Abstracts

Rationale: Extraoperative electrocorticography (ECoG) recording is performed to localize the seizure onset zone and eloquent areas in epilepsy presurgical evaluation.Naming-related high-gamma (70–110 Hz) augmentation has been suggested to be an excellent ECoG measure of cortical activation supporting language function. By averaging high-gamma activity at non-epileptic regions across a large number of patients, we recently generated a whole-brain level 4D atlas of language function, in which the spatial-temporal dynamics of naming-related cortical modulation is animated on a 3D standard surface image as a function of 10 ms. In the present study, we developed a 5D cortical atlas of language function, in which the developmental effect on naming-related high-gamma activity is additionally animated. We then determined how naming-related cortical dynamics are altered throughout the long developmental period from childhood to adulthood. Methods: A total of 103 epilepsy patients who underwent extraoperative ECoG were analyzed (age: 4-20 years, 9297 artifact-free nonepileptic electrodes). Patients estimated to have right-hemispheric language dominance were excluded. Each patient was assigned an auditory naming task, in which he/she overtly answered a series of audible sentence questions with a median duration of 1,800 ms (e.g.: ‘What flies in the sky?’). High-gamma modulation during the task was measured with respect to stimulus onset and offset at each recording site. Mixed model analyses were employed to determine the effect of age on naming-related high-gamma activity at each cortical region of interest (ROI). To maximize generalizability, multiple covariates including gender, number of antiepileptic drugs, proximity of seizure onset zone, handedness, and response time were incorporated in each model. Results: Positive age effects on high-gamma activity were noted 1) in the left posterior superior/middle-temporal gyrus at 100-200 ms following stimulus onset, 2) in the left inferior precentral gyrus at 200-500 ms following stimulus onset, 3) in the left rostral/caudal middle-frontal gyrus at 100-400 ms following stimulus offset. Negative age effects on high-gamma activity were noted in the bilateral anterior lateral temporal neocortex including the superior-temporal gyrus as well as in the right inferior-parietal region between -400 ms and +400 ms relative to stimulus offset. Conclusions: Age-related increases in high-gamma activity in the left temporal-frontal network may reflect the top-down processes with which more mature individuals effectively access lexical semantics. Conversely, age-related decrease in high-gamma activity in the bilateral anterior lateral temporal neocortex may reflect the bottom-up dynamics with which younger children meticulously analyze auditory stimuli during the naming task. Funding: NIH grants NS047550 (to E. A. ) and NS064033 (to E. A. )