| Dynamic muscle functions remain elusive in spite of tremendous advances in the areas of computer simulation and dynamical systems analysis. Forward dynamic simulations of movement provide a general description of dynamic muscle function but are dependent on the performance criterion assumed, the complexity of the underlying biomechanical model, and assumptions made regarding properties of the system (segment inertias, joint degrees of freedom, muscle moment arms and passive structures involved). This dissertation presents two novel electrical stimulation methodologies for the evaluation of dynamic muscle function: one stimulates the muscle from a static limb posture, the other stimulates the muscle during gait. These methodologies were used to assess the dynamic function of the biarticular rectus femoris and semitendinosus muscles during walking. Stimulation of the rectus femoris during pre-swing limited peak knee flexion during swing, with a corresponding reduction in hip flexion. Stimulation of the rectus femoris during early swing had a reduced effect. Stimulation of the semitendinosus during terminal swing caused the knee to flex more during stance, with little or no change at the hip. Stimulation of the semitendinosus during early stance had insignificant effects on hip and knee angles during stance. Our experimental results were compared to those of forward dynamic simulations that were similarly perturbed. Overall, the perturbed simulations reflected the general patterns of induced motion but had much lower variability than the experimental data. Our results confirm some of the non-intuitive predictions reached via dynamic simulations and have implications to our understanding of both normal and impaired walking, such as in stiff-knee and crouch gaits. |
