Association between resting-state functional connectivity, glucose metabolism and task-related activity of neural networks

Simon-Vermot, Lee (2018): Association between resting-state functional connectivity, glucose metabolism and task-related activity of neural networks. Dissertation, LMU München: Graduate School of Systemic Neurosciences (GSN)

Abstract

The brain is organized into several large-scale functional networks. Such networks are primarily characterized by intrinsic functional connectivity, i.e. temporally synchronous activity between the different brain regions of a network. The functional connectivity of networks can be identified via functional MRI during resting state, i.e. without engaging the subject in a particular task. Resting-state fMRI is thus less demanding on the subject and therefore of particular interest from a clinical point of view to detect alterations in brain function. Applied to neurodegenerative disease including Alzheimer’s disease, resting-state fMRI has shown alterations in several resting-state networks, suggesting that basic network function is altered in AD. However, the interpretation of alterations in resting-state fMRI connectivity is inherently limited since no cognitive states are explicitly expressed during fMRI. In this regard, we aimed to elucidate how resting-state fMRI connectivity relates to 1) cognition-related brain activity and 2) markers of pathologically altered brain function in AD. In order to understand at a more basic level the association between resting-state and task-related fMRI, we first examined, in a group of elderly healthy subjects, the association between functional connectivity of major networks assessed during resting-state fMRI with those acquired during memory-task related fMRI, in the same individuals. Secondly, in order to assess whether alterations in AD are associated with already well-established markers of pathological brain function in AD, we compared resting-state fMRI functional network connectivity with that in FDG-PET metabolism in AD. Project 1: We investigated the association between functional connectivity acquired during rest and the level of activation obtained during an episodic memory task that included the encoding and forced-choice recognition of face-name pairs in elderly cognitively normal subjects. Independent component analysis (ICA) was used to identify major resting-state networks in the brain. Next, we applied ICA to the task-fMRI data to determine the components (networks) that were significantly associated with the task regressors of successful vs unsuccessful learning or recognition trials. Spatial correlation analysis between the resulting extracted resting-state and task-related fMRI components showed a spatial match in several components such as medial temporal lobe centered components and posterior components. However, apart from the spatial correspondence, the level of resting state functional connectivity did not predict the level of task-related functional connectivity in spatially matching components. Together these results suggested that particular resting-state networks are activated during a memory task, however, the level of baseline connectivity does not predict to what extent a network becomes activated during a task. Future studies may assess whether pathological resting-state connectivity predicts altered task-related connectivity in the same networks in AD. Project 2: We examined the association between resting-state fMRI functional connectivity within major functional networks and FDG-PET metabolism in those networks, assessed in elderly healthy controls, subjects with prodromal AD (mild cognitive impairment and amyloid PET biomarker confirmed AD etiology) and AD dementia. We found that FDG-PET was generally reduced in all networks in the course of AD. The main finding was that lower network functional connectivity was associated with lower FDG-PET uptake in the Default mode network and fronto-parietal attention network across the whole group and specifically in prodromal AD, suggesting that both modalities are associated in higher networks affected in the course of AD. These results provide insightful comprehension of the hypometabolism patterns that are typically found in AD.

• Dokumententyp: Dissertationen (Dissertation, LMU München)
• Themengebiete: 500 Naturwissenschaften und Mathematik ; 500 Naturwissenschaften und Mathematik > 570 Biowissenschaften, Biologie
• Fakultäten: Graduate School of Systemic Neurosciences (GSN)
• Sprache der Hochschulschrift: Englisch
• Datum der mündlichen Prüfung: 4. Mai 2018
• 1. Bericht­erstatter:in: Ewers, Michael