Beyond Regional Specialization — Using Individualized Networks to Characterize Cortical Brain Function During Emotion Processing

Abstract

The human brain is the ultimate translator – taking in basic physical properties, like light and sound, and transforming them into complex human capacities. This process involves billions of neural transmissions across distributed, specialized brain systems. Recent advances in neuroimaging have allowed researchers to map these systems into coordinated yet discrete functional networks in the cortex, using an approach called functional connectivity analysis. These cortical networks support higher-order cognitive functions, such as recognizing emotion in faces, by integrating activity across specialized regions. Understanding these networks is key to cognitive neuroscience. While functional connectivity analysis has been vital for mapping large-scale networks, traditional group-averaged approaches have masked individual variability in network topography, leading to mischaracterizations of brain function. In response, precision neuroscience has emerged, focusing on the deep sampling of individuals rather than relying on group averaging. Precision approaches better capture individual topography and activity. This dissertation investigates individually defined cortical networks, focusing on how they support emotion processing using a widely adopted emotion processing task. Chapter 1 evaluates the impact of widespread use of group-averaged network atlases (aka. parcellations) in clinical and developmental research for identifying cortical networks and examining the associations of network function with other individual characteristics such as age and cognitive abilities. This chapter highlights the need for alternative approaches that better account for individual differences in network topology. Chapter 2 explores emotion processing using an individualized network parcellation approach. I examine how network-level recruitment during a canonical emotion processing task, originally targeting the amygdala, offers a more comprehensive approach for understanding neural activity patterns during emotional face processing. Chapter 3 investigates whether individualized network parcellations improve the test-retest reliability of neural activity during a canonical emotion processing task, which despite widespread use, exhibits poor reliability. This chapter further explores associations between network function, stress, and psychopathology, examining both individual differences as well as within-person fluctuations in symptoms of psychopathology and stress as potential sources of variability. Collectively, this dissertation highlights the importance of precision neuroscience approaches in studying cortical network function. By moving beyond region-based and group-averaged methods, it offers new insights into how distinct cortical networks, identified at the individual level, contribute to emotion processing. These findings provide a more comprehensive approach for characterizing brain function during emotion processing and underscore the value of precision neuroscience.