Page 14 - TI Journal 18-1

8 SALMINEN ET AL.

the fiber due to fiber walls that act as barriers (80). In The majority of work utilizing DTI has focused
these tissues, water diffusion is anisotropic (i.e., varies on white matter integrity, as changes in white matter
along different directions). Water molecules in pure microstructure can be readily delineated using DTI
water do not encounter barriers and travel quickly metrics. White matter primarily consists of myelin-
and equally in all directions (isotropic diffusion). For ated and unmyelinated axon fibers that restrict water
tissue that is generally isotropic (e.g., gray matter and movement in directions perpendicular to the fibers,
CSF), water molecules follow a random pattern of thus increasing anisotropy (56). Damage to axons
motion, and the signal loss is not dependent on the and/or the myelin sheath reduces anisotropy and
direction of the diffusion-encoding gradient. In these increases the rate of diffusion in directions perpendic-
tissues, a single scalar D is sufficient to characterize ular to the fibers, which is a common result of aging
diffusion (26). By contrast, the signal loss in anisotro- and disease (4,56,82). DTI can also be used to evaluate
pic tissue (e.g., white matter) strongly depends on the changes in gray matter microstructure, though the
direction of white matter fibers and the diffusion-en- biological interpretation of gray matter diffusion is
coding gradients, which complicate interpretation of less clear than in white matter due to the high level
DWI contrast. Thus, the directionally-averaged D (or of isotropic diffusion that is evident in normal-ap-
mean diffusivity (MD) image) is often calculated in pearing gray matter (4,28). Increases in gray matter
clinical practice, which is the average of the computed diffusion have been reported in previous studies of
D images obtained from the orthogonal encoding aging (59,68,75,76), yet the mechanisms underlying
directions. Sometimes the average DWI or isotropic these changes have not been fully delineated.
DWI (DWI iso) is also shown, which is the geometric Fractional anisotropy (FA) and MD are traditional
mean of the corresponding DWIs. In both MD and DTI scalar metrics that measure diffusion processes
DWI iso, the effects of white matter fiber direction have by quantifying the degree of directionality of water
been removed. For a more complete description of the diffusion and the directionally-averaged rate of wa-
directional motion of water molecules in anisotropic ter movement within an image voxel, respectively
tissues, a diffusion tensor is required (9,10,19). (15). Damage to cellular microstructure (e.g., axon
degeneration, myelin loss, etc.) alters the movements
DIFFUSION TENSOR IMAGING of water molecules and typically results in increased
Derivation of the diffusion tensor from the DWIs MD and decreased FA (3). Axial diffusivity (AD) and
allows for the quantification of water diffusion in radial diffusivity (RD) are additional DTI metrics
living brain tissue, thus providing information about that measure water diffusion that occurs parallel and
the underlying tissue microstructure. The diffusion perpendicular to axon fibers (60). Increased RD and
tensor characterizes the three-dimensional spread decreased AD are believed to reflect reduced integ-
of diffusing water molecules from a point source rity as a result of demyelination, axon damage, fiber
using three-dimensional ellipsoids. Various biological rarefaction, and/or gliosis (11).
factors influence the shape of the diffusion ellipsoid,
including the microstructural composition of human TECHNICAL LIMITATIONS OF DTI
brain tissue. DTI scan protocols require the applica- Numerous factors influence accuracy and preci-
tion of at least six non-collinear diffusion-encoding sion of DTI data, including imperfections in scanner
gradients and a reference image (typically a b~0 im- hardware (e.g., RF coil, magnetic field gradients),
age) to capture the full extent of directional water selection of operator controlled parameters (e.g.,
mobility in anisotropic tissue voxels (an anisotropic pulse sequence, number of acquisitions, number of
tissue voxel is a voxel that is, for example, 2x2x5mm diffusion-encoding gradients), and patient variables
in dimensions) (19,63). Modern research scan proto- (e.g., motion) (55). As a result, quality control (QC)
cols tend to utilize many more directions (up to 64) of imaging data is critical for obtaining accurate
to increase the sensitivity of the signal and improve DTI measurements. Various QC procedures can be
measurement accuracy (35). implemented to minimize errors from multiple sources,