In this competing renewal application, the goal is the robust, noninvasive magnetic resonance (MR) imaging of the proximal coronary arteries in humans. MR is a dynamic modality that offers enormous flexibility to deal with the numerous issues involved in coronary artery imaging. These issues include motion, signal-to-noise ratio (SNR), blood vessel contrast, spatial resolution, volumetric coverage, and protocol ease-of- use. Improvements in MR hardware, pulse sequences, and data processing algorithms have been spurring steady advances in coronary MR imaging. Excellent visualization of the coronary arteries on normal subjects is achievable. Performance on subjects with coronary artery disease, while inconsistent, has been improving. Further research is therefore required to raise MR to the level of clinical utility. The specific plan of this research project includes the development and evaluation of new MR methods for coronary artery imaging. The key imaging methods under consideration are breath-hold multi-slice 2D spiral imaging, real-time interactive imaging, and free-breathing 3D cones imaging. As compared to other MR coronary imaging methods, these non-Cartesian scanning methods offer the potential for improved motion and SNR performance in the heart. Several important enhancements will be investigated in this project, including those for improved blood contrast, better motion detection-compensation, and faster volumetric imaging. Following a period of technical development and preliminary evaluation, a comparative study of the optimized coronary MR imaging methods with x-ray blood vessel imaging will take place on patients with suspected coronary artery disease. Because coronary artery disease remains the leading cause of death in the United States, the development of a noninvasive method for imaging the coronary arteries would have a major impact on health care. This project seeks to attain this goal using magnetic resonance imaging, a safe, painless, and radiation-free technology. ? ? ?
Showing the most recent 10 out of 84 publications
