Date on Master's Thesis/Doctoral Dissertation


Document Type

Doctoral Dissertation

Degree Name

Ph. D.


Computer Engineering and Computer Science

Degree Program

Computer Science and Engineering, PhD

Committee Chair

Elmaghraby, Adel S.

Committee Co-Chair (if applicable)

El-Baz, Ayman

Committee Member

Imam, Ibrahim

Committee Member

Nasraoui, Olfa

Committee Member

Li, Xiao-Feng


Heart--Diseases--Diagnosis; Diagnostic imaging; Heart--Imaging


In diagnosing heart diseases, the estimation of cardiac performance indices requires accurate segmentation of the left ventricle (LV) wall from cine cardiac magnetic resonance (CMR) images. MR imaging is noninvasive and generates clear images; however, it is impractical to manually process the huge number of images generated to calculate the performance indices. In this dissertation, we introduce a novel, fast, robust, bi-directional coupled parametric deformable models that are capable of segmenting the LV wall borders using first- and second-order visual appearance features. These features are embedded in a new stochastic external force that preserves the topology of the LV wall to track the evolution of the parametric deformable models control points. We tested the proposed segmentation approach on 15 data sets in 6 infarction patients using the Dice similarity coefficient (DSC) and the average distance (AD) between the ground truth and automated segmentation contours. Our approach achieves a mean DSC value of 0.926±0.022 and mean AD value of 2.16±0.60 mm compared to two other level set methods that achieve mean DSC values of 0.904±0.033 and 0.885±0.02; and mean AD values of 2.86±1.35 mm and 5.72±4.70 mm, respectively. Also, a novel framework for assessing both 3D functional strain and wall thickening from 4D cine cardiac magnetic resonance imaging (CCMR) is introduced. The introduced approach is primarily based on using geometrical features to track the LV wall during the cardiac cycle. The 4D tracking approach consists of the following two main steps: (i) Initially, the surface points on the LV wall are tracked by solving a 3D Laplace equation between two subsequent LV surfaces; and (ii) Secondly, the locations of the tracked LV surface points are iteratively adjusted through an energy minimization cost function using a generalized Gauss-Markov random field (GGMRF) image model in order to remove inconsistencies and preserve the anatomy of the heart wall during the tracking process. Then the circumferential strains are straight forward calculated from the location of the tracked LV surface points. In addition, myocardial wall thickening is estimated by co-allocation of the corresponding points, or matches between the endocardium and epicardium surfaces of the LV wall using the solution of the 3D laplace equation. Experimental results on in vivo data confirm the accuracy and robustness of our method. Moreover, the comparison results demonstrate that our approach outperforms 2D wall thickening estimation approaches.