Скачать или смотреть CORRELATION ANALYSIS OF VISUAL IMAGINARY EEG SIGNALS IN DIFFERENT FREQUENCY BANDS

CORRELATION ANALYSIS OF VISUAL IMAGINARY EEG SIGNALS IN DIFFERENT FREQUENCY BANDS
Haouari Aouatef, University of Badji Mokhtar, Algeria / Rachid Boudour, University of Badji Mokhtar Annaba, Algeria

(Presented at the International Conference on Research in Education and Science (ICRES) which took place on May 18-21, 2023, Cappadocia, TURKEY, https://www.2023.icres.net/ organized by International Society for Technology, Education and Science ((ISTES)-http://www.istes.org)

Brain-computer interface-based EEG devices, are a field that supports and assists challenged individuals to perform daily functions without any motor action, such BCI devices communicate directly to user intention, by decoding EEG signals. There are two types of BCI systems motor imagery and visual imagery, in this article, we will focus on the second one. The EEG signals, in our case, were recorded from 22 subjects, while performing an imaginary selection from an interface with two options(Turn on / Turn off the light). We aimed through this article at finding the best features to extract from EEG signals. The EEG dataset was measured by Pearson's correlation coefficient to find the linear association between feature pairs, in different EEG frequency subbands. NeuroSky headset was used to collect data, the data is displayed in eight frequency bands (delta, theta, alpha 1, alpha 2, beta1, beta 2, gamma 1, gamma2). The results show that the beta 1 and beta2 bands have the highest correlation index than in gamma, theta, and delta bands. The best two features were the mean and standard deviation pair. The mean and standard deviation features pair has the highest linear relationship, the standard deviation and variance pair was relatively high in the beta 2 bands, and the mean and variance features pair correlation was not very significant.