Fingertip Proximity-Based Grasping Pattern Prediction of Transradial Myoelectric Prosthesis

Area of Science:

• Biomedical Engineering
• Rehabilitation Technology
• Prosthetics

Background:

• Transradial amputees often struggle with slow and inaccurate grasp pattern selection for multigrasp prostheses, especially with limited muscle activity.
• Existing electromyography (EMG)-based methods can be insufficient for intuitive and rapid control.

Purpose of the Study:

• To develop and evaluate a novel fingertip proximity sensing system for automatic grasping pattern prediction in transradial prostheses.
• To improve the speed and efficiency of prosthetic hand control for amputees.

Main Methods:

• A fingertip proximity sensor was integrated to create a dataset of five common grasp patterns (spherical, cylindrical, tripod pinch, lateral pinch, hook).
• A neural network classifier achieved 96% accuracy on the proximity training dataset.
• A combined EMG/proximity-based system (PS-EMG) was tested on non-disabled subjects and a transradial amputee performing reach-and-pick-up tasks.

Main Results:

• Non-disabled subjects using the PS-EMG method initiated grasping 1.93 seconds faster and completed tasks 7.30% quicker than with pure EMG methods.
• The transradial amputee subject demonstrated a 25.58% improvement in task completion time with the PS-EMG method compared to switch-based EMG.
• The proposed system effectively reduced reliance on EMG signals for grasp pattern recognition.

Conclusions:

• Fingertip proximity sensing offers a promising approach for rapid and intuitive grasp pattern selection in multigrasp prostheses.
• The PS-EMG method enhances prosthetic control efficiency and reduces the cognitive load on users.
• This technology has the potential to significantly improve the functional capabilities of transradial amputees.