This Is AuburnElectronic Theses and Dissertations

Biventricular Active Mesh Model of the Heart and Analysis of Morphologic Changes Toward Physiology and Pathologies




Schiros, Chun

Type of Degree



Electrical Engineering


Cardiovascular disease (CVD) has been reported as the number one cause of death in the world. It is estimated that by 2030, about 23.6 million people will die from a type of CVD. It is a problem that crosses both gender and ethnicity and is a problem that gets worse with age. Heart failure is usually the end result of most cardiac diseases. Therefore, correct diagnosis and early prevention of CVDs are significantly important. Used as powerful tools for clinical diagnosis, medical imaging techniques have been dramatically developed and improved over the past two decades. Among them, cardiovascular magnetic resonance imaging (cMR) is becoming a leading imaging modality for advanced clinical research, drug studies and patient management due to its high image resolution, minimal invasion and reproducibility, compared with other conventional imaging techniques. Geometric and functional analyses of the left ventricle (LV) of human hearts have been well developed using cMR. However, the development of right ventricle (RV) shape and functional analyses are relatively new territory. Analysis of the RV and interaction between the LV and the RV can provide extra information that might suggest subtle abnormality of cardiac function in patients with normal LV functions. In this dissertation, three major research projects are presented. The first study evaluated the role of the alterations in the changes in LV volume and geometry in achieving elevated stroke volume in endurance athletes’ hearts and patients with chronic compensated mitral regurgitation (MR), which represented physiologic, and pathologic left ventricular volume overload. Cine cMR, with tissue tagging was performed on all subjects of both groups. Three-dimensional data analysis, utilizing in-house software was performed to evaluate the differences in geometry and function between the two groups. The result of this study shows that in the setting of similar increases in LV volumes and stroke volume, marathoners’ hearts maintain a normal LV sphericity, a conically-shaped apex and normal wall thickness with lower LV twist, while, in MR hearts, LV sphericity is increased and the apex is more rounded. The second study further assessed the important role that apex remodeling played in the progression of the severity of mitral regurgitation and its significance as an indicator for timing of surgery. cMR and 3D data analysis were performed on 94 patients with chronic degenerative isolated MR to uncover the importance of volumetric analysis as an indicator of left ventricular remodeling. Among these patients, 35 patients underwent mitral valve repair and each had a 12-month follow-up analysis. The major finding of the study is that LV end-systolic (LVES) dimension does not accurately reflect the extent of LV remodeling, largely due to rounding of the apex and global spherical LV remodeling. The previous two studies highlighted the need for accurate modeling of LV geometry – particularly at the apex. Other research has highlighted a similar need for accurate modeling of RV geometry, which does not have the circular symmetry of the LV. Most current LV models use a polar type of coordinate system with a singularity at the origin, which makes it difficult to model the LV apex and the RV as a whole. Therefore, for the third research project, we proposed a biventricular active mesh model of the human heart that can accurately fit smooth surfaces to both the LV and RV including the LV apex and RV base. The computation time for generating a mesh for a new subject using the proposed algorithm was less than one minute. Moreover, such new meshes can potentially correct the contour errors near both the LV and RV outflow tracts which were usually difficult to determine manually, because the generated new meshes were constrained by the variation from their training sets.