Examining the Organization and Functions of Parallel Association Networks within Individuals

Abstract

In our daily lives, we often recall details of our past experiences, consider friends’ feelings, imagine novel scenarios, and flexibly communicate with others. These impressive cognitive abilities – for memory, social reasoning, prospection and language – are hallmarks of our species. And all have been linked to regions of association cortex in the human brain. Association regions also show disproportionate evolutionary expansion and prolonged postnatal development, making them particularly intriguing targets for understanding how human brains support our advanced cognitive abilities. Prior work commonly attributed broad swaths of association cortex to large, multi-functional networks. More recently, pairing advances in neuroimaging resolution with precise within-individual techniques has revealed multiple interwoven networks distributed across association zones. This raises the question: could dissociable networks support distinct higher-order functions? In this dissertation, I explore this question by examining the organization and functions of individually-defined association networks, with particular focus on three distinct networks that differentially support higher-order functional domains. In three central chapters, I describe growing evidence that distributed, parallel networks occupy association zones within individuals and that three of these networks (termed DN-A, DN-B and LANG) differentially support scene construction, social and language comprehension tasks. Chapter 1 examines how parallel, interwoven networks DN-A and DN-B are differentially recruited by episodic projection and theory of mind task contrasts. Chapter 2 explores the functional properties of the LANG network in relation to DN-A, DN-B and other juxtaposed association networks, including selective response to a language comprehension task. Chapter 3 explores network functions by building on existing neuroimaging data with novel measures of trial-level variation, providing evidence of DN-A’s differentiation from nearby networks and role in scene construction processes. Collectively, these chapters illustrate functional heterogeneity between parallel, distributed networks. These findings, paired with the networks’ similar spatial motif, also raise the possibility that the multiple networks originated from a singular archetype. The final Discussion presents the Expansion-Fractionation-Specialization hypothesis to account for these observations: evolutionary expansion of human association cortex may have allowed for a prototype distributed network to fractionate into multiple specialized networks. Human development may recapitulate fractionation and specialization when these abilities emerge.