Feasibility of Two Different EMG-Based Pattern Recognition Control Paradigms to Control a Robot After Stroke - Case...

Area of Science:

• Rehabilitation Engineering
• Neuroscience
• Biomedical Engineering

Background:

• Stroke frequently causes chronic upper-extremity motor impairment, limiting functional recovery.
• Current therapies, including robotics, show limited profound effects on post-stroke motor deficits.
• Wearable assistive devices require effective real-time control strategies for stroke patients.

Purpose of the Study:

• To evaluate the feasibility and efficacy of two electromyography (EMG)-based controllers for shoulder movement support in stroke survivors.
• To assess the ability of EMG signals to classify intended shoulder movements during functional tasks.

Main Methods:

• Trained a linear discriminant analysis classifier using time-domain and auto-regressive features from EMG data.
• Acquired EMG data during simulated limb weight loading (abduction/adduction).
• Tested EMG-based position and force control paradigms with a custom lab-based robot during lift and reach tasks.

Main Results:

• Participants successfully controlled the robot arm using both position-based and force-based EMG control paradigms.
• The system enabled participants to position the robot arm and perform reaching tasks within specified constraints.
• This case study demonstrates the feasibility of using EMG to classify intended shoulder movements in stroke survivors.

Conclusions:

• Electromyography-based control is a feasible approach for assisting shoulder movement in individuals post-stroke.
• Further research is needed to assess the impact of this technology on improving reaching function.
• Real-time EMG classification holds promise for developing advanced wearable shoulder support systems.