Modeling muscle function using experimentally determined subject-specific muscle properties.

Area of Science:

• Biomechanics
• Muscle Physiology
• Computational Modeling

Background:

• Muscle models often use averaged properties, limiting accuracy.
• Validation against direct muscle force measurements is rare.

Purpose of the Study:

• To test the accuracy of Hill-type muscle models against measured rat medial gastrocnemius muscle forces.
• To evaluate the impact of subject-specific versus population-averaged muscle properties.

Main Methods:

• Collected in-situ and in-vivo isometric, isotonic, and cyclic contraction data from rat medial gastrocnemius muscles.
• Developed Hill-type muscle models using subject-specific and population-averaged force-length and force-velocity parameters.
• Compared model predictions to measured muscle-tendon forces.

Main Results:

• Modeled forces for cyclic contractions showed good agreement with measured forces (r² 0.70–0.86).
• Model accuracy decreased at higher movement and cycle frequencies.
• Subject-specific properties reduced root-mean square error (RMSE) but did not improve r².
• Regional muscle models did not enhance performance over whole muscle models.

Conclusions:

• Hill-type muscle models provide reliable predictions for locomotion-like cyclic contractions.
• Subject-specific parameters offer marginal benefits by reducing prediction error magnitude.
• Anatomical region-specific modeling does not improve overall predictive accuracy.