Abstract
A generalisable understanding of brain function would enable the more effective delivery of neuromodulatory therapies, but interindividual variability and context-dependent neural processes make such an understanding difficult to achieve. This thesis aimed to characterise context-dependent reconfigurations in brain state topology and dynamics that generalise across participants using dynamic functional connectivity and information theoretic methods that capture properties of the brain’s dynamical, complex nature.
To this end, Chapter 3 systematically evaluates factors that influence the performance of a dynamic functional connectivity technique, Leading Eigenvector Dynamic Analyses (LEiDA), applied in Chapters 4 and 6 of this thesis to identify recurrent brain states. Chapter 4 uses LEiDA to identify the shared brain states among four cognitive tasks, and evaluates the task-dependent differences in brain state dynamics, as well as how these metrics of brain state dynamics relate to cognitive ability. Chapter 5 explores how brain state topology reconfigures during a single task, identifying relationships between condition-dependent reconfigurations in brain state topology and behavioural performance. Chapter 6 investigates whether resting state topology and dynamics are influenced by a noninvasive deep brain stimulation technique, Temporal Interference. While Chapters 4, 5, and 6 present studies that observe condition-dependent reconfigurations in brain state topology and/or dynamics, Chapter 7 proposes that computational modelling might offer a more efficient method of developing a mechanistic and predictive understanding of brain function. Chapter 7 describes common generative models of brain function, how they answer mechanistic or predictive questions about the effects of brain stimulation, and with a series of studies intended to advance modelling work from describing brain function to generating data-driven stimulation hypothesis intended for empirical testing.
In summary, this work provides a compelling case that the brain reconfigures its topology and dynamics in line with endogenous mentation and exogenously influenced conditions. Metrics of these reconfigurations relate to trait characteristics, task performance, and brain stimulation condition, encouraging their use in future work.