Patient-specific thrombosis risk in atrial fibrillation by 4D CT imaging of atrial kinetics combined with computational fluid dynamics

点击翻译按钮获取中文摘要

Atrial fibrillation (AF) is the most common arrhythmia, affecting approximately 35 million people worldwide. During AF, the heart's two upper chambers (the atria) beat weakly and irregularly creating regions of slow flow (blood stasis) where clots may form. Clots preferentially form within the left atrial appendage (LAA) and can travel to the brain resulting in stroke. The risk of embolic strokes in AF patients is reduced with anticoagulation medications but, due to the associated increased bleeding risk, these medications are not recommended for all AF patients. Determining if anticoagulation is beneficial requires assessing if patients' risk of stroke outweighs the bleeding risk. However, current methods to risk-stratify AF patients for stroke are not personalized and, for a large number of patients, leave uncertainty as to whether anticoagulation is beneficial. The main objective of this project is to develop novel CT imaging analyses to quantify the personalized risk of LAA thrombosis in AF patients. Our scientific premise is that blood stasis is a key ingredient of thrombosis because it permits thrombogenic reactive species to interact and initiate clot formation. Our preliminary data suggest the spatio- temporal dynamics of blood flow and wall motion in the atrium and LAA strongly correlate with thrombus formation. Our approach consists of three specific aims. In Aim 1 we will develop and validate a computational framework to quantify left atrial blood stasis by 4D CT imaging of atrial kinetics combined with computational fluid dynamics (CFD). We will develop image processing algorithms to quantify left atrial kinetics based on time- resolved CT scans, including the spatio-temporal dynamics of contrast opacification, imaged wall motion, and the non-Newtonian rheology of blood flow in the LAA. In Aim 2 we will establish the relationship between 4D atrial kinetics by multi-heartbeat contrast CT and blood stasis using CFD, in order to facilitate the clinical translation of stasis mapping by CT alone. We will also perform the first rigorous analysis of how uncertainty caused by imaging resolution, modeling assumptions, and physiological variability propagates into predictions of LAA blood stasis. In Aim 3 we will perform an outcome-based clinical pilot study to develop a personalized image-based thrombosis risk score. We will acquire CT data in patients with a history of LAA thrombus or AF- associated stroke and a matched comparison group of AF patients with no history of thrombosis. We will use this unique data set to develop a patient-specific image-based risk score incorporating CT contrast opacification analyses with functional and geometric parameters. Our team includes a cardiologist with a physics background specializing in imaging, an engineer with expertise in CFD analysis, and an engineer with expertise in quantitative analyses of cardiac imaging. Our translational goal is to provide clinicians with a novel image-based tool for personalized risk stratification of patients with atrial fibrillation to guide anticoagulation decisions and improve outcomes.