MR coronary angiography, back to the future?

size and have a complex 3-dimensional (3D) trajectory around the heart. Furthermore, they present a complex motion resulting both from cardiac contraction and respiration. Currently, there are two techniques which compete for such noninvasive coronary imaging: MR and CT. Noninvasive coronary artery imaging by MR (MRCA) was introduced almost 20 years ago [5] and has steadily evolved over time. The first MRCA acquisitions were performed during breathholds. This, however, limited acquisition to single 2D slices allowing only visualization of short segments of proximal coronary arteries. The introduction of free-breathing navigator sequences [6, 7] by the end of the 1990s made it possible to perform longer 3D acquisitions with higher spatial resolution, and offered the possibility to visualize longer segments of the coronary arteries with higher spatial resolution. Initial approaches required individual prescription of multiple coronary artery segments on localizer images, a very time-consuming approach resulting in study durations of about 1 h to visualize all coronary artery segments. In 2005, a faster approach was proposed, i.e. whole-heart MRCA imaging [8–10] . With this approach, instead of localizing individual coronary arteries at the time of acquisition, only one large axial 3D volume encompassing the entire heart with all coronary arteries is acquired. Similar to CT, the individual coronaries are Invasive X-ray coronary angiography (XCA) remains the clinical reference method for the detection of coronary artery disease in patients. Each year, several millions of XCAs are performed worldwide to make the diagnosis of coronary artery disease and to select therapeutic strategies. Although XCA is definitely a very effective diagnostic tool, it is undoubtedly an invasive procedure with a nonnegligible risk of morbidity and mortality. Because of this, it usually needs to be performed in a hospital setting, and is therefore associated with high costs [1–3] . Furthermore, in overall clinical practice, XCA is clearly overused and has a low diagnostic yield. Indeed a recent study [4] reported that only one third of patients without known coronary artery disease who underwent elective XCA have obstructive coronary artery disease, leading to revascularization procedures. Therefore better strategies are required to limit the intense clinical use and the high amounts of nondiagnostic XCAs. Ideally, XCA would only be needed in patients who have significant coronary artery disease, and who will undergo revascularization therapy. The ability to detect or exclude coronary artery disease by noninvasive imaging techniques could be an attractive approach to select patients prior to XCA. Yet such noninvasive detection of coronary artery disease is technically challenging. Indeed, the coronary arteries are of a small Received and accepted: March 7, 2011 Published online: May 10, 2011