Slide Slam B8
The cortical representation of body part concepts
Stephen Mazurchuk1, Leonardo Fernandino1, Jia-Qing Tong1, Colin Humphries1, Lisa L Conant1, Jeffrey R Binder1; 1Medical College Wisconsin
Intro: Category-specific deficits have motivated much of the inquiry on the organization of semantic knowledge. There is an extensive neuroimaging literature on category-specific brain regions engaged during naming visually presented objects (Damasio et al. 1996, Martin et al. 1996), with particular emphasis on the contrasts of animals versus tools, and living versus non-living objects. Despite many decades of interest in the neural representation of the body (Head et al. 1911, Pick 1922), and reports of patients with selective impairment or preservation of body representations (Suzuki et al. 1997, Shelton et al. 1998), there is still little known about the processing of body part concepts as a semantic category (Coslett et al. 2002). FMRI work in this domain has focused almost exclusively on the visual perception of body parts, leaving unclear the distinctiveness of this category when lexically accessed (Devlin et al. 2005). Here we report a study comparing the processing of words denoting body part concepts with words from three other frequently studied object categories. Methods: 15 healthy adult participants were scanned using 3T simultaneous multi-slice fMRI. Stimuli consisted of 200 nouns chosen from the 4 conceptual categories of animals, manipulable artifacts, plants/food, and body parts (50 words per category). Categories were matched on word length, orthographic typicality, word frequency, and lexical decision accuracy. Stimuli were presented visually in an event-related design while participants performed a familiarity judgment task. The entire word list was randomized and repeated 6 times across 3 scanning sessions. Data were preprocessed and projected to a common surface using fmriprep, and mass univariate analysis was performed with AFNI’s 3dREMLFit. These results were then smoothed prior to group level analyses consisting of single-sample t-tests using FSL’s PALM. The reported results are familywise error corrected with a cluster-forming threshold of z > 2.2 (p < .01) and significance level of α < .05. Results: Compared to the other 3 categories, body part concepts elicited stronger activation in the left supramarginal gyrus (SMG), posterior superior temporal sulcus (pSTS), and posterior middle temporal gyrus (pMTG). Conclusion: The regions showing stronger activation for body part concepts in our analysis are similar to areas previously associated with observation and performance of hand actions and object-directed hand movements (Caspers et al. 2010, Grosbras et al. 2012). Activation of the SMG and pMTG was also previously shown to reflect manipulability ratings for noun concepts (Fernandino et al. 2016). We hypothesize that the body part concept representation arises from a combination of input from motor, tactile, proprioceptive, and visual experiential channels that converge in the SMG and pMTG.