Abstracts

Rationale: Cortical language mapping involves the identification of essential language cortex, which is typically spared from resection with the goal of preserving postoperative language function. Object naming is the most widely used task for this purpose; however, when stimulation impedes naming, it is unclear whether this reflects impaired access to word meaning (i.e., semantics), word sound (i.e., phonology), or both. This distinction is clinically relevant, with implications for level of disability and amenability to remediation. Two sets of psycholinguistic tasks were administered at sites where stimulation impaired naming to determine whether semantic vs. phonological processes were disrupted. Access to distinct types of word information is critical in the two tasks: information about word meaning in the semantic task, information about word sounds in the phonological task. We hypothesized that semantic and phonological naming sites would be anatomically distinct. Methods: Subjects were 12 pharmacologically intractable, TLE patients (9 female, mean age = 34.8, SD = 11.1) who underwent extraoperative language mapping prior to surgical resection for seizure control. Stimulation mapping tasks included visual object naming and auditory description naming. At sites positive for naming, two psycholinguistic tasks were administered: 1) Semantic task: patients were presented pictured items during stimulation and indicated (via button press ) whether the item belongs to a particular semantic category (e.g., edible, found indoors); 2) Phonological task: patients indicated whether the item name begins with a particular sound (e.g., p or f ). Results: Across patients, we identified 53 naming sites (38 visual naming, 15 auditory naming). Semantic task performance was impaired at 3 of these sites, phonological task performance was impaired at 14 of these sites, and both semantic and phonological task performance were impaired at 7 of these sites. Topographically, phonological-naming sites were broadly distributed across left lateral temporal cortex, whereas semantic and mixed semantic-phonological naming sites were found primarily in the posterior and inferior left temporal region. There was no clear pattern evident in phonological versus semantic processing related to auditory versus visual naming sites. Conclusions: Results suggest that naming impairment related to anterior temporal abnormalities is due primarily to impaired phonological processing, whereas naming impairment resulting from posterior or inferior temporal damage reflects problems with both semantic and phonological processing. As the anterior temporal region is typically most affected in TLE, we speculate that naming difficulty in left TLE primarily reflects problems accessing information regarding word form, with relatively preserved access to word meaning. Acknowledgment: This research was supported by NIH grant R01 NS35140 (MH).