A physiologically inspired meta-pattern generator bridging intention and motor primitives for human voluntary...

Area of Science:

• Neuroscience
• Biomedical Engineering
• Computational Modeling

Background:

• Understanding neural signal transmission from motor intention to lower limb joint motion is crucial for developing realistic motion generation models.
• Existing models often fail to bridge spinal integration outputs with joint motions or bypass spinal processing altogether, lacking biological relevance.
• This gap hinders advancements in neuroprosthetics and rehabilitation robotics.

Purpose of the Study:

• To propose a physiologically inspired computational model that accurately represents the hierarchical transmission of neural signals for lower limb locomotion.
• To address the challenge of decoupling spinal integration outputs from joint motions.
• To provide a biologically plausible foundation for intention-driven motion generation in prostheses and robots.

Main Methods:

• A dual-layer Meta-Pattern Generator (MPG) was developed to model spinal integration, processing motor intention and joint feedback.
• The MPG generates meta-pattern signals that map one-to-one with joint-specific motor primitives.
• Experimental validation involved eight subjects across five locomotion modes, analyzing electromyographic (EMG) signals and joint angles.

Main Results:

• A clear one-to-one correspondence was found between meta-patterns (from EMG) and motor primitives (from joint angles).
• The model accurately predicted joint angles, closely matching experimental data with an average error below 1.7%.
• The model successfully replicated the hierarchical neural signal transmission from intention to joint execution.

Conclusions:

• The proposed Meta-Pattern Generator (MPG) model effectively bridges the gap between spinal integration and joint motion in lower limb locomotion.
• This physiologically inspired approach offers significant insights into neuromusculoskeletal modeling.
• The model demonstrates potential for advancing the neural control of lower limb prostheses and rehabilitation robots.