Title: EEG-based motor imagery classification in BCI system by using unscented Kalman filter

Authors: Nik Khadijah Nik Aznan; Kyung-Moo Huh; Yeon-Mo Yang

Addresses: Department of IT Convergence Engineering, Kumoh National Institute of Technology, 1 Yangho-dong, Gumi-si, Gyeongsangbuk-do, Korea ' Department of Electronics Engineering, Dankook University, 119, Dandaero, Dongnam-gu, Cheonan, 330-714, Korea ' School of Electronic Engineering, Kumoh National Institute of Technology, 1 Yangho-dong, Gumi-si, Gyeongsangbuk-do, Korea

Abstract: This paper presents the unscented Kalman filter (UKF) to the BCI signal processing to classify the EEG-based motor imagery signals. UKF is applied to the common spatio-spectral pattern (CSSP) filters to improve the feature data extracted from the system. The CSSP is used to extract the related features by applying spatial and spectral filters to the system. Linear discriminant analysis (LDA) is used as the classification method to discriminate between class 1 (left hand) and class 2 (right hand) in the system. The performance criteria of the classification results are accuracy, kappa value, training time and confidence interval. The outputs of classification between our proposed scheme and the previous scheme are compared by using the simulations in Matlab. The simulation results indicate that the results of our proposed schemes outperform the other previous methods.

Keywords: brain-computer interface; BCI; unscented Kalman filter; UKF; EEG; electroencephalograms; motor imagery classification; signal processing; feature extraction; linear discriminant analysis; LDA; simulation; brain signals; right-hand motor imagery; left-hand motor imagery.

DOI: 10.1504/IJICT.2016.079962

International Journal of Information and Communication Technology, 2016 Vol.9 No.4, pp.492 - 508

Available online: 12 Oct 2016 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article