Processing of Epileptic EEG
Processing of epileptic EEG is required for better understanding of the mechanisms underlying the generation and propagation of seizure electrical activity, as well as for mapping of epileptic foci in patients with intractable focal epilepsy who are candidates for surgery. Most processing methods applied to the EEG require linearity; but, before the question of linearity/non linearity is handled, the non stationary character of the EEG signal has to be tackled.
The nonstationary multichannel epileptic EEG has been submitted to adaptive segmentation, dividing the signal into quasi-stationary segments of uneven length. Next, the linear or nonlinear character of the system has been investigated on these quasi-stationary segments using methods for system identification. Vectorial parametric modeling of the multichannel signal has been undertaken on short, quasi-stationary EEG segments, for which no significant deviations from linearity were detected.
Estimation of the delay between various EEG channels by adaptive least-square filtering (a lattice-ladder type) was carried out on EEG from rats with induced focal epilepsy in order to map the epileptic focus. EEG from patients with primary generalized epilepsy was subject to an efficient method for spectral estimation, based on vectorial parametric modeling. Preictal signal segments were processed in order to detect changes which occur prior to the clinical outbreak of the seizure.
KeywordsFocal Epilepsy Epileptic Focus Reference Window Adaptive Segmentation Vectorial Parametric Modeling
Unable to display preview. Download preview PDF.
- 9.George, R. E., and Michael, J. E., 1991, Efficacy of vagal nerve stimulation for the treatment of partial seizures, Epilepsia 32: S3–80.Google Scholar
- 10.Gersch, W., 1987, Non-stationary multichannel time series analysis, in Gevins, A. S. and Remond, A. (eds), EEG Handbook Revised Series Elsevier, Amsterdam, vol. I1: 261–296.Google Scholar
- 14.Kuhlman, W. N.. and Allison, T, 1977. EEG feedback training in the treatment of epilepsy: some questions and some answers. Pavlovian J. Biol. Sci. 12: 112–122.Google Scholar
- 16.Moddemeijer, R., 1989, Delay-Estimation with Application to Electroencephalograms in Epilepsy, Ph.D Thesis, University of Twente, Enschede.Google Scholar
- 17.Ojemann, G. A., Engel, J., 1987, Acute and chronic intracranial recording and stimulation, in J. Engel (ed), Surgical Treatment of the Epilepsies, 263–288, Raven Press, New York.Google Scholar
- 18.Panych, L. P., and Wada, J. A., 1990, Computer applications in data analysis, in Wada, J. A., and Ellingson, R. J., (eds), Clinical Neurophysiology of Epilepsy, Handbook of Electroencephalography and Clinical Neurophysiology Revised Series, Elsevier, Amsterdam, Vol. IV, 361–385.Google Scholar
- 20.Pham, D. T., and Tong, D. Q., 1990, Maximum likelihood estimation for multivariate autoregressive model, Research report RR 807-M, Laboratoire LMC/IMAG, Université Joseph Fourier, Grenoble.Google Scholar