Exploring Multistable Perception to Modulate Steady-State Visual Evoked Potentials using a Computer Graphics Software for a Brain-Computer Interface

Abstract

This study investigated the use of multistable perception to modulate steady-state visual evoked potentials (SSVEPs) using a computer graphics program and an LCD monitor. The Rubin’s vase ambiguous image was employed for this purpose. A visual stimulator was developed in C++ using the Open Graphics Library (OpenGL). EEG was measured on the occipital area (Oz) of the subject, and a Brain-Computer Interface (BCI) was developed to control a robot in a virtual reality environment. Commands were recognized using the Multivariate Synchronization Index (MSI) method. The results demonstrate the feasibility of eliciting and modulating SSVEP response. During online evaluation, the subject were able to command a virtual telepresence robot with over 80% accuracy and an Information Transfer Rate (ITR) of 7 bits/min.