Abstracts

RATIONALE: Advances in technology allowing EEG recordings in the MRI should enable imaging of the epileptic focus using echo planar imaging (EPI) sequences based on the assumption that interictal epileptic spikes cause focal increases in blood flow.
METHODS: fMRI acquisitions were obtained with a 1.5 Tesla magnet using EPI sequences to image the entire brain. Two separate imaging paradigms were used: 1)spike-triggered fMRI acquisition and 2)continuous fMRI acquisition. For the spike-triggered method EPI sequences were obtained 3-6 seconds after epileptic discharges and compared to baseline sequences. For the continuous acquisition protocol, fMRI artifact on the EEG was removed using Fourier filtering (Hoffmann et al. Magn Reson Med 2000) thus enabling accurate recognition of spikes on the EEG. The timing of spikes in relation to the EPI acquisitions was then used to perform statistical analysis. Ballistocardiogram artifact did not interfere with the interpretation of the EEG for either study protocols and, therefore, no ballistocardiogram artifact removal procedure was required.
RESULTS: Twenty studies were performed on 16 subjects. Of the 15 triggered studies, activation was obtained in five. For these five studies, fMRI results were concordant with EEG finding in three and discordant with EEG in two cases. Of the remaining ten cases, four had no spikes during the study, one had a large fMRI artifact due to braces and the remaining five had no fMRI activation despite a technically adequate study. Of the five continuous studies, activation was observed in three. In one of these three cases, only seven spikes were recorded during the study. For all three cases, fMRI activation was concordant with EEG activity. Of the remaining two subjects, spikes were of small amplitude and could not be recognized in the filtered EEG in one, and the other patient had almost continuous spiking without return to baseline. Of the two patients who had both triggered and continuous studies, the studies were concordant for one subject and discordant for the other.
CONCLUSIONS: fMRI activation following interictal epileptic discharges can be measured using both spike-triggered and continuous acquisition protocols. Our experience demonstrates that the continuous acquisition protocol has a higher yield than the spike-triggered protocol. Finally the continuous protocol also has the advantage of not requiring the presence of a skilled EEGer during the fMRI acquisition.
Support: Canadian Institutes of Health Research grant MOP-38079.