Feedback parameters for a closed-loop multiple-input multiple-output model of the upper limb.

Area of Science:

• Neuroscience
• Biomechanics
• Systems Biology

Background:

• Closed-loop and multi-input multi-output (MIMO) models are essential for simulating upper limb motor control.
• Existing models lack comprehensive, large-scale linear neuromusculoskeletal representations that are both closed-loop and MIMO.
• A significant barrier to developing such models is the absence of established feedback parameter values (gains and delays).

Purpose of the Study:

• To introduce a methodology for constructing MIMO models of short-loop afferent feedback in the upper limb.
• To provide estimated average values and ranges for feedback parameters based on existing literature.

Main Methods:

• Integrated feedback parameter data from 26 previous studies.
• Applied principles of system stability and behavior to refine parameter estimates.
• Developed a linear model incorporating 13 superficial muscles and 7 joint degrees of freedom (shoulder to wrist).
• Included homonymous feedback (Golgi tendon organs) and homonymous/heteronymous feedback (muscle spindles).

Main Results:

• Estimated feedback gains and delays for the upper limb neuromusculoskeletal model.
• Validated muscle spindle feedback gains by comparing signs with known central delay differences.
• Validated delay times by comparing estimated delays with measured innervation lengths, showing strong fit for efferent (R=0.88) and moderate fit for afferent (R=0.65) delays.
• Demonstrated the impact of feedback on model behavior and compared it to experimental observations.

Conclusions:

• Successfully developed a method for creating MIMO neuromusculoskeletal models with estimated feedback parameters.
• The model provides a foundation for simulating upper limb motor control with realistic feedback mechanisms.
• Findings contribute to a better understanding of how neural feedback influences upper limb biomechanics and movement.