Contributions to Medical Image Registration
Darko Zikic · mediaTUM – the media and publications repository of the Technical University Munich (Technical University Munich) · 2011
Registration is the process of computing the corresponding points in two images, and constitutes a fundamental step for further processing in many medical applications. Due to the challenges posed by the associated optimization problem arising in medical settings, registration presents an active field of research. In this work, we present several contributions to the field of registration of medical images: 1) A general preconditioning scheme, which leads to efficient optimization for arbitrary difference measures. 2) A method for single-view 2D-3D deformable registration of vascular structures, with applications in angiographic settings. 3) We propose the use of MRF modeling and discrete optimization for linear registration. 4) For generation of statistical deformation models, we propose the use of minimal deformations, from which the linear transformation components are removed.