Improvement in Classification of Tactile Event-Related Potentials using Random-Interval Tasks^∗

Area of Science:

• Neuroscience
• Biomedical Engineering
• Human-Computer Interaction

Background:

• Brain-computer interfaces (BCIs) offer alternative communication and control methods.
• Tactile stimuli present a potential modality for BCI input, particularly for visually impaired individuals using Braille-like displays.
• Investigating event-related potentials (ERPs) elicited by tactile stimuli is crucial for BCI development.

Purpose of the Study:

• To assess the classification accuracy of tactile-stimulus-induced event-related potentials (ERPs) for brain-computer interface (BCI) applications.
• To compare the efficacy of different tactile stimulus patterns in a two-class classification task.
• To determine the influence of attentional focus on BCI performance with tactile stimuli.

Main Methods:

• Mechanical tactile stimuli were delivered to index fingers via piezo actuators, mimicking Braille display.
• A two-class classification task was implemented using three tactile stimulus patterns.
• Subjects focused on infrequent target stimuli amidst frequent non-target stimuli.
• Stepwise linear discriminant analysis was employed for feature classification.

Main Results:

• High classification accuracy was achieved using random interval tactile stimuli compared to constant intervals.
• The task involving random intervals demonstrated superior performance.
• Performance differences suggest a link between task concentration and BCI accuracy.

Conclusions:

• Tactile stimulus-based BCIs show promise, particularly for the visually impaired.
• Attentional concentration significantly influences the accuracy of tactile BCI systems.
• Random interval stimulus presentation may enhance BCI performance by engaging user focus.