FMG Versus EMG: A Comparison of Usability for Real-Time Pattern Recognition Based Control

Area of Science:

• Biomedical Engineering
• Human-Computer Interaction
• Rehabilitation Engineering

Background:

• Force myography (FMG) measures muscle surface pressure, offering a potential alternative to electromyography (EMG) for human-machine interfaces (HMIs).
• While FMG shows promise in offline accuracy, its real-time control usability compared to EMG remains less explored.

Purpose of the Study:

• To comprehensively compare the online usability of FMG- and EMG-based control schemes.
• To evaluate both classification and regression controllers for FMG and EMG in a virtual target acquisition task.

Main Methods:

• Twenty participants performed a two-degree-of-freedom Fitts' Law-style task using FMG and EMG with classification and regression controllers.
• Performance was assessed using standard metrics including throughput, path efficiency, and average speed.

Main Results:

• FMG-based classification significantly outperformed EMG-based classification in throughput and path efficiency.
• FMG-based regression also significantly outperformed EMG-based regression in throughput and path efficiency.
• FMG demonstrated superior performance over EMG across both control types.

Conclusions:

• Force myography (FMG) is a viable and superior alternative to electromyography (EMG) for real-time human-machine interfaces.
• The findings support the broader adoption of FMG in HMI applications due to its enhanced performance.