Deformable models for medical image segmentation are often enhanced by their use of prior shape information. Some problems are well suited to the constraints that global shape information provides, where the shapes of the organs or structures are very consistent and are well characterized by a specific shape model. Other problems involve structures whose shapes are highly variable or have no consistent shape at all and thus require more generic constraints. We describe approaches to these two types of segmentation problems illustrating the varying uses of shape information. For the first, we describe integrated approaches in a maximum a posteriori formulation using parametric models with associated probability densities. For the second, we describe level set methods which incorporate powerful generic shape constraints, in particular, a thickness constraint. These approaches are illustrated with examples from images of the heart and brain. We will discuss the development of these ideas, current methodology and future directions.
Was this article helpful?