Channel Selection of EEG-Based Cybersickness Recognition during Playing Video Game Using Correlation Feature Selection (CFS)

Abstract

Recently, the rapid development of 3D movie or video games, causing the phenomenon of cybersickness. Cybersickness is an unpleasant symptom (dizziness, nausea, vomiting, and disorientation) that occur to humans when exposure in 3D movie or video games within a certain time. It can disrupt psychic and physical condition of the human if not handled appropriately. Many studies have been done to investigate cybersickness using physiological measurements, one of which is EEG. However, earlier studies have not discussed an optimal channel location for identifying cybersickness on EEG. In this paper, we proposed Correlation Feature Selection (CFS) method to select features in order to determine best channel selection. The power percentage (PP) features of alpha (α), beta (β) and theta (θ) bands were extracted on all channels. CFS method obtained 3 optimal channels location on F3, O1, and O2 from PP feature of beta (β) band. The investigating of cybersickness employs three compare classifiers i.e. SVM-RBF, k-NN, and LDA. According to our result, LDA is the best classifier for identifying cybersickness. By using CFS method, it can improve performance accuracy from 83% to 100 %. Hence, we conclude that beta frequency band on frontal and occipital area is suitable to measure EEG-based cybersickness.