Inpainting on diffusion weighted magnetic resonance images. An application to
tractography of multiple sclerosis patients

Vicente Sola, Alex

Inpainting on diffusion weighted magnetic resonance images. An application to tractography of multiple sclerosis patients

License

Citation

Vicente Sola, A. Inpainting on diffusion weighted magnetic resonance images. An application to tractography of multiple sclerosis patients. 2017. handle: http://hdl.handle.net/10230/32913

This citation was generated automatically.

Abstract

Diffusion weighted magnetic resonance imaging is a key modality in medical imaging for neuroscience, which allows to deduce the fiber tracts that conform the brain’s white matter. With this technique, displacement of water molecules in a specific spatial direction is measured, giving information of the shapes depending on the predominant diffusion directions. When this technique is used on images from multiple sclerosis patients there are some areas where water diffusion is equally distributed due to the illness’ effect, losing the necessary information to reconstructing the fiber tracts. In this thesis, we propose a variational method based on the Laplace partial differential equation to estimate the right information on those damaged areas allowing the fiber tract reconstruction (tractography). Finally, we also present experiments over synthetic and real images.