Development and Validation of Subject-Specific 3D Human Head Models Based on a Nonlinear Visco-Hyperelastic Constitutive Framework

Abstract

Computational models of the human head are promising tools for the study and prediction of traumatic brain injuries (TBIs). Most available head models are developed using inputs (i.e., head geometry, material properties, and boundary conditions) derived from ex-vivo experiments on cadavers or animals and employ linear viscoelasticity (LVE)-based constitutive models, which leads to high uncertainty and poor accuracy in capturing the nonlinear response of brain tissue under impulsive loading conditions. To resolve these issues, a framework for the development of fully subject-specific 3D human head models is proposed, in which model inputs are derived from the same living human subject using a comprehensive in-vivo brain imaging protocol, and the viscous dissipation-based visco-hyperelastic constitutive modeling framework is employed. Specifically, brain tissue material properties are derived from in-vivo magnetic resonance elastography (MRE), and full-field strain-response of brain under rapid rotational acceleration is obtained from tagged MRI, which is used for model validation. The constitutive model comprises the Ogden hyperelastic strain energy density and the Upadhyay-Subhash-Spearot viscous dissipation potential. The simulated strain-response is compared with experimental data and with predictions from subject-specific models employing two commonly used LVE-based constitutive models, using a rigorous validation procedure that evaluates agreement in spatial strain distribution, temporal strain evolution, and differences in maximum values of peak and average strain. Results show that the head model developed in this work reasonably captures 3D brain dynamics, and when compared to LVE-based models, provides improvements in the prediction of peak strains and temporal strain evolution.

The white matter fiber tract deforms most in the perpendicular direction duringin vivovolunteer impacts

“…This may be particularly relevant for those models with intended usage to predict the brain response under non-injurious impacts. Emerging attempts have been noted in validating the computational head model against in vivo brain responses in the form of brain-skull relative displacement [69], maximum principal strain [21,22,67,70], maximum shear strain [67], and tractoriented normal strain [21,22]. Our study computes 3D normal and shear strains along and perpendicular to the fiber tract (spatial resolution: 1 mm or 2 mm, temporal resolution: 18 ms, 19.5 ms, or 20 ms) across the whole WM from 44 volunteer impacts.…”

“…In silico head models are increasingly used to quantitatively simulate the localized tissue response during head impacts [16,17]. Modern in silico models feature subject-specific geometry [18,19], anisotropic and heterogeneous brain material [20][21][22], and fluid representation of the cerebrospinal fluid [23,24]. Encouraging correlations have been reported between the response of the in silico model and experimentally quantified histology or clinically diagnosed injury [25][26][27][28].…”