Motor imagery task classification using transformation based features

• Main Authors: Khorshidtalab, Aida, Salami, Momoh Jimoh Eyiomika, Akmeliawati, Rini
• Format: Article
• Language: English; English; English
• Published: 2017
• Subjects: TK Electrical engineering. Electronics Nuclear engineering
• Online Access: http://irep.iium.edu.my/53662/1/2017_aida_biomedical.pdf; http://irep.iium.edu.my/53662/7/53662_Motor%20imagery%20task%20classification%20using%20transformation%20based%20features_SCOPUS.pdf; http://irep.iium.edu.my/53662/8/53662_Motor%20imagery%20task%20classification%20using%20transformation%20based%20features_WoS.pdf; http://irep.iium.edu.my/53662/; http://doi.org/10.1016/j.bspc.2016.12.006

tThis paper proposes a feature extraction method named as LP QR, based on the decomposition of theLPC filter impulse response matrix of the signal of interest. This feature extraction method is inspired byLP SVD and is tested in the context of motor imagery electroencephalogram. The extracted features areclassified and benchmarked against extracted features of LP SVD method. The two applied methods arealso compared regarding the required execution time, which further highlights their respective meritsand demerits. This paper closely examines the contribution of EEG channels of these two informationextraction algorithms too. Consequently, a detailed analysis of the role of EEG channels concerning thenature of the extracted information is presented. This study is conducted on the BCI IIIa competitiondatabase of four motor imagery movements. The obtained results indicate that the proposed method isthe better choice if simplicity is demanded. The investigation into the role of EEG channels reveals thatlevel of contribution each channel can be quite dissimilar for different feature extraction algorithms.