Interpretation of BOLD events using fMRI at 7 Tesla

Abstract

Functional Magnetic Resonance Imaging (MRI) acquired at ultra‐high magnetic field (7 Tesla) provides increased blood oxygen level dependent (BOLD) sensitivity and contrast‐to‐noise ratio. Sparse Paradigm Free Mapping (SPFM) is an fMRI data analysis method recently developed to detect sparse events related to brain activity without prior timing information. This thesis designs and validates methods that interpret BOLD events detected using SPFM at 7 Tesla. Firstly, this thesis validates the use of temporal Independent Component Analysis to decompose SPFM outputs into temporally‐independent BOLD events with spatially overlapping activation maps using a task‐based paradigm. Activation maps of these components can then be used for decoding purposes. This thesis also proposes a method to decode BOLD events by relating their spatial activation maps to a meta‐analysis of previous fMRI studies. A decoding score was derived to relate SPFM outputs to Activation Likelihood Estimation (ALE), which is a coordinate‐based meta‐analysis method. The proposed method was validated against motor task paradigms to decode task‐based and spontaneous motor events in the Sensorimotor Network (SMN). Finally, to investigate the Default Mode Network (DMN), which is a cognitive resting‐state network that has overlapping functions, DMN BOLD events during motor tasks and resting‐state events were investigated. It is shown that there is a small percentage signal change that is close to baseline in DMN nodes when spontaneous events occur in the SMN. In addition, the Precuneus/Post‐Cingulate Cortex (pC/PCC) also co‐activated with the nodes of Dorsal Attention Network (DAN) and Supplementary Motor Area (SMA), further supporting the theory of the DMN being functionally heterogeneous and suggesting a dynamic role of pC/PCC as a functional hub.

Date of Award	1 Jul 2017
Original language	English
Awarding Institution	Univerisity of Nottingham
Supervisor	Yaping Zhang (Supervisor), David Siu-Yeung Cho (Supervisor), Susan Francis (Supervisor) & Penny Gowland (Supervisor)