hide
Free keywords:
-
Abstract:
The spinal cord—as the principal pathway between the brain and peripheral nervous system—plays an important role in sensation, autonomic function, and motor control. Over the last decades, functional magnetic resonance imaging (fMRI) revealed insights into human central nervous system function. However, neuroimaging studies focused more on the brain compared to the spinal cord, possibly owing to methodological challenges of spinal cord fMRI. These challenges are due to several reasons such as the small diameter of the spinal cord, and changes in the magnetic field caused by differences in the magnetic susceptibility between various tissues, and the impact of the physiological noise.
In this thesis, I am aiming to address part of the methodological challenges in fMRI of the spinal cord as well as some of the knowledge gaps in the functional neuroanatomy of the spinal cord that partly arise due to these challenges. The first study aimed to address some of the aforementioned challenges and optimize spinal cord fMRI by developing an automated slice-specific z-shim protocol to compensate for the effect of magnetic field inhomogeneities. First, the relevance of slice-specific z-shimming for spinal cord fMRI was demonstrated (Empirical Investigation 1). To automate the z-shim procedure, two distinct methods were used (Empirical Investigations 2 and 3) one is based on a field map (Empirical Investigation 2), and the other one is based on a reference echo-planar imaging (EPI) measurement (Empirical Investigation 3). These were compared against the manual selection of z-shim values (currently the gold standard) and no z-shimming (serving as a baseline). Finally, all three methods (two automated methods and the manual method) were compared (Empirical Investigation 4). The results suggested that both automated methods were highly beneficial and improved the signal characteristics that are directly relevant to fMRI.
Another challenge of spinal cord neuroimaging is the proneness to physiological noise (i.e., respiratory and cardiac processes) which influence fMRI measurements. Therefore, using the acquisition methods that were developed in the first study, the aim of the second project was to investigate the reliability of the intrinsic spinal networks. Specifically, we investigated test-retest reliability (Empirical Investigation 5) of spinal cord resting-state networks, and how various noise sources impact their reliability (Empirical Investigation 6). The results demonstrated a fair to good reliability in sensory and motor networks and showed how structured physiological noise sources impact the functional connectivity measurements of resting-state spinal networks.
