Abstracts

Rationale: The assessment of hemispheric dominance for language is a critical step in the presurgical evaluation of patients with intractable epilepsy. With this purpose, the merits of Magnetoencephalography (MEG) and Diffusion Tensor Imaging (DTI) have been tested independently, showing that predictions based on the activation/connectivity of Wernicke's area have a high level of concordance with the gold standard procedure for laterality assessments (i.e., the intracarotid amobarbital test, IAT). Here we used MEG and MEG-guided tractography to test the existence of specialized cortical regions and tracts, within Wernicke's area, supporting specific word comprehension subprocesses (i.e., phonological vs. semantic processing); and second, we tested the clinical validity of these measurements in relation to linguistic dominance for receptive language. Methods: Clinical and radiologic data from 8 patients were reviewed. All the patients were diagnosed with refractory epilepsy and underwent presurgical evaluations, including prolonged vEEG monitoring, MEG, DTI and the IAT. During the MEG test all patients performed a lexical decision task while they listened to words and pseudowords. Activity maps associated with word and pseudoword processing were derived for the late components of the evoked magnetic fields (~180 to 800 msecs) and the active areas were used as a seed for tractography analysis. Fractional anisotropy (FA), Diffusivity (mean, axial and radial), tract volume and number of tracts were calculated for the MEG-guided tracts. Individual laterality indexes were computed and compared with IAT estimates. Results: The IAT indicated left dominance for language in 8/8 patients. MEG-derived patterns of activation during word processing showed reliable involvement of 3 regions of Wernicke's area including: the posterior superior temporal gyrus (pSTG) in 8/8 patients, the middle temporal gyrus (MTG) in 6/8 patients and the supramargical gyrus (SMG) in 3/8 patients. Pseudoword processing produced activation in the pSTG (8/8) and sporadically in the MTG (2/8). Laterality indices derived from the tract volume, based on seed regions active during word processing (and not during pseudoword processing), was in complete agreement with IAT judgments of laterality (8/8 patients). FA and MD were not good predictors of laterality. Conclusions: Within Wernicke's area, MEG can be used to dissociate cognitive operations that are critical in word comprehension. MEG-guided tractography confirms the existence of distinct tracts, originating from Wernicke's area, functionally specialized in supporting phonological and semantic processing. Laterality judgments based on the volume of tracts connecting the MTG region (predominantly active during semantic processing) shows complete agreement with the judgments of the intracarotid amobarbital test.