Motion Detection in Dynamic Cardiac Positron Emission Tomography Images

Abstract

Dynamic PET is being increasingly utilized to quantify myocardial blood flow, leading to improved diagnostic and prognostic value for the management of coronary artery disease. MBF quantification assumes constant positioning of anatomical structures throughout image acquisition. This assumption is violated in the presence of patient motion and heart motion. This thesis discusses methods for motion detection in dynamic cardiac 82Rb PET images. First, we determined if positron-emitting fiducial markers can be used as surrogate markers for heart motion in dynamic PET images. Second, algorithms were developed to track fiducial marker motion. Lastly, we tested the idea that principle component analysis may be able to detect heart motion in dynamic 82Rb PET based on patterns in the residue image. Overall, this thesis proposes low-cost, fast, and vendor independent post-reconstruction dynamic 82Rb PET motion detection methods which can be used to detect motion in a clinical environment as a quality assurance tool.