Validation of Participant Specific OpenSim Model for Lower Extremity Gait Biomechanics

• Mack Rubano, Biomedical Engineering, University of Delaware

• Stephanie Cone, Biomedical Engineering, University of Delaware

Athletes with anterior cruciate ligament (ACL) ruptures are 6 times more likely to re-rupture compared to uninjured peers upon return to sport. We aim to inform a previously developed OpenSim model with gold-standard motion capture trajectories for eventual use in assessment of ACL rupture risk. Data from a participant enrolled in a prior study was used to inform this OpenSim model. The participant was fit with 39 retroreflective markers, and 8 Qualisys Oqus 500 cameras were used to track musculoskeletal movement during a walking gait trial on an instrumented treadmill. We used the armless Rajagopal model, which is a highly regarded musculoskeletal figure in OpenSim for data analysis. First, the generic model was scaled to the participant’s musculoskeletal structure. Next, inverse kinematics was performed to compute knee, hip, and ankle joint angles. Finally, inverse dynamics was used to calculate knee, hip, and ankle kinetics using force data from the instrumented treadmill. Comparisons between motion capture and OpenSim measurements revealed exceptional tracking in lower extremity kinematics, specifically at the knee and hip. Root mean square errors (RMSE) for the right and left knee angles were below the minimal clinically relevant range (1.44 and 3.55 degrees, respectively) with high correlation (R ≥ 0.998), while the right and left hip angles showed a low RMSE (2.02 and 1.06 degrees, respectively) and strong correlation (R ≥ 0.993). Conversely, ankle range of motion showed tracking errors, exhibiting high RMSE values (32.26 and 29.88 degrees, respectively) but strong correlations (R ≥ 0.974). Joint kinetics showed greater RMSE, which is likely due to compounding errors in the inverse dynamics calculations. Ultimately, these findings support the hypothesis that the proposed OpenSim model provides a reliable framework for analyzing lower-extremity musculoskeletal dynamics during movement, but could benefit from further development.