Two tasks assessing receptive (verbal memory) and expressive (verbal fluency) language functions were compared.
•
PCA identified activation in the left temporal lobe and left hippocampus with both tasks in addition to frontal activation with the verbal fluency task.
•
In healthy participants and patients, better source localization data and laterality results were obtained with the verbal memory task.
Abstract
Objective
The aim of this study was to develop a presurgical magnetoencephalography (MEG) protocol to localize and lateralize expressive and receptive language function as well as verbal memory in patients with epilepsy. Two simple language tasks and a different analytical procedure were developed.
Methods
Ten healthy participants and 13 epileptic patients completed two language tasks during MEG recording: a verbal memory task and a verbal fluency task. As a first step, principal component analyses (PCA) were performed on source data from the group of healthy participants to identify spatiotemporal factors that were relevant to these paradigms. Averaged source data were used to localize areas activated during each task and a laterality index (LI) was computed on an individual basis for both groups, healthy participants and patients, using sensor data.
Results
PCA revealed activation in the left temporal lobe (300 ms) during the verbal memory task, and from the frontal lobe (210 ms) to the temporal lobe (500 ms) during the verbal fluency task in healthy participants. Averaged source data showed activity in the left hemisphere (250–750 ms), in Wernicke's area, for all participants. Left hemisphere dominance was demonstrated better using the verbal memory task than the verbal fluency task (F1,19 = 4.41, p = 0.049). Cohen's kappa statistic revealed 93% agreement (k = 0.67, p = 0.002) between LIs obtained from MEG sensor data and fMRI, the IAT, electrical cortical stimulation or handedness with the verbal memory task for all participants. At 74%, agreement results for the verbal fluency task did not reach statistical significance.
Significance
Analysis procedures yielded interesting findings with both tasks and localized language-related activation. However, based on source localization and laterality indices, the verbal memory task yielded better results in the context of the presurgical evaluation of epileptic patients. The verbal fluency task did not add any further information to the verbal memory task as regards language localization and lateralization for most patients and healthy participants that would facilitate decision making prior to surgery.
