Displacement and strain imaging of coronary arteries with intraluminal ultrasound.

Abstract

Recently, catheter-based intravascular ultrasound (IVUS) has been an area of great interest, partly due to the widespread incidence of arteriosclerosis. As the number of therapeutic modalities for this disease grows, each with unique advantages and disadvantages, physicians rely increasingly on diagnostic information to select the most appropriate treatment for each patient. This thesis describes the Pressurized Intraluminal Speckle Tracking algorithm (PIST), the goal of which is to enhance coronary artery disease treatments by directly providing diagnostic information about plaque mechanical properties with ultrasound. It has been shown that the most important factor contributing to the success probability of therapies for occluded arteries are the plaque and artery wall elasticities (stiffness). Although recent progress in IVUS has been made possible by advances in catheter technology, many improvements have not directly addressed elasticity. Rather, IVUS improvements have focused mainly on identifying various plaque tissues by associating them with their image brightness, which only weakly correlates with tissue characteristics. Elasticity imaging holds greater promise for distinguishing different plaque types because plaque elasticity spans a broader range than brightness. Elasticity imaging measures how tissue is displaced in response to pressure changes. It involves three critical steps: computation of tissue motion under controlled deformation using correlation-based, phase-sensitive speckle tracking; assessment of 2D tissue strain tensor elements; and elasticity estimation using these strain components. This thesis explores the first two in the context of coronary arteries imaged with IVUS, with the ultimate goal of enhancing diagnostic capabilities in coronary arteries. To demonstrate that displacements and strains can be used to distinguish materials of differing stiffnesses, arteries with soft and hard plaques as well as a homogeneous artery were simulated in a computer experiment. Displacement and strain imaging were successful in identifying each of these cases. Laboratory experiments using a clinical balloon catheter with an integrated ultrasonic imaging unit and a homogeneous tissue-equivalent phantom confirmed the simulation results: the behavior of the homogeneous phantom was azimuthally highly symmetric, while displacement and strain images successfully distinguished soft material from hard in an inhomogeneous phantom.