Por favor, use este identificador para citar o enlazar este ítem:
http://cimat.repositorioinstitucional.mx/jspui/handle/1008/828| ROBUST ESTIMATION OF THE BUNDLE -WISE AXON DIFFUSIVITY PROFILES FROM DW- MRI APPLICATIONS | |
| RICARDO CORONADO LEIJA | |
| Acceso Abierto | |
| Atribución-NoComercial | |
| Difusión MRI White Matter Microstructure | |
| Diffusion-weighted (DW) magnetic resonance imaging (MRI) is an imaging modality which is sensitive to the microscopic Brownian motion of water molecules (i.e. diffusion). Water diffusion in living tissues is affected by their cellular organization. In particular, in the brain white matter, the coherent organization in packed bundles of the neuronal tissue introduces a directional dependence to the microscopic random movements. Therefore, by measuring water molecular displacements using diffusion MRI, it is possible to infer properties of the local tissue micro-structure and the architecture of the neuronal connections in-vivo and in a non-invasive way. Several methods have been proposed to extract multiple fiber intra-voxel configurations from the DW-MR images. However, many of these methods focus only on the estimation of the number of fiber bundles, the percentage of volume of each bundle within the voxel and their orientations, by assuming constant diffusion profile in all the volume; many other methods focus on the estimation of the diffusion properties by assuming the number of fiber bundles equal to one or by factorizing the orientation information; and few methods incorporate the spatial coherence present in the data in order to improve the estimation of the intra-voxel configuration. Also, there are very few methods that estimate different diffusion properties per bundle, which is important in order to study the axonal degeneration for a single fiber bundle in a crossing fiber region. In this thesis, we present a general framework for the parameter estimation of the multi-compartment models; this allows us to improve the estimation of the number and orientation of the fiber bundles, as well as their bundle-wise diffusion profiles. The proposed methodology is based on four key novel ideas: (i) A Multi-Resolution Discrete-Search determines the orientation of the fiber bundles accurately and naturally constrains the sparsity on the recovered solutions; (ii) the determination of the number of fiber bundles using the F-test combined with a Rician bias correction; (iii) a general framework for the estimation of the axial and radial diffusivity parameters independently for each bundle; and (iv) a Simultaneous Denoising and Fitting procedure that exploits the spatial redundancy of the axon bundles to achieve robustness with respect to noise. A new useful evaluation metric is also proposed, which combines the information of the success rate in the est | |
| 28-08-2017 | |
| Trabajo de grado, doctorado | |
| DISEÑO DE SISTEMAS SENSORES | |
| Versión aceptada | |
| acceptedVersion - Versión aceptada | |
| Aparece en las colecciones: | Tesis del CIMAT |
Cargar archivos:
| Fichero | Descripción | Tamaño | Formato | |
|---|---|---|---|---|
| TE 661.pdf | 19.72 MB | Adobe PDF | Visualizar/Abrir |