Which data should be tracked in forward-dynamic optimisation to best predict muscle forces in a pathological...

Area of Science:

• Biomechanics
• Musculoskeletal modeling
• Neuromuscular disorders

Background:

• Accurate prediction of muscle forces is crucial for understanding movement, particularly in individuals with neuromuscular disorders.
• Current forward dynamics optimization methods often rely on joint kinematics or torques, with less emphasis on tracking marker trajectories or electromyography (EMG).

Purpose of the Study:

• To evaluate the effectiveness of different tracking objective functions within a forward dynamics framework for accurately predicting upper-limb muscle forces.
• To determine if incorporating electromyography (EMG) data improves muscle force prediction compared to traditional methods.

Main Methods:

• A musculoskeletal model was developed to simulate a shoulder abduction movement.
• Simulated EMG and marker trajectories, incorporating realistic noise (Gaussian noise for EMG, soft tissue artifacts for markers), were used as reference data.
• Six different non-linear least-squared objective functions were implemented within a forward dynamics approach, varying the tracked data (marker trajectories, joint angles, torques, with and without EMG).
• A direct multiple shooting algorithm was employed to solve the optimization problems.

Main Results:

• The approach combining marker trajectory and EMG tracking yielded the lowest root-mean-square error (RMSe) for muscle forces (18.45 ± 12.60 N).
• This combined approach demonstrated an almost five-fold improvement in force prediction accuracy compared to methods relying solely on joint angles (82.37 ± 66.26 N) or torques (85.10 ± 116.40 N).

Conclusions:

• Integrating electromyography (EMG) data as a complementary tracking objective in forward dynamics significantly enhances the accuracy of muscle force estimation.
• This combined approach shows promise for improving the prediction of muscle forces, especially in complex scenarios and for patients with neuromuscular disorders.