Theses and Dissertations


Xiaodan Shi

Issuing Body

Mississippi State University


Liao, Jun

Committee Member

Butler, James Ryan

Committee Member

To, S.D. Filip

Committee Member

Zhang, Song

Committee Member

Elder, Steven H.

Date of Degree


Original embargo terms


Document Type

Dissertation - Open Access


Biomedical Engineering

Degree Name

Doctor of Philosophy


James Worth Bagley College of Engineering


Department of Agricultural and Biological Engineering


In cardiac biomechanics, there is an apparent knowledge gap in 3D cardiac elastin structure and its biomechanical roles. In this study, we fill this knowledge gap via novel biomedical imaging and bioengineering means. In Aim 1, we created an overall mapping of 3D microstructures of the epicardial elastin fibers on porcine left ventricles (LV) using a laser scanning confocal microscope. We demonstrated the location- and depth-dependencies of the epicardial elastin network. Histological staining was also applied to reveal the patterns of endocardial and interstitial elastin fibers, as well as elastin fibers associated with the Purkinje fibers. In Aim 2, a novel algorithm was developed to better reconstruct the elastin fiber network and extract topological fiber metrics. We created a “fiberness” mask via fiber segmentation and fiber skeletonization to obtain the one-voxel-thick centerline skeleton and remove spurious fiber branches, thus generating topological and geometrical descriptors and bringing the study of cardiac elastin to a new level. In Aim 3, we successfully developed a semi-quantitative approach to characterize the residual stress in the epicardial layer by calculating the total angular change due to curling. Our novel curling angle characterization clearly reveals the existence of residual stress as well as the direction (circumferential vs. longitudinal) and location-dependency of the residual stress. In Aim 4, for the first time we estimated the regional residual stress of the epicardial layer on the intact LV via a four-step methodology: (i) quantify regional residual strains by comparing in situ and stressree marker dimensions; (ii) obtain regional tension-stretch/stress-stretch curves along the circumferential and longitudinal directions; (iii) adjust the biaxial curves to the 0g load reference; (iv) estimate the circumferential and longitudinal residual stresses via residual strains. This method accurately estimates the residual stress in the epicardial layer in various LV anatomical locations. We found that the location-dependency of circumferential and longitudinal residual stresses correlates with the curvature of heart surfaces. Our studies show that the epicardial layer, with its rich elastin content, might function as a balloon that wraps around the heart, and the residual stress sets up a boundary condition that assists with LV contraction.



fiber network reconstruction||3D microstructure||cardiac elastin||bioengineering||residual stress