Por favor, use este identificador para citar o enlazar este ítem:
http://cimat.repositorioinstitucional.mx/jspui/handle/1008/737| ROBUST ALGORITHMS FOR THE ANALYSIS OF THE BRAIN STRUCTURE USING DIFUSSION | |
| ANGEL RAMON ARANDA CAMPOS | |
| Acceso Abierto | |
| Atribución-NoComercial | |
| AXONES | |
| Ones of the most challenging goals in neuroimaging are to infer axonal connectivity patterns and tissue properties on extit{in vivo} brains through water diffusion estimation in cerebral tissue. To this aim, the Diffusion--Weighted Magnetic Resonance Imaging technique is used. This technique allows one to estimate the preferred orientations of water diffusion in brains. Such a diffusion, in the white matter case, is constrained by the axonal structure. That information is very useful in neuroscience research due to the changes in the connectivity patterns associated with neurological disorders and, in general, associated with brain development. There are two important stages to infer axonal connectivity: 1) to analyze the intra--voxel Diffusion--Weighted Magnetic Resonance signals for recovering the local fiber structure, and 2) to apply, over the recovered intra--voxel information, a tractography technique that estimates neuronal tract pathways to determine both global structure and connectivity in human brains. In this work, we propose robust strategies to estimate the intra--voxel fiber structure and the global fiber bundle trajectory. Thus, we present a sparse and adaptive diffusion dictionary based method to estimate the local intra--voxel axonal fiber populations on extit{in vivo} brain data. Our proposal overcomes the following limitations of the diffusion dictionary--based methods: the limited angular resolution and the incapability to estimate tissue properties for different fiber bundles. We propose to iteratively re--estimate the orientations and the diffusivity profile of the atoms independently at each voxel by solving surrogate models based on mathematical simplifications. As a result, we improve the fitting of the Diffusion--Weighted Magnetic Resonance signal. We also propose a new method to estimate axonal fiber pathways from the intra--voxel information. Our method uses the multiple local orientation information for leading stochastic walks of particles. These stochastic particles are modeled with mass, hence, they are subject to gravitational and inertial forces. As result, we obtain smooth, filtered and compact bundle trajectories. This gravitational interaction can be seen as a flocking behavior among particles that promotes better and robust axon fiber estimations because the particles use collective information to move. Both proposals are evaluated on realistic numeric diffusion phantoms, a physical diffusion phantom and on ext | |
| 27-07-2016 | |
| Trabajo de grado, doctorado | |
| CONVERTIDORES ANALÓGICO-DIGITALES | |
| Versión aceptada | |
| acceptedVersion - Versión aceptada | |
| Aparece en las colecciones: | Tesis del CIMAT |
Cargar archivos:
| Fichero | Descripción | Tamaño | Formato | |
|---|---|---|---|---|
| TE 660.pdf | 30.1 MB | Adobe PDF | Visualizar/Abrir |