Muscle synergies for predicting non-isometric complex hand function for commanding FES neuroprosthetic hand systems

Area of Science:

• Biomedical Engineering
• Neuroscience
• Rehabilitation Technology

Background:

• Myoelectric neuroprostheses aim to restore hand function for individuals with paralysis using voluntary electromyographic (EMG) signals.
• Limited EMG signal availability restricts the functionality achievable with current neuroprosthetic systems.

Purpose of the Study:

• To investigate the feasibility of reducing dynamic hand function to three degrees of freedom using a time-invariant muscle synergy model.
• To demonstrate the potential for synergy-based control in neuroprosthetic applications.

Main Methods:

• Extracted time-invariant muscle synergies from electromyographic (EMG) data of able-bodied participants performing static and dynamic tasks.
• Employed non-negative matrix factorization for synergy extraction and task cross-validation.

Main Results:

• Three functional synergies significantly outperformed chance (p < 0.01).
• These synergies accounted for 66.0% ± 4.9% of the variance, compared to 42.5% ± 4.4% for chance levels.

Conclusions:

• The study demonstrates the feasibility of a synergy-based controller for hand neuroprostheses.
• This approach shows potential for enhancing restored hand function by simplifying control signals.