Determination of optimal frequency ranges using common spatial pattern images

Author

Sakkaff, Zahmeeth ; Nanayakkara, A.

Author_Institution

Artificial Intell. & Appl. Electron. Res. Unit, Inst. of Fundamental Studies, Kandy, Sri Lanka

fYear

2013

fDate

17-20 Dec. 2013

Firstpage

133

Lastpage

138

Abstract

In this paper we present brain maps constructed for Visualization of Arrow Movement (VAM) mental tasks with EEG signals recorded from 20 locations on the scalp. Method of Common Spatial Patterns (CSP) is employed to generate the spatial maps after filtering raw EEG signals into nine different frequency bands. Separate CSPs were obtained for six subjects who participated in this study. These brain maps show distinct spatial patterns that can be used for distinguishing two mental tasks in VAM. For every subject, optimal frequency range which is useful in distinguishing mental tasks in VAM was obtained.

Keywords

brain-computer interfaces; electroencephalography; medical signal processing; signal classification; spatial filters; CSP; EEG signal filtering; VAM mental tasks; brain maps; common spatial patterns; optimal frequency range; spatial maps; visualization of arrow movement mental tasks; Conferences; Covariance matrices; Eigenvalues and eigenfunctions; Electrodes; Electroencephalography; Information systems; Random access memory; Common spatial patterns; Electroencephalograph (EEG) classification; Spatial filters; mental tasks;

fLanguage

English

Publisher

ieee

Conference_Titel

Industrial and Information Systems (ICIIS), 2013 8th IEEE International Conference on

Conference_Location

Peradeniya

Print_ISBN

978-1-4799-0908-7

Type

conf

DOI

10.1109/ICIInfS.2013.6731969

Filename

6731969