I. Introduction
The advances in the medical imaging technology allow for in vivo visualization and analysis of human body with unprecedented accuracy and resolution. A diagnosis by a specialist often requires a visit to a radiology department to obtain various images that highlight the suspected pathology. Despite the high resolution of the acquired images, image-based diagnosis often utilizes a considerable amount of qualitative measures. To improve the diagnosis and efficiency, the research in medical image analysis has focused on the computation of quantitative measures by automating some of the error-prone and time-consuming tasks, such as segmentation of a structure.