Doctor of Philosophy
Menon, Ravi S.
Classical models of brain organization have often considered the brain to be made up of a mosaic of patches that are demarcated by discrete boundaries, often defined histologically. In contrast, emerging views have pointed towards an alternative paradigm – referred to as gradients – by conceptualizing brain organization as sets of organizational axes that characterizes spatial variation of differing connectivity principles over the extent of a region. Such organizational axes provide a well-suited framework for elucidating underpinnings of brain connectivity and has garnered widespread attention across various domains of neuroimaging. This work seeks to explore various applications of gradient estimation techniques, in combination with resting-state functional connectivity data, across the fields of basic, comparative, and clinical neuroscience.
First, gradient estimation was performed on resting-state functional connectivity (RSFC) patterns of the primary somatosensory cortex to unveil a secondary organizational axis that spans the region’s anterior-posterior axis, akin to circuitry fundamental to sensory cortical information processing. Second, gradient techniques were used in a cross-species comparison study to unify connectivity principles of humans and marmosets by mapping them simultaneously onto a set of organizational axes. In doing so, this provided a systematic framework to compare the functional architecture of both species, facilitating novel insight of a well-integrated default-mode network in humans, compared to marmosets. Third, connectivity gradients, along with a myriad of other resting-state fMRI features were used to explore the implications of focal lesion pathophysiology on functional organization of the thalamus in individuals with Multiple Sclerosis. A lack of focal changes to resting-state related features was observed suggesting the limited role of focal thalamic lesions to functional organization in MS.
Together, these different avenues of research highlight the capacity for a gradient-centric view in neuroimaging to provide profound insights into brain organization, and its utility across the applications of basic, comparative, and clinical neuroscience.
Summary for Lay Audience
Brain structure maps have typically been described by a set of areas separated by borders, much like countries on the globe. In this manner, brain mapping may overlook more subtle details of the brain – analogous to detailing a country’s cities and neighbourhoods – that are equally important for understanding how brain functions arise. Towards this latter goal, the notion of gradients has been proposed as a way to reveal more nuanced details of the brain. Gradients in brain organization have emerged as a powerful approach for studying different fields of imaging neuroscience. In this thesis, gradients are broadly applied to study finer-grained connectional principles of the brain in domains of basic, comparative, and clinical neuroscience.
The first chapter applies gradients to the primary somatosensory cortex (S1) of the brain, revealing detailed connections that reflect the classical notion of information processing within this brain region. Additionally, this connectivity gradient demonstrated excellent correspondence to S1’s structural properties, which previously have gone underappreciated. Collectively, this insight provides a way to link local brain structure and connectivity properties within S1 to one another, providing a principled approach for studying their interplay in clinical populations.
The second chapter uses gradients to systematically compare the human and marmoset brain. In doing so, differences in default-mode network brain organization were revealed. Given the importance of marmosets as a preclinical animal model, this difference may provide fundamental insights towards the limitations of the marmoset as an animal model for studying cognitive function, for which the default-mode network is thought to play a critical role in.
The third chapter uses gradients (in addition to other approaches) to provide a simplified view of neuronal and connectivity principles in the human thalamus following neural degeneration – specifically from Multiple Sclerosis lesions. No obvious qualitative disruptions of thalamic organization were observed when using a myriad of approaches. This may suggest that lesions in the thalamus do not play a substantial role in mediating thalamic reorganization in Multiple Sclerosis.
Overall, this thesis demonstrates the wide use of gradients to study brain organization across various applications of imaging neuroscience. This demonstrates the collective versatility and receptiveness of gradients as a general investigative tool for studying brain mapping.
Ngo, Geoffrey, "Applications of Gradient Representations in Resting-State fMRI" (2022). Electronic Thesis and Dissertation Repository. 8638.