Using individual-muscle specific instead of across-muscle mean data halves muscle simulation error

Area of Science:

• Biomechanics
• Computational Physiology
• Motor Control

Background:

• Hill-type muscle models are widely used to simulate muscle behavior.
• Parameter values in these models often exhibit significant variation across different muscles.
• Standard practice typically employs across-muscle mean values, potentially limiting model accuracy.

Purpose of the Study:

• To investigate whether using muscle-specific Hill-type parameter values improves muscle model performance compared to using mean values.
• To determine the impact of maximum muscle force (Fmax) parameter specificity on model accuracy.
• To identify which specific parameters are most critical for enhancing muscle model predictive power.

Main Methods:

• Simulated muscle contractions under various motor nerve stimulation paradigms (isotonic and isometric).
• Compared normalized Root Mean Square (RMS) error between models using mean parameter values and muscle-specific parameter values.
• Systematically evaluated the effect of using muscle-specific Fmax versus other muscle-specific parameters on simulation accuracy.

Main Results:

• Using mean parameter values significantly increased simulation error (doubling it from 9 to 18, p < 0.0001).
• Muscle-specific Fmax values alone did not improve model performance compared to using all mean values.
• Using muscle-specific values for parameters other than Fmax significantly reduced error (to 14, p ≤ 0.014), with best performance achieved using all muscle-specific parameters.

Conclusions:

• Muscle-specific measurement of Hill-type model parameters is essential for constructing highly accurate muscle models.
• Improved model performance necessitates muscle-specific values for parameters beyond maximum force.
• Remaining simulation errors may be attributed to complex motor neuron activation and model activation dynamics.