A novel mechanism of contrast in MRI: pseudo super-diffusion of water molecules unveils microstructural details in biological tissues

The goal of this work is to investigate the properties of the contrast provided by Anomalous Diffusion (AD) γ-imaging technique and to test its potential in detecting tissue microstructure. The collateral purpose is to implement this technique by optimizing data acquisition and data processing, with the long term perspective of adoption in massive in vitro, in vivo and clinical studies. The AD γ-imaging technique is a particular kind of Diffusion Weighted- Magnetic Resonance Imaging (DW-MRI). It represents a refinement of conventionally used DW-MRI methods, sharing with them the advantage of being non invasive, since it uses water as an endogenous contrast agent. Besides, it is more suitable to the study of complex tissues, because it is based on a theoretical model that overcomes the simplistic Gaussian assumption. While the Gaussian assumption predicates the linearity between the average molecular displacement of water and the diffusing time, as in case of diffusion in isotropic, homogeneous and infinite environments, a number of experiments performed in vitro and in vivo on both animals and humans showed an anomalous behavior of water molecules, with a non linear relation between the distance travelled and the elapsed time. In particular, the γ-parameter quantifies water pseudo super-diffusion, a peculiarity due to the fact that water diffusion occurs in multi-compartments and it is probed by means of MRI. In fact, a restricted diffusion is rather predicted for water diffusing in biological tissues. Recently, the trick that allows to make the traditional DW-MRI acquisition sequence suitable for pseudo super-diffusion quantification has been unveiled, and in short it consists in performing DW experiments varying the diffusion gradient strengths, at a constant diffusive time. The γ-parameter is extracted by fitting DW-data to a stretched-exponential function. Finally, probing water diffusion in different directions allows to reconstruct a γ-tensor, with scalar invariants that quantify the entity of AD and its anisotropy in a given volume element. In vitro results on inert materials revealed that γ correlates with internal gradients arising from magnetic susceptibility differences (Δ) between neighboring compartments, and that it reflects the multi-compartmentalization of the space explored by diffusing molecules. Furthermore, values of γ compatible with a description of super-diffusive motion were found. This anomaly can be explained considering that the presence of Δ induce an additional attenuation to the signal, simulating a pseudo super-diffusion. Finally, In vivo results on human brain showed that γ is more effective in discriminating among different brain regions compared to conventional DWMRI parameters. These studies suggest that the contrast provided by AD γ-imaging is influenced by an interplay of two factors, Δ -effects on one hand, multicompartmentalization on the other hand, through which γ could reflect tissue microstructure. With the aim to shed some light on this issue I performed AD γ-imaging in excised mouse spinal cord (MSC) at 9.4 T and healthy human brain at 3.0 T. The adoption of MSC was motivated by its current use in studies of demyelination due to an induced pathology that mimics Multiple Sclerosis alterations, and by its simplified geometry. I acquired DW-data with parameters optimized for the particular system chosen: the MSC was scanned along 3 orthogonal directions, thus an apparent γ was derived; for the in vivo studies I used more directions and I extracted a γ-tensor. I found that γ and its anisotropy reflected the microstructure of spinal cord tracts (such as the axon diameters and the axonal density). I investigated both in MSC and human brain the relation between γ and the rate of relaxation (R2*), a parameter well-known to reflect Δ, and found significant linear correlations. Because of this γ was able to differentiate white matter regions on the basis of their spatial orientation, and gray matter regions on the basis of their intrinsic iron content in human brain imaged at 3.0 T. These results suggest that AD γ-imaging could be an alternative or complementary technique to DW-MRI in the field of neuroscience. Indeed it could be useful for the assessment of the bulk susceptibility inhomogeneity, which reflects iron deposition, the hallmark of several neurodegenerative diseases. The part of this thesis work concerning the in vivo experiment in human brain gave rise to a paper published on NeuroImage, a relevant scientific journal in the field of MRI applied to brain investigation.