Annals of Biomedical Engineering Biomedical Engineering
Limitations of Rotational Acceleration for Predicting Concussion: A New Injury Risk Function
Abstract
Abstract Purpose Rotational acceleration of the head has long been hypothesized to be a primary driver of concussion because it can generate shear strains in brain tissue, while linear acceleration has been considered less injurious due to the brain’s near-incompressibility. Instrumented mouthguards (iMGs) now enable direct six-degree-of-freedom measurements of human concussion, making it possible to rigorously test this hypothesis for the first time in vivo . The purpose of this study was to test the hypothesis that rotational acceleration is a more precise predictor of concussion than linear acceleration using iMG measurements of concussion across a multi-sport dataset. Methods A dataset of 47 concussive and 3,804 non-concussive iMG-measured impacts was compiled across male and female athletes in seven sports. Univariate and bivariate logistic regression models were developed using peak linear acceleration, peak rotational acceleration, and peak rotational velocity to assess concussion classification and contribution to injury risk. Models were evaluated using area under the precision–recall curve (AUPRC), F1 score, and odds ratios. Additionally, a novel liquid-filled helmet technology was evaluated during laboratory impacts using the derived risk functions. Results Contrary to our hypothesis, rotational acceleration was not a significantly more precise classifier than linear acceleration (AUPRC = 0.36 vs 0.65; F1 = 0.31 vs 0.50) and was not a significant independent predictor of concussion alongside linear acceleration in a bivariate model (OR rot = 1.3, 95% CI 0.81–1.9; OR lin = 2.3; 95% CI 1.5–3.6). However, peak rotational velocity provided additional predictive value in a bivariate risk function with peak linear acceleration. Applying this bivariate function to laboratory football helmet tests, the liquid-filled padding reduced predicted concussion risk by up to 60% relative to a standard American football helmet, primarily due to attenuating linear acceleration. Conclusion Rotational acceleration was not a significant independent predictor of concussion and was outperformed by linear acceleration in classification, challenging its primacy in injury risk evaluation. Our results support a broader consideration of linear acceleration and rotational velocity in concussion risk assessment and prevention.