Attending to Visual Stimuli versus Performing Visual Imagery as a Control Strategy for EEG-based Brain-Computer...

Area of Science:

• Neuroscience
• Biomedical Engineering
• Human-Computer Interaction

Background:

• Brain-computer interfaces (BCIs) commonly utilize motor imagery.
• Exploring alternative mental tasks like visual imagery can expand BCI applications.

Purpose of the Study:

• To investigate the feasibility of using visual imagery for EEG-based BCIs.
• To differentiate between visual observation and visual imagery tasks using EEG data.
• To determine if specific visual stimuli can be identified during these tasks.

Main Methods:

• 26 participants performed tasks involving observing and imagining visual stimuli (flower or hammer).
• Electroencephalography (EEG) was recorded throughout the trials.
• Machine learning classifiers were used to analyze EEG data for different conditions.

Main Results:

• Classification accuracy between visual imagery and visual observation was 71%.
• Accuracy in distinguishing between different visual observation stimuli was 61%.
• Distinguishing between tasks and resting state yielded accuracies of 77% (imagery) and 75% (observation).

Conclusions:

• Visual imagery generates distinct EEG patterns, particularly alpha power changes.
• Visual imagery is a viable alternative to motor imagery for BCI control.
• These findings broaden the potential control strategies for BCIs.