Abstract:
Numerous human head finite element (FE) models have been developed by various research groups to study the complex biomechanical mechanisms underlying traumatic brain injury (TBI). These models, however, exhibit substantial variations in their features and parameters, leading to potential differences in their predicted intracranial mechanical responses. This study aims to conduct a comprehensive parametric comparison of three independently validated FE models of the human head to assess the inter-model variability in predicted brain tissue-level mechanical responses, thereby contributing to a better understanding of the biomechanics of TBI. The three models selected for this investigation — the Simulated Injury Monitor (SIM), the University College Dublin Brain Injury Model (UCDBIM), and the Global Human Body Models Consortium (GHBMC) brain FE model — were subjected to identical impact conditions covering a range of linear and rotational acceleration profiles. While all three models have demonstrated biofidelic validation against cadaveric experimental data at the head-level response, their intracranial predictions, particularly regarding maximum principal strain (MPS) and maximum principal strain rate (MPSr), exhibited significant differences. Specifically, the results indicated varying sensitivities of MPS and MPSr to key model parameters, with differences in brain material properties, skull-brain interface representation, and element types being identified as major contributors to the observed inter-model variability. This biomechanical analysis revealed that while correlations existed between head-level kinematics and intracranial mechanical responses, the relationships varied significantly across the models, emphasizing the impact of underlying model assumptions on predicted tissue-level deformation and stress. The findings highlight the importance of understanding the influence of model specific features on biomechanical predictions and suggest the need for further efforts in standardization and enhanced validation of intracranial responses to improve the predictive capabilities of human head FE models for TBI risk assessment.
