Topology in Medical Image Segmentation

TOPOLOGY IN MEDICAL IMAGE SEGMENTATION M. C. Jobin Christ1 and R. M. S. Parvathi2 1Lecturer, Department of Biomedical Engineering, Adhiyamaan College of Engineering, Hosur, Tamilnadu, India. E-mail: [email protected] 2Principal, Sengunthar College of Engineering for Women, Tiruchencode, Tamilnadu, India. E-mail: [email protected] Topology is the mathematical study of the properties that are sealed through deformations, twisting, and distending of objects. The topological belongings of objects are often very simple, irrespective of their geometric complexity. Human anatomy follows this rule, even for tremendously complicated shapes like the cerebral cortex or the vasculature. To achieve accurate and topologically correct descriptions of anatomical structures from medical images is a critical step for many clinical applications. Apart from pathological cases, most macroscopic brain structures are fully connected and do not possess any topological defect such as handles or cavities. Human brain has the simple topology of a sphere. Many recent segmentation methods are able to recognize and exactly locate these structures, although without compelling the topology. Being able to achieve exact and topologically correct illustrations of different brain structures is certainly an important goal in medical image processing. In this paper we are giving an overview about topology and medical image segmentation under topological constraints.