Enhancing detection of SSVEP-based surround suppression using canonical correlation analysis
Sicong Chen, Wen Hu, Xin Zhang, Qiyun Huang · 2023
Canonical correlation analysis (CCA), a method for assessing inter-signal correlation that is widely used to enhance the performance of brain-computer interfaces (BCIs), can play an effective role in detecting steady-state visual evoked potentials (SSVEPs) in response to target flicker stimuli. However, it remains unclear whether CCA-based measurements can enhance the quantification of a visual perceptual process that related to human psychophysics. In this study, we used SSVEP to achieve quantitative measurements of the surround suppression effect, a visual perceptual process related to the excitation-inhibition balance in the central nervous system based on electroencephalogram (EEG). The CCA method was compared with two traditional methods for SSVEP quantifications: the power spectral density (PSD) and the signal-to-noise ratio (SNR). Our results demonstrated that CCA had the highest sensitivity among the three to enhance the quantification of SSVEP-based surround suppression effect. Thus, our work confirmed that CCA is an outstanding method for SSVEP measurements, compared to the two other commonly used SSVEP detection methods mentioned below, not only in the field of artificial intelligence, such as the development of BCIs, but also can facilitate vision studies in cognitive and perceptual neuroscience.