Magnetic resonance diffusion tensor analysis and its application to central nervous system

Magnetic resonance diffusion tensor imaging (DTI) has been established as the primary imaging modality for non-invasive characterization of the direction-dependent motion of water molecules in the living tissue. DTI provides each image voxel with a 3x3 tensor matrix that can be processed to probe tissue microstructure and microarchitecture. In this paper, the imaging principle and diffusion tensor calculation are introduced. Examples are given to describe image-processing techniques for DTI visualization, fiber tracking, and applications to central nervous system.