Abstracts

Rationale: To provide new insight about the location and propagation of brain activities observed in patients with an unusual form of benign childhood epilepsy (BCE) using magnetoencephalography (MEG). Methods: Two patients were investigated with simultaneous MEG / EEG recording. Case 1: A 7 year old girl had a single nocturnal generalized tonic clonic seizure. Electroencephalography (EEG) showed sharp waves, abundant during sleep, with stereotyped diphasic morphology and a complex dipolar distribution, manifesting maximum negativity in the left frontal region and lower amplitude positivity in the left parietal region, usually synchronous with lower amplitude sharp waves in homologous regions of the right hemisphere. Case 2: A 5 year old boy had a single autonomic seizure. EEG during sleep showed sharp waves with stereotyped diphasic morphology. The sharp waves had maximum negativity in the left frontal regions and lower amplitude negativity in the left parietal region, usually with synchronous lower amplitude sharp waves in homologous regions of the right hemisphere. The EEG features in these children are consistent with an unusual form of BCE along the spectrum described by Panayiatopoulos. The patients are neurologically normal, each having had only a single seizure without medications at 9 months (Case 1) and 7 months (Case 2). In both cases, interictal activities seen on MEG were not amenable to single equivalent current dipole (ECD) modeling, because two regions (frontal and parietal) were activated simultaneously. Accordingly, multiple ECD modeling and minimum norm modeling were applied to analyze the interaction of these two activities. Results: Multiple ECD modeling revealed activities arising from the parietal region, propagating to the ipsilateral frontal region in both cases, and returning to the parietal region in Case 1. In Case 1, the orientations of the first and the third dipole components in the parietal region were orthogonal. This finding suggests that the epileptic activity in the parietal region evolves from a single ECD into a regional (rotating) source. In Case 2, the orientations and localization of dipoles were variable and complicated even with multiple ECD modeling. As an alternative method to multiple analysis, we used the dynamic significance map from a minimum norm model to create an image of cortical significance as a function of time. It showed relatively broad distribution of activities arising from the parietal region, propagating to the medial aspect of bilateral frontal regions with maximum amplitude in the ipsilateral medial frontal region. These findings helped to explain the complicated EEG pattern of frontal negativity and parietal positivity. Conclusions: MEG with multiple ECD modeling and significance modeling, helps to clarify the complicated brain activities seen in this unusual form of BCE, and provides further insight into the complexity and proper interpretation of scalp EEG recordings.