An embedded hybrid BCI speller

Abstract

The present work proposes an embedded real time hardware-software platform for brain computer interfaces (BCI) based on steady state visual evoked potentials (SSVEP) and alpha rhythm. The complete implementation of the embedded system is described, including the electroencephalogram (EEG) amplifier, signal acquisition and processing stages, and practical implementation. The device is a 25 characters hybrid-BCI speller that uses 5 visual stimulus and visual alpha waves for control instructions. The system focuses on simplicity and portability, using only two electrodes, a simple EEG amplifier and an embedded computer. The speller does not require any training from the user and provides real time biofeedback to increase attention on stimuli. The platform is based on an embedded system with a real time operative system, Windows CE. Experimental results are presented in order to show the feasibility of the proposed system. Working as a speller, the current average speed is around 7 characters per minute.