Graduation Semester and Year

2008

Language

English

Document Type

Dissertation

Degree Name

Doctor of Philosophy in Computer Science

Department

Computer Science and Engineering

First Advisor

Hua-Mei Chen

Abstract

A novel nonrigid image registration algorithm is developed using a well-established mathematic work known as deformation based grid generation. The deformation based grid generation is capable to generate a grid free of mesh folding, which is achieved by devising a positive monitor function describing the anticipated grid point density in the computational domain. Based on this method, a novel nonrigid image registration algorithm is successfully developed with many interesting features. First of all, the functional to be optimized during the image registration process consists of only one term — the similarity term. Thus, no regularization functional is required in this method, not to mention the weight to balance the regularization functional and the similarity functional commonly required in many nonrigid image registration methods. Nevertheless, the regularity (no mesh folding) of the resultant deformation vector field is theoretically guaranteed. Secondly, since no regularization term is introduced in the functional to be optimized, the resultant deformation vector field is highly flexible that large deformation frequently experienced in inter-patient or image-atlas registration tasks can be accurately estimated. We present the detailed description of our proposed nonrigid image registration method with different implementations, alone with several 2D and 3D experimental results evaluating the registration quality, performance, and noise tolerance capability.

Disciplines

Computer Sciences | Physical Sciences and Mathematics

Comments

Degree granted by The University of Texas at Arlington

Download

Share

COinS