Online Compensation of Systematic Effects in Stimuli Generation for XR-Based SSVEP BCIs

Area of Science:

• Neuroscience
• Human-Computer Interaction
• Wearable Technology

Background:

• Brain-Computer Interfaces (BCIs) utilizing Steady-State Visually Evoked Potentials (SSVEPs) show promise for wearable applications.
• Extended Reality (XR) environments present challenges to BCI classification due to stimulus presentation inconsistencies, specifically display refresh rate variations.

Purpose of the Study:

• To introduce and evaluate a novel online compensation method for XR display refresh rate (RR) variations.
• To enhance the classification accuracy of SSVEP-based BCIs in XR environments without additional training or invasive procedures.

Main Methods:

• Developed a non-invasive monitoring module to detect and compensate for frame rate deviations in XR displays.
• Evaluated classification performance using Filter Bank Canonical Correlation Analysis (FBCCA).
• Utilized two datasets comprising 9 and 30 subjects, collected using Moverio BT-350 and HoloLens 2.

Main Results:

• The compensation method significantly improved SSVEP classification accuracy, with improvements correlating to the magnitude of frame per second (fps) deviations.
• Classification accuracy increased by up to 300% in some cases.
• Statistical analyses confirmed the method's reliability across subjects and datasets.

Conclusions:

• The proposed method effectively enhances SSVEP-based BCIs operating in XR environments.
• This approach provides a robust foundation for reliable and practical XR BCI applications.