2018 Volume 3 Pages 203-208
The purpose of this study was to construct a system to estimate brain concussion risk based on video footage of collisions during American football games. In this system, the collision motions of collegiate football players are reproduced from the video by using motion analysis of a whole-body model, and the mechanical parameters thus obtained are input into two helmeted finite-element human head models as the initial condition of the collision analysis. The injury risk of each collision is then estimated using 9 risk predictors based on the kinematic behavior of the head or the deformation of the brain. In this study, 17 head impact accidents (9 concussion cases and 8 non-injured cases) were reproduced using this system to determine the injury risk functions using logistic regression analysis. These analyses led to the conclusion that the estimated risk in the injured cases was higher than in the non-injured cases, when using a new risk predictor that combines translational acceleration of the head and von Mises stress in the brain compared with that when using either of the predictors alone.
