Registration of medical images

Encadrants : Isabelle Bloch (LIP6), Giammarco La Barbera (Hôpital Necker, giammarco.labarbera93@gmail.com)
Disponible : NON
Spécialité : IMA/DIGIT
Nombre d'étudiants : 1
Description :

In medical imaging, very often different types of images are acquired for the same patient, in order to gather complementary information. This is typically the case in magnetic resonance imaging (MRI) where the type of information carried by the images depends on the acquisition parameters. In this project, anatomical and diffusion images will be considered, since they are very useful, for instance for surgical planing. To combine both types of information, the images have to be registered. Registration is the process of aligning two or more images by finding an optimal transformation that makes the main features of each input images align. The aim of this project is to develop and compare several registration methods, using different types of transformations (rigid, affine…), and apply them to MRI images of the pelvis.

Pré-requis : Image analysis and machine learning
Travail demandé : 1. Literature review.
2. Development of image registration methods, with different types of transformations, comparison and application.