constitutive model was defined as constant or design parameters for the optimization procedure, with given values or design domain. Uniform Latin Hypercube algorithm was applied to generate the initial group of FE models, and an improved multi-objective genetic algorithm (MOGA-II) was applied as the optimization strategy. Results indicates that the strain-stress curves predicted by the optimized FE models were located in experimental corridors when the strain higher than 15% for both strain rates of 30 s-1 and 90 s-1. Thus, the proposed visco-hyperelastic constitutive law could be applied in FEHMs for the prediction of intracranial mechanical responses at injury levels.