Characterizing non-Gaussian diffusion by using generalized diffusion tensors.

摘要: Diffusion tensor imaging (DTI) is known to have a limited capability of resolving multiple fiber orientations within one voxel. This mainly because the probability density function (PDF) for random spin displacement non-Gaussian in confining environment biological tissues and, thus, modeling self-diffusion by second-order breaks down. The statistical property diffusion process characterized via higher-order (HOT) coefficients reconstructing PDF displacement. Those HOT can be determined combining series complex diffusion-weighted measurements. signal equation an MR experiment was investigated theoretically generalizing Fick's law partial differential (PDE) obtained Kramers-Moyal expansion. A relationship has been derived between PDE and cumulants Monte-Carlo simulations restricted with different geometrical shapes were performed, strengths weaknesses both established analysis techniques investigated. generalized formalism capable accurately underlying structures, which neither conventional DTI nor at high angular resolution (HARD) capable. method helps illuminate some restrictions that are characteristic these other methods. Furthermore, direct q-space also established.