An EEG-based Imagined Speech Database for comparing Paradigm Designs.

Area of Science:

• Neuroscience
• Human-Computer Interaction
• Cognitive Science

Background:

• Brain-computer interfaces (BCIs) are advancing computationally but often overlook human factors like fatigue and attention.
• These factors significantly influence brain signal modulation, impacting BCI performance.
• Existing BCI paradigms may not adequately account for user engagement and cognitive states.

Purpose of the Study:

• To propose and evaluate a video game-based paradigm for imagined speech BCIs.
• To enhance user engagement, reduce fatigue, and increase attention within BCI tasks.
• To investigate the impact of human-centered design on imagined speech signal quality.

Main Methods:

• Recruited 15 volunteers to compare two imagined speech paradigms: a traditional abstract scenario and a novel video game paradigm.
• Collected and analyzed imagined speech signals evoked under both conditions.
• Focused on the quality of brain signals modulated by user engagement and cognitive load.

Main Results:

• The study provides a dataset comparing imagined speech signals from a traditional BCI paradigm versus a human-factors-informed video game paradigm.
• Differences in signal quality between the two paradigms were investigated, highlighting the impact of engagement.
• Preliminary findings suggest potential improvements in signal quality with the video game approach.

Conclusions:

• Human factors, such as engagement and attention, are critical for optimizing brain-computer interface performance.
• Video game-based paradigms show promise for improving user experience and data quality in imagined speech BCIs.
• The dataset can inform the development of more effective BCI decoding models and personalized BCI designs.