Comparison of the accuracy of two segmentation methods in medium rate EEG transitions

Author

Hiltunen, J.K. ; Kaipio, Jari P.

Author_Institution

Dept. of Appl. Phys., Kuopio Univ., Finland

Volume

4

fYear

2000

fDate

2000

Firstpage

3018

Abstract

Two methods for the estimation of medium rate transitions of nonstationary EEG are compared. The first method is based on basis constrained least squares time-varying autoregressive model (TVARLS). This model is reparametrized so that a single parameter determines when the change occurs. The second method is a segmentation algorithm (Bodenstein-Praetorius segmenter). The method is based on the comparison of changes of power spectra between two short EEG epochs. The comparison of these methods is performed using realistic EEG simulations

Keywords

electroencephalography; medical signal processing; spectral analysis; Bodenstein-Praetorius segmenter; EEG analysis; basis constrained least squares time-varying autoregressive model; electrodiagnostics; power spectra changes comparison; reparametrized model; segmentation algorithm; signal segmentation; Adaptive algorithm; Brain modeling; Delay estimation; Electroencephalography; Frequency synchronization; Least squares methods; Neuroscience; Physics; Rhythm; Steady-state;

fLanguage

English

Publisher

ieee

Conference_Titel

Engineering in Medicine and Biology Society, 2000. Proceedings of the 22nd Annual International Conference of the IEEE

Conference_Location

Chicago, IL

ISSN

1094-687X

Print_ISBN

0-7803-6465-1

Type

conf

DOI

10.1109/IEMBS.2000.901515

Filename

901515