The registration of two medical images is usually based on the assumption that corresponding regions exist inboth images. If this assumption is violated by e. g. pathologies, most approaches encounter problems. Thehere proposed registration method is based on the use of probabilistic correspondences between sparse imagerepresentations, leading to a robust handling of potentially missing correspondences. A maximum-a-posterioriframework is used to derive the optimization criterion with respect to deformation parameters that aim tocompensate not only spatial differences between the images but also appearance differences. A multi-resolutionscheme speeds-up the optimization and increases the robustness. The approach is compared to a state-of-theartintensity-based variational registration method using MR brain images. The comprehensive quantitativeevaluation using images with simulated stroke lesions shows a significantly higher accuracy and robustness of theproposed approach.
展开▼
