Towards an SSVEP Based BCI With High ITR

Abstract

A brain-computer interface (BCI) provides the possibility to translate brain neural activity patterns into control commands without movement by the user. In recent years, there has been increasing interest in using steady-state visual evoked potential (SSVEP) in BCI systems; the SSVEP approach provides currently the fastest and most reliable communication paradigm for the implementation of a non-invasive BCI system. However, many aspects of current system realizations need improvement, specifically in relation to speed (in terms of information transfer rate as well as time needed to perform a single command), user variability and ease of use. With these improvements in mind, this paper presents the Bremen-BCI, an online multi-channel SSVEP-based BCI system that operates on a conventional computer making use of the minimum energy combination method for extraction of power information associated with the SSVEP responses. An additional advantage of the presented methodology is that it is fully online, i.e., no calibration data for noise estimation, feature extraction, or electrode selection is needed, the system is ready to use once the subject is prepared. The SSVEP-based Bremen-BCI system with five targets, an adaptive time segment length between 0.75s and 4s, and six EEG channel locations on the occipital area, was used for online testing on 27 subjects. ALL participants were able to successfully complete spelling tasks with a mean accuracy of 93.83% and an information transfer rate (ITR) of 49.93 bit/min.