Levitis, Elizabeth; (2025) Integrating Multimodal Neuroimaging and Computational Approaches to Investigate Atypical Brain Development in Psychiatric and Neurodevelopmental Disorders. Doctoral thesis (Ph.D), UCL (University College London).

Text EL_UCL_PhD_Thesis_Corrections_Final.pdf - Accepted Version Access restricted to UCL open access staff until 1 July 2026. Download (36MB)

Abstract

Studying atypical brain development in psychiatric and neurodevelopmental disorders is complex due to the variability in clinical manifestations, neuroimaging characteristics, and genetic influences. This complexity often leads to poorly defined diagnostic categories, making traditional diagnostic approaches and supervised learning methods insufficient. To address this, my PhD research employs multimodal neuroimaging and advanced computational techniques to investigate brain abnormalities associated with these disorders. This thesis introduces a comprehensive approach to understanding these abnormalities by leveraging multimodal neuroimaging datasets. The first project applies a ”genetics-first” framework to explore brain changes in genetically defined disorders, such as XXY, XYY, and Down Syndrome. By profiling morphometric, microstructural, and functional brain alterations, this project aims to identify both common and distinct patterns of brain vulnerability associated with these genetic conditions. The second project extends this approach to idiopathic neurodevelopmental disorders using the Adolescent Brain Cognitive Development (ABCD) dataset. This project employs deep-learning based normative modeling to first assess how neuroimaging measures deviate among individuals with varying degrees of psychopathology. By focusing on normative deviations, the research aims to move away from group-level comparisons to identify individual-level deviations from normative ranges of brain structural and functional measures. This approach enables the identification of transdiagnostic neurobiological substrates that cut across traditional diagnostic boundaries. Furthermore, the approach is flexible enough to be applied to the aforementioned genetically defined disorders, enabling the exploration of potentially convergent patterns of extreme multimodal deviations in both behaviorally and genetically defined disorders. Overall, this thesis contributes to a deeper understanding of brain development and psychiatric risk by integrating genetic, neuroimaging, and normative modeling approaches.

Download activity - last month

Export as XMLCSVJSON

Download activity - last 12 months

Export as XMLCSVJSON

Downloads by country - last 12 months

Export as CSVXMLJSON

Archive Staff Only

View Item