Quantitative two-dimensional echocardiography
Degree GrantorUniversity of Canterbury
Degree NameDoctor of Philosophy
This thesis reports studies of the motion and shape of the walls of the heart using data extracted from two-dimensional echocardiograms. The relevant literature and clinical practice are first reviewed at length. The development of hardware and software systems for digitally recording two-dimensional echocardiograms (scans) is described. Computer techniques for enhancing scans of the left ventricle of the heart are presented. Methods of interactively extracting the endocardial outlines seen in the scans are described. The outlines of the left ventricular walls were extracted from the scans of two groups of subjects. One group consisted of nine normal volunteers, the other consisted of nine coronary artery disease (CAD) patients known to have abnormal left ventricles. The outlines obtained were used to study methods of analysing left ventricular wall motion and shape under exercise and drug interventions. Two cardiac drugs were used: nifedipine and metoprolol. Wall motion was measured in terms of the change in distance of the walls from the ventricular centroid. CAD patients with apical infarctions were found to have hyperkinetic regions of the walls adjacent to their infarctions. The hyperkinesia decreased when the patients were subjected to isometric exercise. Under the influence of nifedipine the hyperkinesias increased. However under the influence of metoprolol it decreased. These observations are consistent with the known effects of exercise and the drugs on the motion of the heart. Ventricular shape was described using a trigonometric Fourier series. Descriptors derived from the coefficients of the series were developed. The values of some of the descriptors of the CAD patients with apical infarctions were found to be larger than those of the normal volunteers. Under exercise the value of the descriptors of the normal volunteers did not change, while those of the apical infarction patients increased further. Nifedipine was found to increase the values of the descriptors of the patients. Metoprolol in contrast, reduced the values. These findings are consistent with the expected effects of exercise and the drugs on the shape of the left ventricle. Therefore, the methods developed may be used to quantitatively assess the effects of other exercise and drug interventions on ventricular wall motion and shape. The quantitative analysis of ventricular shape may also be used as a means of identifying patients with ventricular wall abnormalities.