Cardiac motion synchronization for three-dimensional cardiac ultrasound imaging

Three-dimensional imaging of cardiac structures and coronary arteries can potentially provide reliable volume measurements of cardiac chambers, coronary lumen and plaques, for effective cardiac disease management. The overall goals of this research are: to determine factors that affect the accuracy and precision of 3D cardiac ultrasound measurements; and to develop techniques that can eliminate the major source of variability, due to cardiac motion artifacts, in 3D cardiac ultrasound images. This thesis is based on four specific objectives.; The first objective was to investigate the intrinsic accuracy and precision of 3D-echocardiographic measurements of LV volume and mass under ideal in-vitro conditions. This study showed that 3D-echocardiography had twice the accuracy and half the variability of 2D-echocardiography in LV volume and mass measurement. Based on theoretical analyses, it was concluded that intrinsic variability in volume measurement depends on: the number of image slices used for measurement, and the uncertainty in determining cardiac boundaries in the 3D image.; In in-vivo 3D images, larger measurement variabilities are expected due to cardiac motion artifacts. ECG-gating, commonly used to resolve cardiac motion artifacts, causes temporal jitter in 3D images. The second objective was to simulate and quantify temporal jitter in 3D-echocardiography using in-vitro studies. Temporal jitter increased with increasing phantom velocity and could be accurately predicted from the motion waveform.; The third objective was to develop an image-based retrospective gating technique for 3D cardiac ultrasound imaging, to overcome limitations of ECG-gating. This technique was evaluated using in-vitro and in-vivo studies, by quantifying and comparing the reduction in cardiac motion and pulsation artifact in retrospectively-gated and non-gated 3D images. Residual artifact in 3D images using image-based gating was compared with predicted residual artifact using ECG-gating. Image-based gating reduced cardiac motion artifacts by 90% and performed substantially better than ECG-gating.; The fourth objective was to develop a realistic dynamic vessel phantom to evaluate the image-based gating technique. The phantom pulsation could be accurately controlled to simulate normal human coronary pulsation with an error <7%.; The techniques described in this thesis can potentially improve measurement precision in 3D cardiac ultrasound images by significantly minimizing cardiac motion artifacts.