Abstracts

Synchrony Changes between Hippocampal and Neocortical EEG Signals Precede Seizure Activity Induced by Intrahippocampal NMDA Application in Freely Behaving Rats

Abstract number : 1.055
Submission category : Clinical Neurophysiology-Computer Analysis of EEG
Year : 2006
Submission ID : 6189
Source : www.aesnet.org
Presentation date : 12/1/2006 12:00:00 AM
Published date : Nov 30, 2006, 06:00 AM

Authors :
1Hans von Gizycki, 2Shirn L. Baptiste, 2Geza Medveczky, 2Ruben I. Kuzniecky, 2Orrin Devinsky, and 2Nandor Ludvig

This study aimed to determine the changes in synchrony between hippocampal and neocortical EEG signals prior to hippocampal seizures in rats. Synchrony between EEG channels has usually been calculated using cross-correlation and coherence. The limitations of these approaches are: (a) the assumption that EEG signals exhibit a certain degree of amplitude and phase stationarity, and (b) the production of a combined measure of synchrony where the contributions of phase and amplitude cannot be parsed out. To rectify these limitations, we utilized and advanced newer synchrony analysis methods., Rats were prepared with chronic hippocampal and neocortical electrodes, and artifact-free EEG recordings were conducted during behavior. Via microdialysis or minipump-assisted microinjection, 500 microM NMDA was delivered into the hippocampus to induce seizures. EEG data were stored for: (1) pre-NMDA state, (2) post-NMDA state prior to seizure, and (3) post-NMDA state during seizure. Synchrony index (SI) and instantaneous phase synchrony distribution (IPSD) were calculated, as follows: EEG signals at 8, 14, 20, 40 and 60 Hz were extracted from the EEG data using a continuous wavelet transform (CWT), applying a first order complex Gaussian wavelet. The complex results of the CWT for both channels were transformed to an angular term and unwrapped, producing a phase time series for each EEG channel. Then the phase distance between the phase time series derived from each EEG channel was calculated so that these distances can be seen as instantaneous phase synchrony. Next, the averaged normalized phase distances were calculated, yielding the SI. Finally, the IPSD was calculated by determining the digression of phase distances from a Gaussian distribution, specifically by estimating the kurtosis of the distribution of the distances. Statistical analysis consisted of repeated measures factorial ANOVA for SI and IPSD, with EEG frequency and NMDA treatment states as factors., Main effects for NMDA treatment states and the interaction between NMDA states and EEG frequency were significant for IPSD, but not for SI. Post-hoc analysis revealed that at 8 Hz the IPSD in the post-NMDA state prior to seizures (-1.048) significantly differed from the IPSD measured before NMDA treatment (-1.210)., In the present seizure model, pre-seizure synchrony changes could not be captured by using SI. Instead, these synchrony changes could be captured with measuring the alteration of the properties of distribution of instantaneous phase synchrony, as expressed in the IPSD values. Thus, IPSD may be used as a sensitive marker for revealing the critical phase synchrony changes that precede seizure activity., (Supported by NYU/FACES.)
Neurophysiology