Using speech for mode selection in control of multifunctional myoelectric prostheses

Area of Science:

• Biomedical Engineering
• Rehabilitation Technology
• Neuroprosthetics

Background:

• Electromyogram (EMG) signals from residual muscles are crucial for controlling motorized prostheses.
• High-level amputations limit residual muscle availability, hindering flexible control of multi-degree-of-freedom myoelectric prostheses.
• Current control methods like sequential mode-switching can be inefficient.

Purpose of the Study:

• To propose and evaluate a novel control strategy for multifunctional myoelectric prostheses.
• To enhance prosthetic arm control flexibility by integrating speech signals with EMG signals.
• To improve the operational efficiency of myoelectric prostheses, especially for individuals with limited residual muscle function.

Main Methods:

• A new control strategy was developed, utilizing speech signals for mode (joint) selection.
• Electromyogram (EMG) signals were then used to determine and execute motion classes within the selected joint.
• The strategy replaces traditional sequential mode-switching methods.

Main Results:

• Preliminary results from three able-bodied subjects and one transhumeral amputee showed high mode-selection accuracy.
• The proposed strategy demonstrated enhanced operational efficiency compared to traditional methods.
• The findings suggest a potential improvement in the control performance of commercial myoelectric prostheses.

Conclusions:

• Integrating speech signals with EMG offers a more flexible and efficient control solution for myoelectric prostheses.
• This strategy addresses the limitations of reduced muscle availability in high-level amputations.
• The approach shows promise for improving the usability and performance of advanced prosthetic devices.