Detection of characteristic wave in EEG using locally stationary AR model

Author

Fukami, Tadanori ; Akatsuka, Takao ; Saito, Yoichi

Author_Institution

Fac. of Eng., Yamagata Univ., Japan

Volume

2

fYear

2001

fDate

2001

Firstpage

1857

Abstract

It is much important to detect the characteristic wave in EEG clinical diagnosis. They are classified into three groups. One is a stationary wave such as an alpha or beta wave, the others are burst wave (semi-transient wave) and transient wave, nonstationary wave, such as hump wave. The final goal of our research is labeling of EEG wave in short section. In this research, we tried to detect a hump wave by using a locally stationary AR model as a first trial. We employed this method for clinical EEG data. The accuracy of detection showed a 76% level.

Keywords

electroencephalography; medical signal detection; medical signal processing; physiological models; EEG characteristic wave detection; alpha wave; autoregressive model; beta wave; burst wave; clinical EEG data; detection accuracy; electrodiagnostics; hump wave detection; locally stationary AR model; nonstationary wave; transient wave; Brain modeling; Electroencephalography; Equations; Filtering; Fluctuations; Frequency; Heart rate; Kalman filters; Shape; Time series analysis;

fLanguage

English

Publisher

ieee

Conference_Titel

Engineering in Medicine and Biology Society, 2001. Proceedings of the 23rd Annual International Conference of the IEEE

ISSN

1094-687X

Print_ISBN

0-7803-7211-5

Type

conf

DOI

10.1109/IEMBS.2001.1020585

Filename

1020585