Neural regions supporting lexical processing of objects and actions: A case series analysis

• ¹ Johns Hopkins University, Cognitive Science, United States
• ² Johns Hopkins University, Radiology, United States
• ³ Johns Hopkins University, Neurology; Physical Medicine & Rehabilitation; Cognitive Science, United States

Introduction. Linking semantic representations to lexical items is an important cognitive process for both producing and comprehending language. Past research has suggested that the bilateral anterior temporal lobes are critical for this process (e.g. Patterson, Nestor, & Rogers, 2007). However, the majority of studies focused on object concepts alone, ignoring actions. The few that considered actions suggest that the temporal poles are not critical for their processing (e.g. Kemmerer et al., 2010). In this case series, we investigated the neural substrates of linking object and action concepts to lexical labels by correlating the volume of defined regions of interest with behavioral performance on picture-word verification and picture naming tasks of individuals with primary progressive aphasia (PPA). PPA is a neurodegenerative condition with heterogeneous neuropathological causes, characterized by increasing language deficits for at least two years in the face of relatively intact cognitive function in other domains (Gorno-Tempini et al., 2011). This population displays appropriate heterogeneity of performance and focal atrophy for investigating the neural substrates involved in lexical semantic processing of objects and actions. Method. Twenty-one individuals with PPA participated in behavioral assessment within six months of high resolution anatomical MRI scans. Behavioral assessments consisted of four tasks: picture-word verification and picture naming of objects and actions. Performance on these assessments was correlated with brain volume measured using atlas-based analysis in twenty regions of interest that are commonly atrophied in PPA and implicated in language processing. Results. Impaired performance for all four tasks significantly correlated with atrophy in the right superior temporal pole, left anterior middle temporal gyrus, and left fusiform gyrus. No regions were identified in which volume correlated with performance for both object tasks but neither action task or vice versa. Conclusions. Our results suggest that the right superior temporal pole is involved in linking both object and action concepts to their respective lexical items. This contradicts the hypothesis that the temporal poles are implicated in processing of object but not action concepts. We found several other regions that are related to lexical semantic processing as well: the left anterior middle temporal gyrus and left fusiform gyrus, consistent with earlier work implicating the left inferior and anterior temporal lobe in lexical processing. We were not able to identify areas that were differentially involved in object vs. action processing. Differences may be masked due to the fact that static depictions of actions necessarily involve objects (e.g. a picture of “write” involves a pen and piece of paper).