Relationship of semantic representations to connectivity-based brain architecture

University Of Washington, Seattle WA

Investigators

Linked publications & trials

Abstract

? DESCRIPTION (provided by applicant): Decades of research provide evidence that the left anterior temporal lobe (ATL) plays a critical role in the retrieval of semantic (conceptual or word meaning) representations during naming. However, the nature of these representations and how they relate to the functional organization of the ATL remains elusive. The goal of this project is to combine newly emergent imaging techniques (multi-voxel pattern analysis fMRI and diffusion MRI-based cortical parcellation) to explicitly investigate the organization of semantic knowledge support within the left ATL. Specifically, we will focus on addressing the prevailing theories that the left ATL 1) supports domain general (i.e. category-independent) representations, 2) supports representations specifically for persons (or social entities), or 3) supports representations for unique entities (e.g. famous faces or landmarks). We hypothesize that there are functional subdivisions within the ATL, with different cortical sectors supporting each of these representations. Furthermore, we hypothesize that the functional subdivisions within the ATL will have distinct patterns of anatomical connectivity patterns (reflecting the nature of information binding), and that connectivity-defined cortical regions are an appropriate level of organization to assess semantic structure. To test our hypotheses, we will analyze data from two event-related fMRI experiments in which participants named entities from several different conceptual categories (animals, famous faces, unique TV characters, and landmarks). We will use diffusion MRI data to parcellate the left ATL into sub-regions based on the homogeneity of their anatomical connectivity profiles and multi-voxel pattern analysis fMRI to interrogate the information content in order to reveal the semantic structure of the representations within each parcel. These approaches have the potential to shed light on the nature of semantic representations and to establish a link between their emergence and underlying connectivity-based brain architecture. The findings and approaches proposed here have significant potential for application to clinical populations with altered semantic representations like degenerative dementia, or schizophrenia, and altered language abilities like aphasia, autism and developmental dyslexia. Moreover, our theoretical and methodological framework offers a means to measure the integrity of semantic knowledge in patients and to track disease progression, and treatment or training responses.