Abstracts

RATIONALE: The character of cerebral sources of scalp EEG ictal rhythms has seldom been directly confirmed. Misconceptions are therefore likely regarding the extent of cortical activity required to generate seizure patterns, such as those in temporal lobe epilepsy.
METHODS: We recorded simultaneously 26 channels of scalp EEG and 46 to 98 channels of intracranial EEG in 16 epilepsy surgery candidates. Cerebral ictal discharges and interictal spikes with and without a scalp EEG correlate were identified, and the area of the cortical generator was estimated from the spatial extent of electrode contacts demonstrating concurrent depolarization.
RESULTS: Only a fraction of intracranial EEG spikes were associated with recognizable scalp potentials. Similarly, ictal discharges when restricted to a few electrode contacts often resulted in no scalp EEG rhythms. Synchronous or at least temporally overlapping activation of 10-20 cm2 of gyral cortex was a common substrate for scalp-recordable spikes and ictal patterns. Cerebral generators with areas of less than approximately 6 cm2 did not produce scalp potentials. The onset of cerebral ictal activity was usually not reflected on scalp EEG until sufficient cortex was recruited into ictal activity. This commonly took several seconds. Propagation most often appeared as a moving patch of cortical depolarization. The changing geometry of the activated cortex was reflected in evolving scalp voltage topography.
CONCLUSIONS: Brain sources of scalp EEG ictal rhythms are larger than commonly thought. A large area of cortex must be recruited into synchronous activity for scalp potentials to be evident. Accordingly, scalp EEG ictal onset seldom reflects intracranial EEG ictal onset. Propagation further obscures the actual ictal origin. Any analysis of ictal rhythms to define seizure origin should be performed on the earliest scalp EEG ictal pattern.