Proposed Reformulation of Brain Injury Criteria (BrIC) Using Head Rotation-Induced Brain Injury Thresholds Simulated and Derived Directly from A Subhuman Primate Finite Element Model.

Abstract

Recent studies have found that Brain Injury Criteria (BrIC) grossly overpredicts instances of real-world, severe traumatic brain injury (TBI). However, as it stands, BrIC is the leading candidate for a rotational head kinematics-based brain injury criteria for use in automotive regulation and general safety standards. This study attempts to understand why BrIC overpredicts the likelihood of brain injury by presenting a comprehensive analysis of live primate head impact experiments conducted by Stalnaker et al. (1977) and the University of Pennsylvania before applying these injurious conditions to a finite element (FE) monkey model. Data collection included a thorough analysis and digitization of the head impact dynamics and resulting pathology reports from Stalnaker et al. (1977) as well as a representative reconstruction of the Penn II baboon diffuse axonal injury (DAI) model. Computational modeling techniques were employed on a FE Rhesus monkey model, first introduced by Arora et al. (2019), to derive risk related brain tissue strain thresholds from the laboratory data. The existing critical velocities proposed for BrIC were then scaled until the target strain level associated with each severity level of diffuse brain injury was reproduced in the FE model of the human brain. Overall, this study provides a comprehensive understanding of these two historical non-human primates (NHP) models and predicts a strain based diffuse tissue injury threshold (MPS99.9) of 1.0 and 1.6 for concussion (mild TBI) and DAI (severe TBI), respectively. The findings indicate scale factors of 1.6 to 5.9 times the original BrIC critical velocities, depending on the loading duration, are required to predict severe (AIS 4+) diffuse brain injury. These results allude to a necessity for including angular acceleration and duration as kinematic parameters in an injury criterion that can accurately predict real-world, diffuse brain injuries. This study also attempts to evaluate and recommend a methodology for post-processing strain parameters produced by head models, settling on the use of MPS99.9 and CSDM50.