Advanced approaches for medical image segmentation

Abstract

Image segmentation, i.e., dividing an image into its constituent's regions, is a decisive phase in plentiful medical imaging studies to extract meaningful information such as shape, volume, motion, and abnormalities and to quantify changes of the human organs by radiologists and investigators, which can be facilitated by several automated computational procedures. Several efficient approaches for medical image segmentation have been developed till now based on hard and soft computing models such as thresholding, clustering, graph cut approaches, fuzzy-based approaches, neural network approaches, and many more. Tremendous success of deep learning nowadays has achieved state-of-the-art performance for instinctive medical image segmentation. This chapter provides the brief introduction about medical image segmentation and several current researches for the precise dissection. Further, it will provide the information about the deep learning used as an advanced approach presently for accurate segmentation of medical images.