Category-specific representation of animal, food, tool, and vehicle concepts in the brain
Poster D67 in Poster Session D with Social Hour, Friday, October 7, 5:30 - 7:15 pm EDT, Millennium Hall
Jiaqing Tong1, Leonardo Fernandino1, Stephen Mazurchuk1, Lisa L. Conant1, Jeffrey R. Binder1; 1Medical College of Wisconsin
Many functional imaging studies have addressed the neural representation of distinct object categories, providing evidence for differential responses in high-level visual cortex to images of faces, scenes, body parts, tools, animals, and other categories. Studies using words to elicit concepts in different categories have been less numerous and less conclusive. We examined object category effects in a large fMRI study using word stimuli to elicit concept retrieval while controlling for a range of word form nuisance variables. A second aim was to test a prediction of an experiential account of category specificity, which claims that category effects arise from systematic differences in experiential content between categories, and thus there should be no residual category effects on brain activation after accounting for item-level experiential content. Thirty-nine participants were shown 160 English nouns consisting of 40 items in each of 4 object categories using a fast event-related design during 3T fMRI. Data were also obtained for 160 event nouns from 4 categories, which are not used in the current analyses. Each stimulus was presented 6 times across 3 sessions on separate days. Participants rated the familiarity of each word on a 1 to 3 scale. MRI data were preprocessed with the HCP pipeline. Functional data were processed under a general linear model that included each of the object and event categories as 8 regressors of interest. Other lexical variables coding length, orthographic and phonological properties, and word frequency, were included as nuisance regressors. Paired t tests were performed for all possible object category pairs. Thresholded (p < .001) t maps were corrected for multiple comparisons via permutation testing (α < 0.05). As a strict test for category-specific effects, a conjunction map was created for each category compared to the other 3 object categories. In a second analysis, we included 65 experiential (sensory, motor, affective, temporal, spatial, social, etc.) feature dimensions, derived from crowd-sourced ratings of each word, as nuisance regressors. Conjunction maps for each category compared to the other categories were created as before. Tool words activated the left posterior middle temporal gyrus (pMTG) compared to the other categories. Food words activated bilateral insula and orbital frontal cortices (OFC). Vehicle words activated bilateral parahippocampal cortex and right posterior cingulate gyrus. There were no regions where animal words produced stronger activation than the other categories. When the 65 experiential feature regressors were included in the model – but not when 65 random regressors were used instead – all category-specific activation differences disappeared. Category-specific representation for tool concepts in pMTG and food concepts in OFC are consistent with several prior studies using word stimuli. Insular activation for food concepts likely reflects involvement of this region in gustatory and olfactory processing. Vehicle concepts activated regions previously identified with processing visual scenes. We found no evidence for category-specific animal concept representation. When experiential content was taken into account, category-specific effects disappeared. These results suggest that the preferential representation of these categories may be the result of cortical specialization for the representation of specific experiential features.
Topic Areas: Meaning: Lexical Semantics, Multisensory or Sensorimotor Integration