Coronary artery disease accounts for more than half of all cardiovascular death in the United States (US) and approximately 7.9% of US adults have a history of coronary artery disease. Coronary magnetic resonance angiography (MRA) has been established as a valuable tool for non-invasive visualization of proximal coronary arteries without any ionizing radiation, which is a major drawback of X-ray and computed tomography. However, the clinical utility of coronary MRA is mainly limited to the proximal coronary arteries, which has been attributed to both inadequate temporal and spatial resolution. An improved spatial and temporal resolution can only be obtained if the signal-to-noise ratio (SNR) is sufficiently high. For these reasons, this dissertation focuses on novel approaches to improve SNR in coronary MRA by exploiting multi-channel phased array signal receive coils, parallel imaging, and contrast media.

The SNR enhancing techniques developed in this dissertation exploit sensitivity encoding for enhanced SNR at no extra cost in scanning time. With a higher sensitivity encoding acceleration factor, extra signal sampling time is made available and is utilized for signal acquisition rather than for radiofrequency transmission. Thus, the image noise is reduced and a substantial SNR benefit is obtained. The method has been implemented for coronary MRA with a sensitivity-encoding factor of 2, a high temporal resolution of 40ms and the theoretically predicted SNR improvement of 50% was consistently found in vitro and in vivo. When combined with a T1 shortening exogenous contrast agent, the sensitivity encoding acceleration further amplifies both SNR and blood-myocardium contrast-to-noise ratio (CNR). For adequate analyses and quantitative comparisons, new image processing tools had to be integrated together with the novel MRA approach. At the end of this thesis, this combined methodology was made available for a translational patient study as part of an international collaboration where a significant SNR and CNR improvement could successfully be documented. The SNR and CNR advantages have been traded for an enhanced spatial resolution both in vitro and in vivo. The proposed methodology is easy-to-use and is expected to benefit not only coronary MRA specifically but also motion suppressed imaging with high temporal and spatial resolution more generally.

This dissertation was prepared under the guidance of Prof. Matthias Stuber. The dissertation defense committee is comprised of Prof. Matthias Stuber (dissertation reader), Assistant Prof. Daniel Herzka (dissertation reader), Prof. Elliot McVeigh. Prof. Robert. G. Weiss, and Prof. Susumu Mori.