Anisotropy Creases Delineate White Matter Structure in Diffusion Tensor MRI

Abstract

Abstract. Current methods for extracting models of white matter architecture from diffusion tensor MRI are generally based on fiber tractography. For some purposes a compelling alternative may be found in analyzing the first and second derivatives of diffusion anisotropy. Anisotropy creases are ridges and valleys of locally extremal anisotropy, where the gradient of anisotropy is orthogonal to one or more eigenvectors of its Hessian. We propose that anisotropy creases provide a basis for extracting a skeleton of white matter pathways, in that ridges of anisotropy coincide with interiors of fiber tracts, and valleys of anisotropy coincide with the interfaces between adjacent but distinctly oriented tracts. We describe a crease extraction algorithm that generates high-quality polygonal models of crease surfaces, then demonstrate the method on a measured diffusion tensor dataset, and visualize the result in combination with tractography to confirm its anatomic relevance.

Delineating white matter structure in diffusion tensor MRI with anisotropy creases

“…We believe that removing the smallest components increases the anatomic significance of the crease surface results by removing the pieces that are most apt to vary as a function of image noise and scale selection. This geometric postprocessing is a significant improvement over the method presented in our previous work (Kindlmann et al, 2006).…”

“…Given the ubiquity of FA as a quantitative variable in the diffusion tensor literature, we start by detecting creases in FA, and term these anisotropy creases (Kindlmann et al, 2006). We propose that the ridge surfaces and ridge lines of FA coincide with the interiors of white matter fiber tracts, and that valley surfaces of anisotropy delineate the interfaces between fiber tracts that are adjacent but distinctly oriented (such as between the corpus callosum and the cingulum bundles).…”

Biomedical Visualization

“…In collaboration with G. Kindlmann and C.-F. Westin from Harvard Medical School we have recently proposed a compelling alternative to that approach [114] which does not require the potentially errorprone integration of fiber tracts in noisy tensor data. Leveraging concepts from the field of computer vision we have devised a method that computes Fig.…”

“…From [114] within biomedical computing. For this reason, visualization researchers have concentrated on developing effective ways to visualize large-scale computational fields.…”

Interactive Volume Rendering of Diffusion Tensor Data

“…While local direction information does not have medical importance as, e.g., neural fibers are defined as integral curves in a certain volume, in many areas of the brain, neighboring fiber bundles have different directions but their anisotropy values only change marginally. Kindlmann et al [KTW06] showed that FA ceases can be found after pre-processing the data set but these methods cannot be easily transferred to volume rendering. Nevertheless, these areas can be found easily by looking at the local change of direction.…”