TABLE OF CONTENTS TONG-YEE LEE [*] and CHAO-HUNG LIN
Computer Graphics Group/Visual System Lab
Department of Computer Science and Information Engineering
National Cheng Kung University, Taiwan, ROC
In the past, a variety of interpolation approaches have been proposed for medical images. The shape-based interpolation is a well-known and most commonly used method that can be implemented efficiently and achieve reasonable results. In this chapter, the morphology-based interpolation is first introduced and it can solve drawbacks in the shape-based interpolation method. Next, a powerful feature-guided image interpolation is presented. This method automatically finds feature-line segments and integrates image-warping technique to interpolate shapes. In comparison with the shape-based and morphology-based methods, the feature-guided method can manage more general image cases and generate better shape interpolation.
Clinicians exploit computer graphics tools to enable them to visualize, manipulate, and quantitate the 3D internal structures of patients. Major sources of data in these medical applications are gathered from 2D medical-imaging devices such as CT, MRI and PET. A 3D image, formed by stacking a contiguous series of 2D images, can be used to visualize complex structures in 3D. However, generally, the number of image slices generated from these instruments is not adequate enough to produce high-quality 3D images. Therefore, the 3D reconstruction must be accomplished by appropriate interpolation methods to fill gaps between available image slices.
A variety of approaches have been proposed to reconstruct 3D objects. Among these works, the simplest method is to linearly interpolate the gray values in the slices to fill in...
