On the Use of Ultrasound Phase Data for Arterial Characterization

Atherosclerosis is one of the major causes of death in the world. It is an inflammatory disease in the arteries which causes a gradual increase in arterial wall thickness and stiffness. Eventually a plaque could be formed, protruding into the artery and partially occluding the blood flow. Rupture of such a plaque could cause cardiac infarction or stroke. This dissertation revolves around finding better methods to detect and diagnose different stages of atherosclerosis and also to better understand the physiology of arteries. The methods developed and presented in this dissertation are all based on non-invasive ultrasound images. Ultrasound is the most common tool to image, particularly, the carotid artery in clinics today. The first method utilizes the time domain phase of the ultrasound radio frequency data for plaque/tissue characterization. The center frequency of the backscattered signal is estimated using phase differences measured in the time domain. We have shown that there is a clear correlation between the center frequency shift (CFS) and scatter size using a series of agar phantoms with well-defined sizes of glass beads. We have also shown that the CFS can be used in vivo to determine carotid plaque vulnerability. The method could potentially become a useful tool to identify patients at risk for development of acute cardiovascular events as well as to monitor response to interventions. The second and third methods are based on motion tracking (one is again based on time domain phase data) and could be used to measure e.g. the longitudinal movement of the arterial wall. The longitudinal movement is a relatively recent discovered physiological phenomena and could potentially be used as an image-derived biomarker for vascular dysfunction. Measurements of the longitudinal movement are also interesting from a physiological point of view since the mechanism behind it is unclear. The fourth and fifth methods measure the thickness and diameter of the arterial wall during the cardiac cycle. Arterial wall thickness measurements are a standard procedure in cardiovascular research and may be used for early diagnosis of atherosclerosis. Although there is a large number of automatic methods developed for this purpose, most of these measurements are still made manually. One of our methods is the first to automatically measure the thickness and diameter of arteries in a small animal model