Investigation of multiple frequency recognition from single-channel steady-state visual evoked potential for efficient brain–computer interfaces application

摘要: In this study, the authors have examined a single-channel electroencephalogram from O z for identification of seven visual stimuli frequencies with multivariate synchronisation index (MSI) and canonical correlation analysis (CCA). Authors investigated feasibility in three case studies varying overlapped as well non-overlapped window lengths. The ≤10 Hz are considered study I >10 Hz II. Case III contains both All revealed that CCA outperforms MSI reference signals constituting fundamental, one subharmonics, super-harmonics. results accuracy improves 50% overlap algorithms. Further, recognition is studied combination sub- super-harmonics overlap. perform better fundamental twice frequency compared traditional power spectral density (PSDA). addition to accuracy, information bit transfer rate also higher relative PSDA.