Jacob Thomas, Rebecca Bliss and Trent Guess

University of Missouri, Columbia, MO

Current methods used to screen for concussions are subjective and not based on movement tasks which simulate the demands of sport. PURPOSE: The purpose of this study was to assess the ability of a novel movement-based concussion screening tool (MPASS) to quantify differences between individuals with and without concussion. METHODS: 17 healthy individuals (22.59±1.66 yrs.) participated in this study. 5 suffered a concussion within 1-month before data collection. 12 had no concussions within the past year. All participants completed the same battery of tasks: walking (control, serial subtraction by seven dual-task, and head-shaking), Romberg balance tests (firm-surface and foam-surface with eyes-open, eyes-closed, and eyes-closed head-shaking), and reaction time. For walking tasks, lower extremity spatiotemporal parameters were collected via Kinect depth-sensing camera. For balance tasks, a force plate recorded center of pressure (CoP) and Kinect recorded center of mass. Reaction time was recorded using an Arduino-based interface board while CoP was simultaneously recorded via force plate. This resulted in 133 unique variables for each participant. Principal component analysis (PCA) was used to reduce the dimensionality of the data. All statistical analyses were completed in RStudio (v4.2.0). RESULTS: 5 retained PCs explained 65.35% of dataset variance. Visual analysis of the first 3 PCs demonstrated clear separation between the two groups (concussion/control) (Fig. 1). Contribution analysis revealed significant contributions from center of mass data during balance tasks, stride and step length during walking, and step length during head-shaking walking. This suggests results from these tasks are most vital in distinguishing individuals with concussion from those without concussion. Further samples will be required before this can be confirmed. CONCLUSION: This study demonstrates promising initial results for concussion screening using MPASS.

This document is currently not available here.