Minimal-assisted SSVEP-based brain-computer interface device

Steady-state visual evoked potential (SSVEP)-based brain computer interface (BCI) device is one of the most accurate assistive technologies for the persons with severe disabilities. However, for the existing systems, the persons with disabilities still need the assistance for the long period of time as well as the continuous time usages. In order to minimize this problem, we propose the SSVEP-based BCI system that the persons with disabilities can enable/disable the BCI device by alpha band EEG and control the electrical devices by SSVEP. A single-channel EEG (O1 or O2) is employed. Power spectral density via periodogram at the four stimulated frequencies (6, 7, 8, and 13 Hz) and their harmonics are used as the features of interest. Simple threshold-based decision rule is applied to the selected features. With the minimal need for assistance, the classification accuracy of the proposed system ranged from 75 to 100%.