Computationally efficient modeling of proprioceptive signals in the upper limb for prostheses: a simulation study

Area of Science:

• Biomechanics
• Neuroscience
• Prosthetics Engineering

Background:

• Accurate modeling of proprioceptive neural patterns is crucial for developing intuitive prosthetic limbs for amputees.
• Existing neuro-musculoskeletal models are often computationally intensive, limiting their application in portable prosthetic devices.

Purpose of the Study:

• To combine efficient biomechanical and proprioceptor models for generating human-like muscular proprioceptive signals.
• To enable real-time signal generation for experimental prosthesis feedback systems.

Main Methods:

• Development of a 7-degree-of-freedom, 17-muscle upper limb neuro-musculoskeletal model.
• Integration of an inverse dynamics tool (static optimization) to estimate unknown muscle activation and excitation levels.
• Implementation of approximations and optimizations for real-time operation on portable hardware.

Main Results:

• The model successfully generates real-time estimates of muscle spindle and Golgi Tendon Organ neural firing patterns.
• Proposed optimizations make the computationally demanding modeling feasible for portable prosthetic applications.

Conclusions:

• The developed model and optimizations represent a significant step towards creating intuitive proprioceptive neural prostheses.
• Identified technical obstacles and model limitations provide a roadmap for future research in advanced prosthetic feedback.