Leveraging Peripheral Visual Stimuli for Enhanced SSVEP-Based BCIs in Fast Calibration Scenario

The decoding method based on training sample is utilized to effectively improve the recognition ability for steady-state visual evoked potential (SSVEP)-based brain computer interfaces (BCIs). However, these methods are not adapted to fast calibration scenarios because their performance degrades as the number of calibration trials decreases. For fast calibration scenarios, we proposed a multi-visual source aliasing matrix estimation task related component analysis (MVSAME-TRCA) method, which can simultaneously exploit the common information of the source aliasing matrix and the spatial filter across a peripheral visual stimulus (PVS). Experiment results show that the MVSAME-TRCA method outperforms the state-of-the art (SOTA) method such as eTRCA, ePRCA, same-eTRCA in a fast calibration scenario in the two public datasets. Notably, our approach only needs 28s to calibrate 40 targets in the two public datasets and achieves an averaged information transfer rates (ITRs) of 248.78 ± 86.47 bits/min and 186.86 ± 100.32 bits/min, respectively. This research greatly reduces the need for individualized calibration for SSVEP system and is conducive to the development of practical plug and play SSVEP-BCI.