BRAIN CONTROL OF ROBOTIC ARM USING AFFECTIVE STEADY-STATE VISUAL EVOKED POTENTIALS

Abstract

The recent emergence of successfully performing BrainComputer Interfaces (BCI) has given new hope to the disabled and elderly populations for improvements in quality of life. However, most non-invasive BCI systems which allow direct brain-to-machine control still have rather limited capabilities. Visual-flicker-based SSVEPBCI systems have received increased attention due to their potential capability to provide a large number of commands with high reliability. Yet, researchers have had limited success improving further the single-trial extraction of the weak SSVEP oscillations buried in strong brain ‘noise’. In a previous study we have shown that the optimization of the stimulus properties is essential for the enhanced performance of SSVEP–based BCI designs. This paper presents substantial enhancements in the brain response and the processing algorithms necessary for reliable multi-command SSVEP-BCI systems. The visual perception of flickering emotional video, instead of neutral checkerboards, enhanced substantially the measurable SSVEP responses of the brain. Furthermore, a single-trial phase-locking gradient measure was found to be more reliable and resulted in decreased variability when compared to wavelet energy changes. When BCI users operated a robotic arm with this Hybrid BCI platform, the speed, reliability, and information transfer rates were substantially improved when utilizing the proposed affective flicker paradigm.