Continuous Versus Discrete Simultaneous Control of Prosthetic Fingers

Area of Science:

• Biomedical Engineering
• Neuroscience
• Robotics

Background:

• Modern prosthetic hands offer individual digit control, but this is underutilized due to challenges in estimating finger motion from surface electromyography (EMG).
• Existing regression methods for closed-loop control (position, velocity, force) have limitations.
• A need exists for robust and effective control schemes for dexterous prosthetic hands.

Purpose of the Study:

• To propose and evaluate an alternative open-loop, action-based control scheme for prosthetic hands using simultaneous finger motion classification.
• To compare the efficacy of continuous closed-loop versus discrete open-loop control for operating a robotic hand with five degrees of actuation (DOAs).

Main Methods:

• Eight normally-limbed subjects teleoperated a dexterous robotic hand using a data glove.
• Subjects utilized both continuous closed-loop and discrete open-loop control schemes.
• Performance was assessed by matching target hand postures presented on a screen.

Main Results:

• The performance of continuous closed-loop and discrete open-loop control methods was found to be comparable.
• Subject experience significantly improved performance over time, irrespective of the control scheme used.
• Myoelectric classification demonstrated potential for discrete, action-based control in prosthetic finger actuation.

Conclusions:

• Prosthetic finger control in a continuous space may be achievable through myoelectric classification and discrete, action-based control.
• Further research is encouraged to explore the potential of this open-loop control approach.
• The findings support the development of more intuitive and effective prosthetic hand control systems.