Single-Channel EEG SSVEP-based BCI for Robot Arm Control

Abstract

Brain-Computer Interfacing (BCI) systems can be used to improve the quality of life of disabled individuals. Electroencephalography (EEG) Steady-State Visual Evoked Potentials (SSVEP) based BCI systems provide a non-invasive, inexpensive method of communication and control with minimal user training. Among many applications of BCI systems, robot control is widely used to restore motor functions of individuals with severe neuromuscular disabilities. In this paper, a low-cost, single-channel, SSVEP based BCI system is implemented to control the motion of the wrist and the gripper of a robot arm. The SSVEP user commands are generated by focusing the user's gaze on a set of light-emitting diodes (LED) flickering at different frequencies. For the identification of user intent to generate control commands, a classification algorithm is proposed, which is based on Euclidean distance measurement between prominent peaks of the SSVEP Fast Fourier Transform (FFT) spectrum, and the fundamental and harmonic spectral content of stimulus frequency.