Muscle Synergy-Based Iterative Learning Control for Upper Limb Functional Electrical Stimulation in Stroke...

Area of Science:

• Neuroscience
• Rehabilitation Engineering
• Biomedical Engineering

Background:

• Functional Electrical Stimulation (FES) is crucial for stroke patient rehabilitation, enabling complex movements.
• Controlling multiple muscles with FES presents challenges in coordination and efficiency.
• Existing methods struggle with high-dimensional muscle control and natural motor strategies.

Purpose of the Study:

• To develop a novel FES control framework integrating muscle synergy theory, LSTM, and ILC.
• To reduce the dimensionality of muscle control for more natural movement patterns.
• To enhance the precision and efficiency of FES-driven upper limb movements in stroke rehabilitation.

Main Methods:

• Utilized muscle synergy theory to predict synergy activation coefficients via LSTM networks.
• Employed Iterative Learning Control (ILC) to refine electrical stimulation intensities based on tracking errors.
• Integrated LSTM and ILC for a synergy-based FES control framework.

Main Results:

• Significantly reduced joint angle tracking errors (RMSE) in healthy subjects over 10 iterations.
• Demonstrated substantial RMSE reduction in elbow and wrist movements during a drinking task.
• Achieved a coefficient of determination (R² > 0.96), indicating high tracking accuracy and stability.

Conclusions:

• The proposed synergy-based ILC method enhances FES control for upper limb movements.
• The framework shows potential for improving natural motor control strategies in stroke rehabilitation.
• Preliminary results suggest adaptability and clinical potential for stroke patients.