Landmark guided spatial normalization of Diffusion Tensor Images in the presence of large deformations
Verma R, Davatzikos C
University of Pennsylvania
We propose a method of correspondence detection in diffusion tensor images by using a unique morphological signature of each voxel which oncorporates the anisotropy, diffusivity and orientation information provided by tensors. This proves to be a key step in enhancing existing spatial normalization routines for diffusion tensor images (DTI), by adding the white matter context to the information provided by other scalar measures computed from DTI or other MR modalities. This is especially useful in the case of registering brain tumor images to a healthy template brain, as is required in large population studies.