Musculoskeletal Injuries: The Role Of Micro-doppler Radar As A Screening Tool

PURPOSE: Musculoskeletal injuries (MSKI) cause significant health and economic costs. No clinically efficient validated MSKI risk assessment tool exists to identify the high-risk groups and guide limited preventative resources. We assessed the feasibility of using micro-Doppler radar (MDR) as a musculoskeletal injury screening tool for collegiate athletes. METHODS: We completed an IRB approved prospective cohort study on injured and uninjured collegiate athletes from two campuses during preseason physicals. Athletes performed single leg balance, single leg hops, sit to stand, long jumps, and drop box jumps in front of MDR. Multilinear discriminant analysis based features in a nearest neighbor classifier were used to classify the data. RESULTS: Data on 113 participants was collected. 79 (69.9%) athletes were male, 34 (30.0%) were female, and the mean age was 19.76. Athlete breakdown by sport were 26 (20.0%) from men’s soccer, 18 (15.9%) from men’s football, 14 (12.3%) from men’s baseball, 11 (9.7%) from men’s cross country, 10 (8.8%) from women’s soccer, and 9 (7.9%) from women’s volleyball. Other sports included track and field, basketball, field hockey, tennis, and softball. Average data collection time per participant was 4 minutes. The designed classifier had an 83% accuracy in identifying the injured vs un-injured athletes when the training data was included as test samples and a 73% accuracy when training data was excluded. CONCLUSION: MDR is a feasible screening tool capable of accurately identifying a high-risk group for MSKI. It can be operated without the need for special training, is portable, efficient, inexpensive, and rapid. No special marker sets or sensors need to be placed on participants and minimal preparation time is needed before data collection. This technology's speed and practicality is superior to the use of motion capture systems or other current risk assessments. Current and future research will aim to improve data collection efficiency and machine learning algorithm accuracy. Supported by: Funding: Penn State College of Medicine: Artificial Intelligence and Biomedical Informatics Pilot Funding 2023.