ABSTRACT
This chapter focuses on the geometric deformable models and how to deform it under a particular stochastic force till it captures the object properties of interest. It introduces the density distribution of the input image gray values are estimated using the modified expectation maximization algorithm. The chapter presents a novel and promising level set based technique for segmentation and that is dependent on the information from both the shape and the intensity. Shape delineation is the primary task in shape analysis. Such representation is essential in the computer vision field and in a number of medical imaging applications, such as registration and segmentation. To enhance the segmentation accuracy, expected shapes of goal objects are constrained with a probabilistic shape prior. Radiation-induced lung injury is the main side effect of radiation therapy for lung cancer patients. The severity of radiation-induced lung injury ranges from ground-glass opacities and consolidation at the early phase, to fibrosis and traction bronchiectasis in the late phase.
