Meeting Abstract
Although a large amount of effort has been put towards modeling muscle forces, muscle models are still unable to predict forces during dynamic animal movements, especially with perturbations (e.g. moving over obstacles). In this study, we use a novel model inspired by the winding-filament hypothesis to predict muscle forces in running guinea fowl. Lateral gastrocnemius lengths, activations and forces were measured using sonomicrometry, implanted EMG, and tendon buckles, respectively. Guinea fowl (n = 2) were recorded running on a treadmill during level running and running over 5cm and 7cm obstacles. Muscle morphology parameters included pennation angle, muscle mass and muscle fascicle resting length. The EMG was smoothed, transformed to a percentage of maximum activation, and shifted by a time delay to account for excitation-contraction coupling. The winding filament model (WFM) consists of a contractile element in series with a spring. The contractile element is also in series and parallel with a second spring and damper, representing the titin protein, which wraps around a pulley representing actin thin filaments. Muscle length and activation are inputs to the model, and muscle force is predicted in each time step. The predicted forces were compared to measured forces. The free parameters (n = 6), including an activation factor that varied from trial to trial, were optimized locally and globally using a high-performance computer. Results show that the WFM-based model more accurately predicts forces during perturbed and level gaits (R2 = 0.72-0.86) than published results using complex Hill-type models. Biological relevance of the model was assessed by evaluating input parameters, internal model variables, and sensitivity analysis.
