Results 1 
3 of
3
Tracking of Nonstationary EEG with the Polynomial Root Perturbation
 in EMBS’96
, 1996
"... A method for estimation of the change point in event related desynchronization test is presented. The method is based on tracking of a single system pole of the timevarying ARMA model. The pole is approximated using perturbation theory. I. Introduction The end to automatic analysis of EEG is ofte ..."
Abstract

Cited by 4 (3 self)
 Add to MetaCart
(Show Context)
A method for estimation of the change point in event related desynchronization test is presented. The method is based on tracking of a single system pole of the timevarying ARMA model. The pole is approximated using perturbation theory. I. Introduction The end to automatic analysis of EEG is often segmentation of the EEG into stationary epochs. The parameters of autoregressive (AR) and autoregressive moving average (ARMA) models have been found to exhibit reasonably good discrimination efficiency in many cases [1]. Several algorithms exist for calculation of the model parameters. Since EEG is a nonstationary signal, a natural choice is to use an adaptive algorithm, such as the recursive least squares (RLS) algorithm [2]. Sometimes the model roots are more suitable for making further inference than the model parameters directly [3]. The use of model roots has been proposed earlier for the classification of stationary epochs of EEG [4]. The calculation of all the roots from model para...
TimeVarying ARMA modelling of Nonstationary EEG using Kalman Smoother Algorithm
"... An adaptiveauti egressive moving average (ARMA) modelling of nonstationary EEG by means of Kalman smoother is presented. The main advantage of the Kalman smoother approach compared to other adaptive algorithmssuo as LMSor RLSis that the tracking lag can be avoided. This advantage is clearly presente ..."
Abstract

Cited by 1 (0 self)
 Add to MetaCart
An adaptiveauti egressive moving average (ARMA) modelling of nonstationary EEG by means of Kalman smoother is presented. The main advantage of the Kalman smoother approach compared to other adaptive algorithmssuo as LMSor RLSis that the tracking lag can be avoided. This advantage is clearly presented with simu477S40 Kalman smoother is also applied to tracking of alpha band characteristics of real EEG duS ing an eyes open/closed test. The observed tracking ability of Kalman smoother, compared to other methods considered, seemed to be better. 1 INTROTRfi164 In the analysiso nosis1EEbbM1 EEG the interest is o1 tento estimate the timevarying spectralpro ertieso the signal. Atraditiow@ approt h to this is the spectro1M / metho d, which is basedo FoMM@: transfo5 matiof Disadvantageso thismetho d are the implicit assumptioo statioo/: y within each segment and the rather po o time/frequencyresofreque A better appro1 h is to use parametric spectral analysis metho ds basedo e.g. timevarying...
Report No. 5/96
, 1996
"... Estimation theoretical background of root tracking algorithms with applications to EEG ..."
Abstract
 Add to MetaCart
(Show Context)
Estimation theoretical background of root tracking algorithms with applications to EEG