MRI based detection of time-of-day variation in white matter microstructure - are more complex diffusion models better?

Diffusion Weighted Imaging (DWI) is a MRI technique probing the diffusion of water molecules in the human body. Its ability to measure structural and functional changes beyond the nominal image resolution level, has made it possible to quantify tissue microstructure and map biophysical parameters non-invasively. Particularly DWI has proved to be an effective method for mapping the neuronal network, i.e the white matter (WM) tracts in the brain. Aspirations to quantify brain microstructure has led to a wide range of diffusion models trying to accurately represent the biological parameters. Even experienced MR-researches lack full insight in this material and there is no common understanding for which model-based approach that sufficiently provides meaningful biological insight and at the same time justifies the large amount of data required. Two well-established and robust diffusion models are Diffusion Tensor Imaging (DTI) and Diffusion Kurtosis Imaging (DKI). DTI is based upon the assumption that water molecules diffuse with a Gaussian probability distribution (free diffusion). The DKI model is an expansion of the DTI model with an addition "kurtosis term", quantifying the amount of non-Gaussian diffusion. Both models provide maps of the standard diffusion metrics, namely: fractional anisotropy (FA), mean diffusivity (MD), axial diffusivity (AD) and radial diffusivity (AD). These indices can be related to microstructural properties affecting diffusion, thus serve as biomarkers for both normal and pathological processes. This project was based upon a larger study where a DKI analysis revealed diffusion related WM changes from morning to evening in healthy young volunteers. There were two objectives in this theses, 1) to investigate systematic differences in DTI metrics obtained with the DTI model in comparison to the DKI model and 2) to investigate whether similar morning-evening changes in diffusion metrics are obtained with the two models. Two DTI analyses were performed; firstly, investigating the quantitative differences between the DKI-derived and DTI-derived diffusion indices, secondly, investigating morning-to-evening changes by replicating the former study with a V CONTENTS conventional DTI analysis. It was found that both models are able to detect morning-to-evening changes i diffusion indices. However, DKI provided additional significant findings and the results suggests that the morning-to-evening variations in WM microstructure are associated with intra- and extra-cellular changes. Thus it is considered that the improved performance of DKI justifies the need for a larger set of data and a more sophisticated acquisition scheme