Enhanced Spatial Division Multiple Access BCI Performance via Incorporating MEG With EEG.

Area of Science:

• Neuroscience
• Biomedical Engineering
• Signal Processing

Background:

• Spatial division multiple access (SDMA) in Brain-Computer Interfaces (BCI) is limited by the poor spatial resolution of Electroencephalography (EEG).
• Magnetoencephalography (MEG) offers better spatial resolution, but its integration with EEG for SDMA-BCI has not been fully explored.

Purpose of the Study:

• To develop and evaluate a 16-command SDMA-encoded BCI system using a fusion of MEG and EEG signals.
• To investigate the spatiotemporal characteristics of MEG and EEG signals in the occipital region.
• To compare the performance of EEG-only, MEG-only, and MEG-EEG fusion modalities.

Main Methods:

• Synchronous acquisition of MEG and EEG signals from 10 subjects during a visual stimulus task.
• Analysis of spatiotemporal features in the occipital region.
• Fusion of MEG and EEG modalities without prior signal processing.
• Evaluation using the multi-class discriminative canonical pattern matching (Multi-DCPM) algorithm.

Main Results:

• MEG demonstrated superior offline classification accuracy (27.81% improvement over EEG) for 16 fixation points with a 4s data length.
• The MEG-EEG fusion modality achieved an average offline accuracy of 91.71%, significantly outperforming MEG (p<0.01, ANOVA).
• The fusion modality reached an average information transfer rate (ITR) of 60.74 bits/min with a 1s data length, a 14% increase over MEG.

Conclusions:

• MEG-EEG fusion significantly enhances spatial features and overall performance of SDMA-encoded BCIs.
• The developed fusion modality BCI system shows high potential and feasibility for practical applications.
• This study provides valuable insights for advancing SDMA applications in BCI technology.