Comparison of Point-and-Click Performance Between the Brainfingers BCI and the Mouse

Area of Science:

• Human-Computer Interaction
• Biomedical Engineering
• Rehabilitation Technology

Background:

• Brain-computer interfaces (BCIs) offer alternative control methods for computing.
• Non-invasive BCIs utilize biosignals like EMG, EOG, and EEG.
• Evaluating BCI performance against conventional input devices is crucial for assessing usability.

Purpose of the Study:

• To quantitatively compare the performance of a hybrid BCI (Brainfingers) with a conventional mouse.
• To assess device efficiency using Fitts' law metrics in 1D and 2D pointing tasks.
• To explore the potential assistive technology value of non-invasive BCIs.

Main Methods:

• A commercial wearable BCI (EMG, EOG, EEG) was tested against a standard optical mouse.
• Participants performed 1D and 2D target acquisition tasks following ISO/TS 9241-411 standards.
• Pointer movement data was analyzed using Fitts' law to calculate throughput.

Main Results:

• The BCI device demonstrated substantially lower performance than the mouse.
• Mean throughput for the BCI was 0.35 bits/s (1D) and 0.43 bits/s (2D), versus 6.03 bits/s (1D) and 5.17 bits/s (2D) for the mouse.
• Significant performance differences were observed between the BCI and mouse.

Conclusions:

• Despite lower performance in non-disabled users, the evaluated non-invasive BCI may hold assistive value.
• Further research could explore BCI applications for individuals with severe motor disabilities.
• The study highlights the performance gap between current non-invasive BCIs and conventional input devices.