A high-performance SSVEP-BCI system based on high-frequency flickers in the peripheral visual field.

Area of Science:

• Neuroscience
• Biomedical Engineering
• Human-Computer Interaction

Background:

• Existing steady-state visual evoked potential (SSVEP)-based brain-computer interfaces (BCIs) often use central visual field flickers (8-20Hz), leading to noticeable flicker perception.
• This limitation impacts user experience and system usability.

Purpose of the Study:

• To develop a high-performance, low-flicker-perception SSVEP-based BCI system.
• To leverage high-density electrodes and high-frequency peripheral visual stimulation.

Main Methods:

• Utilized a custom EEG cap with high-density electrodes for enhanced data acquisition.
• Implemented high-frequency (32.00-36.68Hz) peripheral visual stimulation with annuli (2.1°-4.1° angular range) to encode 40 targets.
• Applied task-discriminant component analysis (TDCA) for signal decoding.

Main Results:

• Verified system feasibility through online experiments.
• Achieved an average classification accuracy of 83.22 ± 11.95%.
• Attained a high information transfer rate (ITR) of 178.21 ± 43.84 bits/min, the highest for peripheral visual field SSVEP-BCIs.
• Demonstrated that high-density electrodes improve EEG data quality and classification accuracy.

Conclusions:

• The proposed system offers a novel approach for designing SSVEP-BCIs with reduced flicker perception.
• Provides valuable insights for the future application of high-density electrodes in SSVEP-BCIs.