ASFNR Guidelines for Clinical Application of Diffusion Tensor Imaging

1. Definitions: a. Diffusion Tensor Imaging (DTI) refers to the acquisition and reconstruction of MR images with diffusion-encoding gradients applied in a minimum of 6 non-collinear directions, followed by voxel-wise calculation of the tensor representation of a 3D Gaussian model of the diffusion profile and subsequent mapping of any number of tensor parameters. Most commonly these would include the scalar parameters mean diffusivity (the trace of the diffusion tensor, equivalent to the directionally averaged apparent diffusion coefficient [ADC] common to conventional diffusion-weighted imaging) and fractional anisotropy [FA] (one of several available measures of the degree to which the diffusion tensor deviates from a spherical shape, i.e. the degree to which the diffusion profile is directionally dependent). b. Directional color mapping refers to the additional DTI post-processing step in which a color scheme is used to encode the direction of the major eigenvector (i.e. the direction of maximum diffusivity). The most widely used color scheme assigns the colors red to left-right, green to anterior-posterior, and blue to superior-inferior directions with proportionate combinations of these colors assigned to directions in between these principal directions. The brightness or intensity of color is typically used to indicate the FA (ranging from 0 to 1); such maps are known as " Directional FA " or " Color FA " maps. c. Diffusion tensor tractography, also known as fiber tracking, refers to the additional DTI post-processing step in which a predefined algorithm is used to generate computer graphical representations of anatomical fiber bundles as follows: Starting from selected seed points, multiple trajectories or streamlines are traced though the imaged volume by iteratively stepping in directions determined by the tensor shapes and orientations encountered as the algorithm proceeds. Tractography algorithms may be constrained by predetermined stopping criteria in order to terminate trajectories likely to be erroneous; such criteria typically include thresholds for minimum anisotropy and maximum turning angle.