Medical Image Segmentation

Image data is of immense practical importance in medical informatics. Medical images, such as Computed Axial Tomography (CAT), Magnetic Resonance Imaging (MRI), Ultrasound, and X-Ray, in standard DICOM formats are often stored in Picture Archiving and Communication Systems (PACS) and linked with other clinical information in EHR clinical management systems. Research efforts have been devoted to processing and analyzing medical images to extract meaningful information such as volume, shape, motion of organs, to detect abnormalities, and to quantify changes in follow-up studies. Automated image segmentation, which aims at automated extraction of object boundary features, plays a fundamental role in understanding image content for searching and mining in medical image archives. A challenging problem is to segment regions with boundary insufficiencies, i.e., missing edges and/or lack of texture contrast between regions of interest (ROIs) and background. To address this problem, several segmentation approaches have been proposed in the literature, with many of them providing rather promising results. In this chapter we focus on two general categories of segmentation methods widely used in medical vision, namely the deformable modelsand the machine learning-based classification approaches. We first review these two categories of methods and discuss the potential of integrating the two approaches, and then we detail on two of the most recent methods for medical image segmentation: (i) the Metamorphs, a semi-parametric deformable model that combines appearance and shape into a unified space, and (ii) the integration of geometric models with a collaborative formulation of CRFs in a probabilistic segmentation framework. We show different examples of medical data segmentation, we draw general conclusions from the methods described in this chapter, and we give future directions for solving challenging and open problems in medical image segmentation.