Abstracts

RATIONALE: Functional cortical mapping prior to seizure surgery has traditionally relied upon intraoperative or extraoperative cortical stimulation mapping. However, this procedure is often limited by afterdischarges and by stimulated seizures. Evoked potentials have been routinely used to localize the central sulcus, but not surrounding sensorimotor cortex. An alternative electrophysiological method for functional mapping consists of event-related spectral analysis, in which functional cortical activation is correlated with event-related changes in the EEG power spectrum. Prior scalp EEG studies of these changes have focused on event-related power suppression in the alpha (8-13 Hz) band. However, during unilateral self-paced finger movements alpha suppression occurs over bilateral central scalp regions. In subdural ECoG recordings during visually-paced fist-clenching (Crone et al., [italic]Brain[/italic] 1998;121:2271-2315) we observed a similar pattern of alpha suppression, but also discovered a novel event-related power increase in a broad [dsquote]high gamma[dsquote] band (~80-100 Hz). This index of cortical activation occurred exclusively over contralateral sensorimotor cortex. In the present study we further investigate the anatomical specificity of ECoG gamma with a self-paced finger movement task similar to those used in scalp EEG studies.
METHODS: ECoG was recorded in 6 patients during a self-paced button press task (both hands tested in 3 patients, only the contralateral hand in the other 3). Event-related changes in ECoG band power were calculated for alpha and high gamma bands. In all patients ESA results were compared with clinical cortical stimulation maps. In two subjects event-related power changes were also compared with somatosensory evoked potentials (SSEPs) elicited by contralateral median nerve stimulation.
RESULTS: Topographical patterns of gamma augmentation were more discrete and consistent with cortical stimulation maps than were those of alpha suppression. In the three subjects in whom both hands were tested, gamma augmentation was observed only during contralateral finger movements, while alpha suppression was observed with both contralateral and ipsilateral finger movements. Gamma topography was also consistent with that of the median SSEP in two patients.
CONCLUSIONS: This study illustrates the potential clinical utility of ECoG spectral analysis. The topographic specificity of gamma power augmentation make it particularly well suited to serve as an index of sensorimotor cortex activation prior to seizure surgery. Although this novel index may not supersede SSEPs for localization of the central sulcus, its localizing accuracy for sensorimotor cortical function provides support for its use in other functional-neuroanatomic domains such as language.
Support: 1R01NS40596-01A1 and
National Epifellows Foundation