SKDCPM algorithm can improve the single-trial decoding performance of very similar error-related potentials

Area of Science:

• Neuroscience
• Biomedical Engineering
• Signal Processing

Background:

• Error related potentials (ErrPs) are crucial brain signals for brain-computer interfaces (BCIs).
• Existing ErrP decoding methods struggle to differentiate subtle variations in error types.
• Accurate decoding of detailed error information is vital for enhancing BCI functionality and user feedback.

Purpose of the Study:

• To develop a novel algorithm for decoding very similar ErrPs.
• To compare the performance of the new algorithm against established ErrP decoding methods.
• To assess the capability of distinguishing different degrees of errors from electroencephalography (EEG) signals.

Main Methods:

• Proposed a new algorithm: shrinkage discriminant canonical pattern matching (SKDCPM).
• Collected EEG data from 18 subjects under four conditions: correct (0°) and errors (45°, 90°, 180° deviations).
• Compared SKDCPM with Linear Discriminant Analysis (LDA), Shrinkage LDA (SKLDA), Stepwise LDA (SWLDA), Bayesian LDA (BLDA), and Discriminant Canonical Pattern Matching (DCPM).

Main Results:

• SKDCPM demonstrated high balanced accuracy (BACC) in distinguishing correct from incorrect responses (0° vs. others).
• SKDCPM achieved a significant grand averaged BACC of 69.54% (up to 74.25%) in decoding very similar ErrPs (45° vs. 90° vs. 180°).
• The proposed SKDCPM significantly outperformed all other tested algorithms in differentiating subtle error variations.

Conclusions:

• The novel SKDCPM algorithm offers a significant advancement in decoding subtle ErrPs.
• This method enhances the potential for providing more granular feedback in ErrP-based BCI systems.
• SKDCPM provides a promising new decoding approach for developing more sophisticated BCI applications.