Multigesture Electromyographic Control Complexity in Upper Limb Prostheses Actuated via Single Sensor Input...

Area of Science:

• Biomedical Engineering
• Rehabilitation Technology
• Human-Computer Interaction

Background:

• High prosthetic rejection rates are linked to functional limitations.
• Electromyography (EMG) upper limb prostheses use residual limb muscle contractions for control.
• Expanding EMG prostheses to multigesture control presents challenges in accuracy, usability, training time, and cognitive load.

Purpose of the Study:

• To assess the feasibility of learning multiple EMG gestures in a single training session.
• To investigate the impact of increasing gesture complexity on user performance, usability, and cognitive workload.

Main Methods:

• Participants (n=54) with intact upper extremities were fitted with a surface EMG device (Flex Controller).
• A custom training app presented 1, 3, or 5 control zones, each corresponding to a specific prosthesis gesture.
• Outcomes measured included task performance, perceived usability (System Usability Scale), and cognitive workload (NASA-Task Load Index).

Main Results:

• Task performance significantly decreased as the number of control zones increased (P<.001).
• Perceived usability was rated lower by the 5-gesture cohort compared to a progressive training group (P=.05).
• Cognitive workload and perceived difficulty increased with a higher number of gestures.

Conclusions:

• Progressive training for three EMG gestures is feasible and recommended.
• A three-gesture progressive learning approach balances device functionality, user intention, perceived usability, and cognitive demands.
• This method optimizes the user experience for advanced EMG upper limb prostheses.