ABSTRACT
In general, the term segmentation denotes the process of assigning sets of pixels to one or more distinct groups that are defined by the needs of the respective image processing task. Regarding medical imaging, volumetric segmentation is based on the classification of voxels to regions, which usually correspond to objects or organs in the data set. The chapter focuses on binary segmentation since most medical imaging classification techniques target at a clear distinction of the detected structures. The goal of registration is to provide a mapping of two different images that show comparable content. Finding the appropriate transformation between the images is the challenge of every registration algorithm. The term fusion denotes the combined visualization of registered data sets. Since the goal of registration is the comparison of images that show different information, methods are needed that present the clinically important structures from both data sets while hiding irrelevant details.
