Estimation of muscle activity using higher-order derivatives, static optimization, and forward-inverse dynamics

Area of Science:

• Biomechanics
• Musculoskeletal modeling
• Computational neuroscience

Background:

• Accurate estimation of muscle activity is crucial for understanding human movement and diagnosing neuromuscular disorders.
• Existing methods often face challenges with computational cost or accuracy, especially during dynamic activities.

Purpose of the Study:

• To develop a straightforward, accurate, and computationally efficient method for estimating muscle activity.
• To investigate the impact of higher-order derivatives of joint angles and muscle dynamics on estimation accuracy.

Main Methods:

• The proposed method integrates inverse skeletal dynamics, forward muscular dynamics, and static optimization of neural excitation signals.
• It utilizes joint angle and its first to fourth time derivatives as external inputs.
• Constraints for load-sharing problems are calculated using these higher-order derivatives.

Main Results:

• Feasibility demonstrated through simulations of a single-joint musculoskeletal model.
• Analysis showed the influence of muscular dynamics and higher-order derivatives on muscle activity estimation.
• Results highlight the impact of small activation dynamics time constants and the omission of third and fourth derivatives.

Conclusions:

• The novel method provides a promising approach for improving the accuracy of muscle activity estimation.
• It shows particular potential for analyzing highly dynamic human motions.
• The findings underscore the importance of incorporating higher-order kinematic information and muscle dynamics for precise biomechanical analysis.