In medical imaging, large amounts of data are created during each patient examination, especially using 3-dimensional image acquisition techniques such as Computed Tomography. This data becomes more and more difficult to handle by humans without the aid of automated or semi-automated image processing means and analysis. Particularly, the manual segmentation of target structures in 3D image data is one of the most time consuming tasks for the physician in the context of using computerized medical applications. In addition, 3D image data increases the difficulty of mentally comparing two different images of the same structure. Robust automated organ segmentation and registration methods are therefore needed in order to fully utilize the potentials of modern medical imaging. This thesis addresses the described issues by introducing a new model based method for automated segmentation and registration of organs in 3D Computed Tomography images. In order to be able to robustly segment organs in low contrast images, a volumetric model based approach is proposed that incorporates texture information from the model’s interior during adaptation. It is generalizable and extendable such that it can be combined with statistical shape modeling methods and standard boundary detection approaches. In order to increase the robustness of the segmentation in cases where the shape of the target organ significantly deviates from the model, local elasticity constraints are proposed. They limit the flexibility of the model in areas where shape deviation is unlikely. This allows for a better segmentation of untrained shapes and improves the segmentation of organs with complex shape variation like the liver. The model based methods are evaluated on the liver in the portal venous and arterial contrast phase, the bladder, the pancreas, and the kidneys. An average surface distance error between 0.5 mm and 2.0 mm is obtained for the tested structures which is in most cases close to the interobserver variability between different humans segmenting the same structure. In the case of the pancreas, for the first time, an automatic segmentation from single phase contrast enhanced CT becomes feasible. In the context of organ registration, the developed methods are applied to deformable registration of multi-phase contrast enhanced liver CT data. The method is integrated into a clinical demonstrator and is currently in use for testing in two clinics. The presented method for automatic deformable multi-phase registration has been quantitatively and qualitatively evaluated in the clinic. In nearly all tested cases, the registration quality is sufficient for clinical needs. The result of this thesis is a new approach for automatic organ segmentation and registration that can be applied to various clinical problems. In many cases, it can be used to significantly reduce or even remove the amount of manual contour drawing. In the context of registration, the approach can be used to improve clinical diagnosis by overlaying different images of the same anatomical structure with higher quality than existing methods.