Files in this item

FilesDescriptionFormat

application/pdf

application/pdf9305488.pdf (7MB)Restricted to U of Illinois
(no description provided)PDF

Description

Title:Left ventricle motion and shape modeling, analysis, and visualization from image sequences
Author(s):Chen, Chang Wen
Doctoral Committee Chair(s):Huang, Thomas S.
Department / Program:Engineering, Biomedical
Engineering, Electronics and Electrical
Discipline:Engineering, Biomedical
Engineering, Electronics and Electrical
Degree Granting Institution:University of Illinois at Urbana-Champaign
Degree:Ph.D.
Genre:Dissertation
Subject(s):Engineering, Biomedical
Engineering, Electronics and Electrical
Abstract:This thesis addresses the problems of modeling, analysis, and visualization of left ventricle motion and deformation based on two types of image sequences. The first type of image sequence used in this research is the coronary cineangiograms. The 3D coordinates of a set of bifurcation points obtained from the coronary cineangiogram are used to estimate the global motion and deformations as well as local motion and deformations of the left ventricle. The second type of image sequence is the dynamic CT volumetric data obtained from the unique DSR scanner. The left ventricle chambers extracted from the volumetric data are used to estimate the global motion and global deformations of the left ventricle. The estimated results of left ventricle shape, motion and deformations from both types of image data are then used to generate the animation sequences to effectively analyze the estimated numerical results.
A hierarchical motion model is proposed for developing a model based approach to the estimation of left ventricle motion and deformation from the image sequences. Our hierarchical motion model is the first attempt to include global motion and deformations as well as the local motion and deformations. Based on this model, the hierarchical decomposition of motion and deformation analysis is described. This decomposition leads to computationally efficient implementations of the seemingly complex estimation algorithms.
Two surface-modeling primitives are presented to parameterize the global and local deformations. The superquadric modeling primitives are used to characterize the global deformations including expansion, contraction, tapering, bending and twisting. The spherical harmonic modeling primitives are used to characterize the local surfaces that cannot be modeled by the superquadric surfaces. Through surface modeling, the estimation of the left ventricle deformations is accomplished by fitting the 3D data to the modeling primitives. This surface-fitting-based deformation analysis extends previous research that has been using only simple surface models, such as cylinders and ellipsoids.
Based on the proposed motion and shape models, the algorithms for the estimation of left ventricle motion and deformation from both angiographic data and CT data are developed. The estimation algorithms are implemented in a coarse-to-fine fashion through the application of the hierarchical decomposition of the complex motions. This model-based approach has been successfully applied to left ventricle motion and deformation estimation from the angiographic image sequences as well as the CT volumetric image sequences. The animation sequences generated using the estimated left ventricle shape and motion parameters show the apparent motion patterns of the left ventricle dynamics.
Issue Date:1992
Type:Text
Language:English
URI:http://hdl.handle.net/2142/23531
Rights Information:Copyright 1992 Chen, Chang Wen
Date Available in IDEALS:2011-05-07
Identifier in Online Catalog:AAI9305488
OCLC Identifier:(UMI)AAI9305488


This item appears in the following Collection(s)

Item Statistics