Spatial decoupling of targets and flashing stimuli for visual brain-computer interfaces

Area of Science:

• Neuroscience
• Biomedical Engineering
• Human-Computer Interaction

Background:

• Code-modulated visual evoked potentials (c-VEPs) offer high information transfer rates for noninvasive brain-computer interfaces (BCIs).
• Current c-VEP paradigms necessitate direct foveal fixation, causing user discomfort and fatigue, limiting practical BCI application.
• A need exists for more ergonomic and user-friendly c-VEP BCI designs.

Purpose of the Study:

• To present a novel c-VEP BCI paradigm that spatially decouples visual stimuli from target locations.
• To investigate the feasibility of non-foveal target classification using spatially separated stimuli.
• To enhance the number of selectable targets without compromising BCI performance.

Main Methods:

• Developed a c-VEP paradigm where flashing stimuli form a ring around non-flashing targets.
• Utilized non-foveal visual field stimuli to induce EEG changes for target classification.
• Incorporated boundary positioning for a subset of targets to decouple them from single stimuli.

Main Results:

• Achieved practical classification accuracies in the non-foveal condition, comparable to direct-foveal methods at longer observation times.
• Online testing with 5 subjects yielded an average accuracy of 95.6% for a 4-target condition.
• Offline analysis demonstrated that boundary-positioned targets performed as accurately as traditional superimposed targets.

Conclusions:

• c-VEPs can be effectively classified for BCI applications without requiring direct foveation, reducing user irritation.
• The number of targets can be increased beyond the number of stimuli without performance degradation.
• This research paves the way for more practical and ergonomic BCIs with high information transfer rates.