A dynamic optimization technique for predicting muscle forces in the swing phase of gait

Area of Science:

• Biomechanics
• Human Movement Science
• Computational Biology

Background:

• The muscle force sharing problem is crucial for understanding human locomotion.
• Previous studies often employed static optimization, which may oversimplify muscle dynamics.

Purpose of the Study:

• To compare dynamic and static optimization methods for solving the muscle force sharing problem during the swing phase of gait.
• To investigate the influence of internal muscle dynamics on muscle force control histories.

Main Methods:

• Dynamic optimization using a combined tracking error and metabolic energy consumption objective function (muscle work + enthalpy).
• Static optimization using the sum of cubed muscle stresses, preceded by inverse dynamics calculations.
• Numerical differentiation with a low-pass digital filter for obtaining joint velocities and accelerations.
• Specific algorithms: Fletcher-Reeves conjugate gradient for dynamic, Gradient-restoration for static optimization.

Main Results:

• Dynamic optimization revealed distinct muscle force control histories compared to static optimization.
• Results demonstrated significant sensitivity to the numerical differentiation procedures used in the static approach.
• Internal muscle dynamics were shown to influence the calculated muscle forces during gait.

Conclusions:

• Dynamic optimization provides a more comprehensive understanding of muscle force sharing during gait by incorporating internal muscle dynamics.
• The choice of optimization method and data processing techniques significantly impacts the results of muscle force analysis.