Key result
IF-VMD-EMD method outperforms relevant alternatives for single-channel SSVEP datasets with ~91% average accuracy.
Why the study?
Multichannel setups for SSVEP-based detection are complex and cause discomfort during prolonged use, creating a need for accurate single-channel detection methods.
The proposed IF-VMD-EMD method improves accuracy and information transfer rate for single-channel SSVEP-based brain-computer interface systems.
May enhance single-channel SSVEP BCI accuracy; leaves open prospective clinical validation.
• An accurate and user-friendly method for SSVEP-based BCI classification using a single channel. • The window length of the moving average filter is optimally selected based on maximum SNR. • Variational mode decomposition is used to decompose the SSVEP signal into harmonics and noisy components. • Introduces impulse factor of cross-correlation function (IFCCF) method for optimally selecting the SSVEP harmonic components. • Empirical mode decomposition method is used to further enhance the SNR of SSVEP signal. • The proposed method represents improved accuracy and information transfer rate. The steady-state visual evoked potential (SSVEP) is a brain response to visual stimuli, flickering at the target frequency and its harmonics, in the occipital cortex. SSVEP-based detection methods use multichannel or single channel setups. Multichannel setups are complex and can cause discomfort to the user, especially disabled individuals, during prolonged use. So, there is a need to develop accurate single channel SSVEP based detection method. In this study, we developed a novel method called impulse factor-based Variational Mode Decomposition, Empirical Mode Decomposition (IF-VMD-EMD) for single-channel SSVEP datasets. For preprocessing of recorded SSVEP signals, the Moving Average Filter (MAF) was applied to improve the Signal-to-Noise Ratio (SNR). The VMD then decomposes the SSVEP signal into harmonic Variational Mode Functions (VMFs) and noisy VMFs. Moreover, we introduced the Impulse Factor of Cross Correlation Function (IFCCF) method to optimally select harmonic VMFs. The addition of EMD with IF-VMD further enhances the SNR by selecting the contributing EMD-Intrinsic Mode Function based on the Pearson correlation coefficient, aiming to remove noise from the restored signal and extract sub-signals with rich parent signal information. The proposed method (IF-VMD-EMD) performed better than other relevant methods and achieved an average accuracy of 92.32 %, 89.62 %, 90.01 % and an Information Transfer Rate (ITR) of 64.80 bits/min, 77.56 bits/min, 63.11 bits/min for three publicly available datasets consisting of seven, twelve and seven classes, respectively. The proposed method (IF-VMD-EMD) exhibits better accuracy, improved ITR, and is user-friendly in real world applications.
No takes yet. Share an insight, caveat, or question.
Wahab et al. (2025) studied SSVEP-based BCI classification. IF-VMD-EMD method vs. Other relevant methods was evaluated on Average accuracy and Information Transfer Rate (ITR). The proposed IF-VMD-EMD method for single-channel SSVEP datasets achieved average accuracies of 92.32%, 89.62%, and 90.01% across three public datasets, outperforming other relevant methods.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: