3D Ultrasound Image Segmentation: Interactive Texture-Based Approaches

Abstract

The recent breakthroughs in 3D medical imaging technologies open new promising perspectives in the health domain. Significant efforts have being carried out and the precision of acquisition systems keeps on being improved. These devices are becoming widespread in the hospitals and the constantly increasing queues for this kind of exams prove the interest in these technologies. On the contrary, it seems that the evolution of 3D image analysis tools does not generate the same interest. It may be because, for a long time, people believed that considering 3D images as a succession of 2D frames was a sufficient and efficient way to make precise analysis. In our opinion, it is not the case and huge advancements can be obtained by considering 3D images with their entire complexity. This approach needs the development of specific models which can deal with multiple knowledge and are able to process huge information quantities. As a consequence, this chapter will present two original and efficient approaches of computer science for 3D image analysis and more particularly 3D ultrasound images.