|dc.description.abstract||The human brain is an incredibly complex system, consisting of billions of neurons, interacting over trillions of synaptic connections. These interactions give us consciousness and intelligence, control our core functions, and allow us to manipulate our environment. Functional magnetic resonance imaging, or fMRI, allows us to measure blood flow throughout the brain over time, from which we may infer neural activity.
fMRI data have traditionally been analysed using model based techniques, by searching for regions of the brain which have been activated in response to an experimental stimulus. Model free analysis methods are becoming more prevalent however, allowing us to explore not only those regions of the brain which were activated by a stimulus, but also the ways in which different regions interact with each other whilst performing a task, or at rest. These interactions are referred to as functional connectivity, and when calculated between every pair of regions in the brain, form a functional network, the properties of which may be explored using techniques derived from graph theory.
Some success has been achieved in the use of fMRI and functional connectivity analysis methods to explore the ways in which brain function changes with advancing age, and with the presence of neurological disorders such as Alzheimer's Disease (AD). However, the ways in which the properties of functional networks change with age, and the ability of functional network analysis to distinguish between healthy aging and aging in the presence of AD, are still unknown.
In this PhD Thesis, we explore the broad hypothesis that the neurodegenerative changes which are associated with aging and in particular with AD must, in some way, have a measurable effect upon functional connectivity, and upon the properties of functional networks, derived from fMRI data. We attempt to comprehensively explore differences in the properties of functional networks reflecting both task related and resting state brain activity and interpret them in the context of known facts about neurodegenerative changes associated with aging and AD. At the same time, we shall explore the effects of some methodological approaches to functional network analysis.||