Modeling, simulation and analysis of the heart from four-dimensional cardiac tagged-MR images
Alteration in heart shape and motion is a reasonable indicator of heart diseases. Insights with regard to normal physiology and dysfunction can help in understanding effects and their causes. In the last decade, there has been much progress in developing techniques for studying heart motion with cardiac imaging. Most existing model-based techniques are in the preliminary stages. Statistical model-based techniques deal with the huge variations of shape and motion of the whole heart but these methods have no temporal and multiple subjects' correspondences, so it is difficult to build a model from training sets and analyze different hearts and their motion. Previous parameterized model-based methods could only handle the left ventricle (LV) or up to mid right ventricle (RV) although a model including both the left and right ventricles up to the basal area is needed for comprehensive understanding of cardiac physiology and anatomy. The thesis uses a whole heart model, including LV RV and up to the basal area, for the functional analysis of heart motion. The model, based on a blended parameterized deformable model, is generic enough to deal with different hearts. A generic heart model is coupled with the finite element method to reconstruct heart motion from tagged magnetic resonance (MR) images. Tagged MR is the most promising non-invasive technology to characterize myocardial deformation during the heart cycle because it provides temporal correspondence of material points inside heart walls and enable tracking of these material points over time. The resulting parameters are used for assessing and analyzing global and local cardiac functions. The quantitative analysis derived the complicated but typical patterns of motion and strain on ten test subjects. The significant distinct motion and strains are found from RV hypertrophy patients. Enough sets of experiments can yield parameter values associated with typical normal subjects, identify abnormally functioning area such as infarct regions, and provide diagnostic information by associating altered parameter values with different stages of disease.
Computer science|Biomedical research
Park, Kyoungju, "Modeling, simulation and analysis of the heart from four-dimensional cardiac tagged-MR images" (2005). Dissertations available from ProQuest. AAI3165742.