dMRI:Diffusion Magnetic Resonance Imaging as a Window onto Structural Brain Networks and White Matter Microstructure

Diffusion magnetic resonance imaging (dmri) can be used to probe the connectivity and microstructure of human brain tissue non-invasively in vivo. The diffusion-weighted mr signal has sensitivity to micrometer-scale tissue properties averaged over the imaging voxel size, for example, intra-cellular and extra-cellular volumes and the 3d orientations of axonal bundles. It derives its contrast from sensitivity to the bulk displacement of water through what is known as thermal motion, brownian motion, or passive (self-)diffusion. In the brain, cell membranes, organelles, and myelin sheaths create barriers and form biological compartments that constrain the displacement of water molecules, modifying the statistical behavior of bulk diffusion over time. This review therefore focuses on the use of dmri for estimates of the local orientations and estimates of microstructure of fiber tracts, and on an understanding of dmri signal mechanisms and appropriate signal processing and modeling for this purpose. It first discusses basic diffusion mri acquisition principles and diffusion contrast and the constraints the acquisition places on the modeling of diffusion. It then sets out the diffusion tensor model, the most used model in dmri that underlies diffusion tensor imaging, in a way which prepares a discussion of its limitations. The next sections set out advances in dmri beyond dti as focusing on tractography and connectomics, with a need to accurately model spatially complex fiber configurations and on diffusion microstructure, with a need to accurately model restricted diffusion and compartmentalization. Throughout, the emphasis is on a thorough understanding of basic principles and assumptions underlying techniques, as well as their possibilities and limitations for inference of brain connectivity, with a minimum of technical detail and mathematics. This review ends with an outlook on future developments emanating from current trends