Abstracts

Generalised spike wave activity (GSW) is the surface EEG hallmark of absence seizures. Recent combined EEG-fMRI studies during GSW have shown negative and positive BOLD changes. Although there was variability across subjects, a common pattern was found at the group level, with negative BOLD changes in the so-called [quot]default mode[quot] brain areas and positive BOLD changes in the thalamus. We investigated these findings further at 3T, using an arterial spin label (ASL) sequence to measure the perfusion changes during GSW in addition to a standard BOLD sequence to assess reproducibility., Four patients with frequent GSW who had shown significant negative BOLD changes in a 1.5T study were recruited for investigation at 3T. Two patients had idiopathic generalised epilepsy and two had secondary generalised epilepsy. We used a 3T Siemens Allegra (Germany) and MR-compatible EEG system(Brainproducts, Munich, Germany) for simultaneous fMRI and EEG data acquisition. Patients underwent a 30 minute BOLD session (TR/TE 3120/40 ms, 48 slices, 500 volumes), followed by a 30 minute ASL session for perfusion images (QUIPPS2 sequence: TR/TE 2300/30, 6 slices, 800 volumes label and control images). A time-series of difference images was dereived from the perfusion series. SPM2 was used for image preprocessing and analysis. Images were spatially realigned and normalised; perfusion images were co-registered to corresponding BOLD images; both BOLD and perfusion images were smoothed with an 8mm FWHM Gaussian kernel. EEG was visually coded to mark the start and stop of GSW events. These were used to create a boxcar model convolved with a canonical HRF and its temporal derivative for analysis of the BOLD and perfusion MRI data according to the general linear model., Frequent GSW discharges were observed during all data acquisition sessions. Negative BOLD changes were seen in bilateral frontal (2 patients) and parietal cortices (2 patients), and in the posterior cingulate (2 patients). Thalamic (1 patient) and frontal activation were seen in 2 patients. The spatial distribution of activations was very similar to that seen previously at 1.5T in all cases. The time courses of the ASL changes matched closely those of the BOLD decreases. The decrease in perfusion at the global statistical maximum was between 11.1 to 37 percent., Although there was inter-subject variability, GSW-related BOLD increase and decrease activation patterns were found to be reproducible between 1.5T and 3T for each case studied. On the whole, GSW-related BOLD decreases was associated with a decrease in cerebral perfusion., (Supported by Wellcome Trust, Medical Research Council.)