A technique used to visualize elements with directional structure, such as bundles of myelinated axons, in the brain. The technique derives its name from the fact that a tensor is computed for each voxel, where a tensor is a mathematical array describing the orientation and magnitude of diffusion in all three spatial dimensions (i.e., the diffusion coefficient). In conventional MRI, what in essence is provided are images of free water, which differs in concentration according to tissue type (e.g., white matter consists of about 70% water, gray matter 80%, and cerebrospinal fluid 99%). The contrasts in tissue types observable with structural MRI derive from such differences in water content. Diffusion tensor imaging (DTI) goes further than this through measuring differences with which water molecules move within a tissue type, as well as the amount and orientation of their diffusion, particularly in the case of white matter. Moreover, DTI produces multivariate (i.e., tensor-valued) images, whereas most MRI methods deliver univariate (i.e., scalar) images such as T1 or T2 maps. Consequently, DTI provides high-resolution images of white matter fiber tracts, assuming appropriate data collection and processing techniques, and thus the possibility of detecting their degradation as, for example, in the case of per ventricular leucomalacia and related white matter lesions that are relatively common in very preterm infants. An important constraint on DFI at present is that it cannot represent multiple bundle fibers with different orientations. Another is that it lacks the ability to distinguish ‘kissing’ from ‘crossing’ fibers. With regard to kissing fibers, the two fibers merge into a single voxel before they cross over each other, and the direction detected by DTI could be the same in each case. Finally, recent research has demonstrated that DTI measurements correlate better with measures of executive function than conventional MRI measures.
See Brain (neuro-) imaging, Cerebrospinal fluid (CSF), Diffusion anisotropy, Diffusion tensor magnetic resonance imaging, Executive function (EF), Gray matter, Myelin, Myelination, T1 signal-weighted (T1-W) technique, T2 signal-weighted (T2-W) technique, Periventricular leucomalacia (PVL), Preterm infant, Structural magnetic resonance imaging (sMRI), Voxel, White matter