An Adaptive Task-Related Component Analysis Method for SSVEP Recognition

Abstract

Steady-State Visual Evoked Potential (SSVEP) recognition methods use a subject's calibration data to differentiate between brain responses, hence, providing the SSVEP-based brain-computer interfaces (BCIs) with high performance. However, they require sufficient calibration EEG trials to achieve that. This study develops a new method to learn from limited calibration EEG trials, and it proposes and evaluates a novel adaptive data-driven spatial filtering approach for enhancing SSVEP detection. The spatial filter learned from each stimulus utilizes temporal information from the corresponding EEG trials. To introduce the temporal information into the overall procedure, a multitask learning approach, based on the Bayesian framework, is adopted. The performance of the proposed method was evaluated into two publicly available benchmark datasets, and the results demonstrated that our method outperformed competing methods by a significant margin.

Keywords: EEG; brain–computer interfaces; multitask learning; spatial filtering; steady-state visual evoked potentials; task-related component analysis.

Figure 1

Average Classification over all subjects (a) for the Speller dataset and (b) the EPOC dataset 14 using the basic configuration with respect to the EEG channels. In both cases, the time window ranges from 0.5 s to 4 s (0.5 s interval). * indicates statistically significant difference between the TRCA and adTRCA methods, using paired sample t-test for Speller dataset and Wilcoxon signed rank test for the EPOC dataset ($p<0.05$).

Figure 2

Average Classification over all subjects (a) for the Speller dataset and (b) for the EPOC dataset using the EEG channels covering the occipital areas. In both cases, the time window ranges from 0.5 s to 4 s (0.5 s interval). * indicates statistically significant difference between the TRCA and adTRCA methods, using paired sample t-test for the Speller dataset and Wilcoxon signed rank test for the EPOC dataset ($p<0.05$).

Figure 3

Average Classification over all subjects by using for the Speller dataset with (a) 9 channels and (b) 3 channels and for the EPOC dataset with (c) 14 channels and (d) 2 channels, respectively. In both cases, the time window ranges from 0.5 s to 4 s (0.5 s interval). * indicates statistically significant difference between the two methods using paired sample t-test for the Speller dataset and Wilcoxon signed rank test for the EPOC dataset ($p<0.05$).

Figure 4

Average Classification in Hairless case for the Speller dataset. The time window ranges from 0.5 s to 4 s (0.5 s interval). * indicates statistically significant difference between the two methods using paired sample t-test ($p<0.05$).