Relation of canonical correlation analysis and multivariate synchronization index in SSVEP detection. (March 2022)

• Record Type: Journal Article
• Authors: Tong, Changkai; Wang, Huali; Wang, Yida
• Is Part Of: Biomedical signal processing and control. Volume 73(2022)
• Publication Date: 2022-03
• Subjects: Canonical correlation analysis -- Multivariate synchronous index -- Steady-state visual evoked potentials -- Taylor expansion; Signal processing -- Periodicals; Biomedical engineering -- Periodicals; Signal Processing, Computer-Assisted -- Periodicals; Image Processing, Computer-Assisted -- Periodicals; Biomedical Engineering -- Periodicals; 610.28
• DOI: 10.1016/j.bspc.2021.103345
• Languages: English
• ISSNs: 1746-8094

Abstract

Abstract: Canonical correlation analysis (CCA) and multivariate synchronization index (MSI) are two most extensively signal processing algorithms to detect steady-state visual evoked potentials (SSVEP) in brain-computer interfaces (BCI). However, there is no standard principle to choose the proper algorithm in a practical system. Conventionally, algorithm selection is determined by experimental comparisons. A drawback of this method is that the corresponding reference significance reduces sharply as the experimental conditions are changed. In this paper, the conversion relationship between the above algorithms is verified through a rigorous mathematical derivation, which also provides the basis for future signal processing algorithm selection. In addition, almost all the approaches designed to enhance CCA can directly transfer to MSI by the proposed method. Finally, the conclusions are validated with a benchmark data set.