Abstracts

Rationale: Using whole-brain 4D maps of language function based upon event-related modulation of electrocorticography (ECoG) signals, we determined how the spatial-temporal dynamics of language-related cortical modulations differed between visual and auditory domains as well as between young children and older individuals. Methods: Patients estimated to have right-hemispheric language dominance were excluded. Thus, this study included 79 patients with focal epilepsy who underwent two-stage epilepsy surgery with extraoperative ECoG recording. An older group consisted of 68 patients at least 10 years of age (6267 artifact-free nonepileptic electrodes), whereas a younger group included 11 children younger than age 10 (1084 electrodes). ECoG amplitude changes of high-gamma (70-110 Hz) activity during picture and auditory naming tasks were animated on the average surface image in a 4D manner. Thereby, high-gamma augmentation was treated as a summary measure of cortical activation. Results: Following onset of presentation of a picture, high-gamma augmentation involved the occipital lobes, inferior surfaces of the temporal lobes, inferior-parietal lobules, and inferior precentral gyri of both hemispheres. At the same time, high-gamma attenuation involved the inferior-frontal gyri, bilaterally. Prior to overt responses to pictures, high-gamma augmentation involved the pars opercularis of the left inferior-frontal gyrus as well as bilateral superior-frontal and inferior precentral/postcentral gyri. High-gamma attenuation then involved right anterior inferior-/middle-frontal regions. Picture naming, compared to auditory naming, elicited high-gamma augmentation preferentially in the temporal regions infero-medial to the inferior-temporal sulcus, whereas auditory naming did so supero-lateral to that sulcus. Picture naming, compared to auditory, elicited a smaller extent of high-gamma augmentation in the left frontal lobe. We found a double dissociation between patient age groups and high-gamma profiles; left-hemisphere dominant high-gamma augmentation was noted in the frontal-parietal regions in the older group, whereas in the anterior fusiform gyrus in the younger group. Conclusions: Our 4D brain maps revealed contrasting characteristics in the spatial-temporal profiles of language-related cortical modulations between visual and auditory domains as well as between young children and older individuals. ECoG mapping using a picture naming task alone may not sufficiently localize the language areas in the left frontal lobe. Our ECoG observations support the notion that young children recognize objects primarily with bottom-up processing via the left ventral visual language pathway, whereas mature individuals exert picture naming with more top-down processing via the left frontal-parietal regions. Funding: NIH grants NS064033 (to Eishi Asano)