Enhanced Brain-Controlled Mobile Robot Based on SE-VEP Paradigm With Single Stimulus

Area of Science:

• Neuroscience
• Biomedical Engineering
• Human-Computer Interaction

Background:

• Steady-state visually evoked potentials (SSVEPs) are efficient for brain-computer interfaces (BCIs) but cause visual fatigue and stimulus interference.
• Limitations of traditional SSVEP methods necessitate innovative BCI paradigms for improved user experience and performance.

Purpose of the Study:

• To introduce and validate a novel spatial encoding-visually evoked potential (SE-VEP) BCI paradigm.
• To address the limitations of traditional SSVEP BCIs, specifically visual fatigue and stimulus interference.
• To optimize target point placement for efficient electroencephalogram (EEG) encoding within a single stimulus block.

Main Methods:

• Developed a SE-VEP model using four optimized target points for gaze restriction around a stimulus block.
• Employed a Riemann kernel-based support vector machine (R-SVM) for classifying electroencephalogram (EEG) data with varying eccentricities.
• Validated the paradigm's feasibility through an online brain-controlled robotic virtual system and evaluated user fatigue and information transfer rate (ITR).

Main Results:

• Achieved a classification accuracy of up to 86.11% using the R-SVM approach.
• Determined an optimal time window length of 1.2 s for online BCIs based on ITR evaluation.
• Demonstrated significant reductions in user fatigue (2.8 ± 0.5 vs. 4.1 ± 0.6) and improved stimulus block utilization (24.6 ± 2.3 vs. 8.2 ± 1.1 bits/min) compared to traditional BCI systems.

Conclusions:

• The proposed SE-VEP paradigm is effective for online BCI control, offering a viable alternative to traditional SSVEP systems.
• The SE-VEP model successfully mitigates user fatigue and enhances the efficiency of BCI applications.
• Optimized spatial encoding and classification methods contribute to improved BCI performance and user experience.