Development and Validation of a New Anisotropic Visco-Hyperelastic Human Head Finite Element Model Capable of Predicting Multiple Brain Injuries

Ding Lyu; Runzhou Zhou; Chin-Hsu Lin; Priya Prasad; Liying Zhang

doi:10.3389/fbioe.2022.831595

Development and Validation of a New Anisotropic Visco-Hyperelastic Human Head Finite Element Model Capable of Predicting Multiple Brain Injuries

Front Bioeng Biotechnol. 2022 Mar 24:10:831595. doi: 10.3389/fbioe.2022.831595. eCollection 2022.

Authors

Ding Lyu¹, Runzhou Zhou¹, Chin-Hsu Lin², Priya Prasad³, Liying Zhang¹

Affiliations

¹ Department of Biomedical Engineering, Wayne State University, Detroit, MI, United States.
² General Motors R&D Center, Warren, MI, United States.
³ Prasad Engineering, LLC, Plymouth, MI, United States.

Abstract

This paper reports on the latest refinement of the Finite Element Global Human Body Models Consortium 50th percentile (GHBMC M50) adult male head model by the development and incorporation of a new material model into the white matter tissue of the brain. The white matter is represented by an anisotropic visco-hyperelastic material model capable of simulating direction-dependent response of the brain tissue to further improve the bio-fidelity and injury predictive capability of the model. The parameters representing the material were optimized by comparing model responses to seven experimentally reported strain responses of brains of postmortem human subjects (PMHS) subjected to head impact. The head model was subjected to rigorous validation against experimental data on force-deflection responses in the skull and face, intracranial pressure, and brain strain responses from over 34 PMHS head impact experiments. Crash-induced injury indices (CIIs) for facial bone fracture, skull fracture, cerebral contusion, acute subdural hematomas (ASDHs), and diffuse brain injury were developed by reconstructing 32 PMHS and real-world injury cases with the model. Model predicted maximum principal strain (MPS) and stress were determined as fracture CIIs for compact bone and spongy bones, respectively, in the skull and face. Brain responses in terms of MPS, MPS rates, and pressure distribution in injury producing experimental impacts were determined using the model and analyzed with logistic regression and survival analysis to develop CIIs for brain contusions, diffuse brain injuries, and ASDH. The statistical models using logistic regression and survival analysis showed high accuracy with area under the receiver operating curve greater than 0.8. Because of lack of sufficient moderate diffuse brain injury data, a statistical model was not created, but all indications are that the MPS rate is an essential brain response that discriminates between moderate and severe brain injuries. The authors stated that the current GHBMC M50 v.6.0 is an advanced tool for injury prediction and mitigation of injuries in automotive crashes, sports, recreational, and military environments.

Keywords: GHBMC human head model; brain injury; brain strain validation; crash induced injury index; facial injury; finite element analysis; risk function; skull injury.