Title: Graph-based deformable registration: slice-to-volume mapping and context specific methods

Abstract: Image registration methods, which aim at aligning two or more images into one coordinate system, are among the oldest and most widely used algorithms in computer vision. A particular type of registration algorithm, known as graph-based deformable registration, has become popular during the last decade given its robustness, scalability, efficiency and theoretical simplicity. In this thesis, we extend this flexible framework to new scenarios, and propose novel methodological contributions.

We start by formulating, within the graph-based deformable registration framework, the challenging slice-to-volume registration problem. We introduce a scalable, modular and flexible formulation accommodating low-rank and high order terms, which simultaneously selects the plane and estimates the in-plane deformation through a single shot optimization approach. The other two contributions included in this thesis are related to how semantic information can be encompassed within the registration process. Currently, most of the methods rely on a single metric function explaining the similarity between the source and target images. We argue that incorporating semantic information to guide the registration process will further improve the accuracy of the results, particularly in the presence of semantic labels making the registration a domain specific problem.