Comprehensive Validation of Cardiac SPECT Reconstruction

点击翻译按钮获取中文摘要

DESCRIPTION (Provided by Applicant): The diagnostic and quantitative accuracy of cardiac SPECT are limited by depth-dependent blurring, attenuation, scatter, and image noise. Approaches to correct for these physical effects have been described, but are not widely used clinically, primarily due to lack of comprehensive validation. We propose to pooi the talents of leading investigators in nuclear cardiology to validate the most promising techniques, as follows: Aim 1. To utilize a rigorous implementation of OSEM that includes models of attenuation, collimator blur, and shift-variant scatter (Tsui and colleagues), with our own approach, which is less rigorous but has produced excellent human multi-center trial results, as a baseline for comparison. Aim 2. To assess the quantitative accuracy of these approaches in a series of computer simulations and actual phantoms: the MCAT and NCAT code to produce a series of simulations of different "defects" (in terms of location, size, and degree of "ischemia"), with varying degrees of attenuation, scatter, and blur, and the newest Data Spectrum cardiac phantom, which is capable of "beating." The "true" activity concentration in the object (i.e., the model prior to simulation) or phantom will serve as the baseline for comparison. Aim 3. To assess the quantitative accuracy of these approaches in a series of dog studies with a model of experimentally-induced ischemia via variable coronary artery occlusion, with different ischemic regions (in terms of location and severity). The true distribution of myocardial perfusion, as measured by microspheres counted ex vivo, will serve as the baseline for comparison. Aim 4. To assess the effects of these approaches on the diagnostic accuracy of visual interpretation and polar-plot quantification of human multi-center clinical SPECT studies in the diagnosis of coronary artery disease. The results of coronary angiography will serve as the baseline for comparison. Hypotheses: The OSEM approach will have significantly better diagnostic and quantitative accuracy than other approaches, clinically important in magnitude, especially in mild disease patients.