Abstracts

Rationale: EEG acquired during functional MRI (EEG-fMRI) provides a method to interrogate the structures involved in the generation of generalised spike and wave (GSW) in idiopathic generalised epilepsy (IGE). We used statistical mapping and time course analysis to study epileptiform activity in a cohort of children with typical absence seizures (AS). Our aim was to identify cortical and sub-cortical regions involved in the genesis of spike and wave and to explore the timing of activity in these regions. Methods: Twelve children with typical AS recorded on routine EEG, who were not currently taking medication, underwent EEG-fMRI. Absence seizures were studied in 5 children and paroxysmal generalised spike and wave (GSW) in 8 children (10 successful studies) using standardized analysis methodology. Regions of interest (ROIs) were identified from the statistical maps and the event related time course of the blood oxygen level dependent (BOLD) signal was analyzed. Results: The group analysis of epileptiform activity showed increased blood oxygen level dependent (BOLD) signal change strongest in the thalamus but also in left lateral and midline frontal cortex. Reduced BOLD was seen in the lateral and mesial parietal lobe, caudate nuclei and brainstem reticular formation. Event related time course analysis was performed on ROIs in the thalamus, pons, caudate nuclei and parietal cortex revealing three patterns. In the thalamus BOLD signal change conformed closely to the standardised haemodynamic response function(HRF) while in the other sub-cortical regions (caudate and pons) an inital small increase in BOLD was followed by a more pronounced reduction in BOLD signal. In the parietal cortex BOLD signal appeared to rise gradually prior to a rapid fall following the onset of spike and wave. Longer events consistent with AS had delayed onset of BOLD signal change in sub-cortical structures, particularly the thalamus, when compared to short events. Conclusions: In children with typical AS epileptiform activity is not only associated with BOLD changes in the thalamus, caudate and parietal lobe but also in the reticular structures of the brainstem. In the sub-cortical structures BOLD signal change occurs at, or immediately after, electrographic onset while in the parietal cortex (and not in other cortical regions) there is a gradual positive drift in BOLD signal for 10 seconds prior. This suggests the parietal cortex may have a role in the initiation of epileptiform activity. Furthermore delays in the thalamic response, particularly in longer events, suggests the thalamus becomes involved following electrographic onset. These results raise important questions about the interaction between neural structures during spike and wave discharge and suggest event related BOLD analysis may provide a means to examine cortical and sub-cortical interactions during GSW.