Evaluation of command algorithms for control of upper-extremity neural prostheses

Area of Science:

• Biomedical Engineering
• Neuroscience
• Rehabilitation Technology

Background:

• Upper-extremity neural prostheses aim to restore grasping function.
• Current control algorithms for neural prostheses have limitations in grasp force regulation.
• Advanced control strategies are needed to improve prosthesis dexterity and user control.

Purpose of the Study:

• To develop and evaluate novel command control algorithms for enhanced grasp force control in upper-extremity neural prostheses.
• To compare the performance of new algorithms against the existing proportional control with lock algorithm.
• To assess the efficacy of a video-simulation tool for initial evaluation of neural prosthesis control algorithms.

Main Methods:

• Five new command control algorithms were designed, focusing on grasp force modulation.
• Able-bodied subjects used a shoulder controller with a video-simulated hand to perform standardized tasks.
• Data analysis involved a generalized estimating equations-based linear model to compare algorithm performance.
• Algorithms were ranked using contrast analyses of linear model coefficients.

Main Results:

• Algorithms enabling post-lock command adjustment or decreased controller gain after locking outperformed the standard algorithm.
• Algorithms that modified command based on time performed worse than the proportional control with lock method.
• The video-based computer simulator demonstrated utility as an effective preliminary evaluation tool.

Conclusions:

• Novel algorithms offer improved grasp force control for neural prostheses compared to current methods.
• Specific algorithmic features, such as post-lock adjustment and gain decrease, are beneficial for performance.
• Video simulation is a valuable tool for the initial assessment of neural prosthesis control strategies.