Feasibility of hybrid BCI using ERD- and SSSEP- BCI

To increase information transfer rate (ITR) in brain-computer interface (BCI), many approaches better extracting new and robust features, or effectively combining such features have been proposed. Recently, hybrid BCI combining features simultaneously acquired from two different experimental paradigms has been reported. In this work, subjects performed tasks typically used in two BCI approaches, namely event-related de-synchronization (ERD) and steady-state somatosensory evoked potentials (SSSEP). Subjects imagined the movement of left or right hand (ERD) and gave more attention to the vibration (having different frequencies) of left or right finger (SSSEP). Offline analyses were found that half of the subjects did well ERD BCI rather than SSSEP BCI and information that was invisible in ERD pattern appeared in SSSEP pattern. As a preliminary study to develop a hybrid BCI using both ERD and SSSEP, we investigated how different information could be extracted from ERD and SSSEP, and further discussed how the hybrid ERD-SSSEP BCI is feasible.