Recognition of Eye Movements Based on EEG Signals and the SAX Algorithm

Abstract

For patients with motor disabilities and difficulties to interact with computer and devices, brain–computers interface (BCI) may provide them with new ways to solve this problem. The patients may use a portable electroencephalography (EEG) device to instruct a computing device via eye movements. We propose the use of commercial off-the-shelf (COTS) EEG devices and pattern classification as a potential solution. In this paper, we investigate simple eye movement recognition using the symbolic aggregate approximation (SAX) algorithm and compare its suitability and performance against known classification algorithms such as support vector machine (SVM), k-nearest neighbour (KNN) and decision tree (DT). The SAX-based recognition performed better than the three classification algorithms. The SAX-based recognition was also able to achieve higher accuracy with only one single simple feature. These results showed that SAX could be a suitable and efficient technique to perform simple eye movement recognition using EEG signals.