Assessing Muscle Driven Optimization Techniques in OpenSim for Long Distance Running and Deep Muscle Adaptations

Understanding how deep muscles contribute to running performance and adapt to accumulated distance remains underexplored in biomechanics. This study addresses this gap by validating muscle-driven optimization techniques in OpenSim, focusing on the behavior of five deep muscles–adductor magnus ischial, sartorius, semitendinosus, vastus intermedius, and extensor digitorum longus–during long-distance running. Biomechanical data were collected for run distances of 0 km, 40 km, 80 km, and 120 km through motion capture, vertical ground reaction force (VGRF) measurement, and electromyogram (EMG) analysis. Muscle-driven simulations were conducted using Static Optimization (SO), Computed Muscle Control (CMC), and MocoInverse techniques in OpenSim, with validation against experimental EMG data. The results demonstrated that CMC provided the most accurate muscle force estimations, exhibiting the lowest root mean square error (RMSE) and highest correlation, though at the cost of increased computational time. Analysis revealed significant changes in muscle force generation at 80 km, indicating the body's adaptation to accumulated running distance. Muscles like the sartorius and semitendinosus exhibited compensatory force generation, while the adductor magnus ischial showed adaptive shifts between stance and swing phases. In conclusion, CMC was identified as the most reliable optimization technique for muscle-driven simulations in OpenSim. The study highlights the biomechanical adjustments during long-distance running and the role of deep muscles in maintaining running efficiency as the body adapts to prolonged activity.