An automated and fast approach to detect single-trial visual evoked potentials with application to brain–computer interface

•A joint spatial–temporal–spectral filter combining common spatial pattern and wavelet filtering can significantly increase the signal-to-noise ratio of single-trial visual evoked potentials.•The proposed approach can obtain robust and reliable visual evoked potentials in an automated and fast manner, thus satisfying the requirements of practical brain–computer interface systems.•The proposed approach can be potentially used to achieve real-time and automated detection of single-trial evoked potentials or event-related potentials in various paradigms.

ObjectiveThis study aims (1) to develop an automated and fast approach for detecting visual evoked potentials (VEPs) in single trials and (2) to apply the single-trial VEP detection approach in designing a real-time and high-performance brain–computer interface (BCI) system.MethodsThe single-trial VEP detection approach uses common spatial pattern (CSP) as a spatial filter and wavelet filtering (WF) a temporal–spectral filter to jointly enhance the signal-to-noise ratio (SNR) of single-trial VEPs. The performance of the joint spatial–temporal–spectral filtering approach was assessed in a four-command VEP-based BCI system.ResultsThe offline classification accuracy of the BCI system was significantly improved from 67.6 ± 12.5% (raw data) to 97.3 ± 2.1% (data filtered by CSP and WF). The proposed approach was successfully implemented in an online BCI system, where subjects could make 20 decisions in one minute with classification accuracy of 90%.ConclusionsThe proposed single-trial detection approach is able to obtain robust and reliable VEP waveform in an automatic and fast way and it is applicable in VEP based online BCI systems.SignificanceThis approach provides a real-time and automated solution for single-trial detection of evoked potentials or event-related potentials (EPs/ERPs) in various paradigms, which could benefit many applications such as BCI and intraoperative monitoring.