Date of Award

2011

Degree Type

Thesis

Department

Chemical and Biomedical Engineering

First Advisor

Setser, Randolph

Subject Headings

Heart -- Left ventricle -- Magnetic resonance imaging -- Technique, Myocardium, MRI Tagging analysis techniques, LV deformation measurement, Radial tagging MRI

Abstract

The extent of myocardial abnormalities is related to cardiac disease. The assessment of myocardial function of the left ventricular (LV) described by the LV myocardial deformation plays an important role to understand the effects of cardiovascular disease and therapeutic interventions on LV motion. Cardiac magnetic resonance imaging (CMR) is widely agreed as the most accurate noninvasive imaging modality for the assessment of left ventricular (LV) function. In addition, myocardial-tagging technique can track the myocardial deformation by labeling specific regions. Although the grid tagging is popular in cardiac research use, it has not been adopted into routine clinical examination with unreliable data between frames. We hypothesized that a radial-tagging methods has geometric advantages over grid tagging method and it is the most direct way for the circumferential strains.In this study, a semi-automatic radial tagging analysis tool was developed using MATLAB to measure the circumferential strain and rotation angle. The LV MRI short-axis image series (from end of diastole to end of systole) was used to develop and validate the analysis tool from 9 volunteers. The radial tagging analysis tool computes the strain and rotation angle by automatic calculating the mid-wall myocardium and the LV centroid. The results of radial tagging analysis were evaluated the accuracy by manual mathematical calculating of 2 volunteers and repeating tool operations 3 times to all volunteers. Also the results were compared with the grid tagging analysis and the manual radial tagging analysis tool. To visualize the differences among tools, the RMS (root-mean-square) value of the strains and rotation angle of all volunteers by the time ( of systole) was used.The results show that the strains and rotation angles of radial tagging analysis have no significant difference with P>0.05 compared to grid tagging analysis, which is conventional use, and the measured values are closer than manual radial tagging analysis. Also the tool is easy to

COinS