Latency Aligning Task-Related Component Analysis Using Wave Propagation for Enhancing SSVEP-Based BCIs

Area of Science:

• Neuroscience
• Biomedical Engineering
• Signal Processing

Background:

• Steady-state visual evoked potential (SSVEP) is crucial for high-speed brain-computer interfaces (BCIs) due to its artifact robustness.
• Task-related component analysis (TRCA) is an effective spatial filtering method for SSVEP-based BCIs.
• Existing TRCA methods can be further optimized for improved performance.

Purpose of the Study:

• To introduce a novel method, Latency Aligning TRCA (LA-TRCA), for enhancing SSVEP detection.
• To improve the accuracy of phase information extraction in SSVEP signals.
• To validate the effectiveness of LA-TRCA compared to existing TRCA-based methods.

Main Methods:

• Proposed Latency Aligning TRCA (LA-TRCA) by incorporating visual latency alignment.
• Estimated phase velocity using channel phase shifts to determine visual latencies.
• Applied TRCA to inter-channel aligned data epochs for target recognition.
• Validated LA-TRCA against other TRCA expansions on two SSVEP datasets.

Main Results:

• LA-TRCA demonstrated superior classification performance compared to other TRCA-based expansions.
• The proposed method effectively aligns visual latencies across channels.
• Accurate phase information was obtained from task-related SSVEP signals.

Conclusions:

• Latency Aligning TRCA (LA-TRCA) is an effective enhancement for SSVEP detection.
• Aligning visual latencies significantly improves the performance of TRCA in SSVEP-based BCIs.
• The findings support the practical application of LA-TRCA for advanced brain-computer interface development.