Coronary MR Angiography: Current Status

© Urban & Vogel 2000

431Herz 25 · 2000 · Nr. 4 © Urban & Vogel

Herz

Coronary MR Angiography: Current StatusPeter G. Danias1, Warren J. Manning1, 2

AbstractSince first described in the early 1990s, coronary magnetic reso-nance angiography (MRA) has evolved as a promising nonin-vasive modality for imaging of the coronary arteries and evalu-ation of coronary artery disease. Despite technical limitations,coronary MRA has established value for imaging of anomalouscoronary arteries and assessment of bypass graft patency. Cur-rent research focuses on the development of optimal respirato-

Aktueller Stand der koronaren MR-Angiographie

ZusammenfassungSeit ihrer Erstbeschreibung Anfang der 90er Jahre hat sich diekoronare Magnetresonanzangiographie (MRA) zu einemvielversprechenden nichtinvasiven Verfahren zur Darstellungder Koronararterien und zur Beurteilung der koronarenHerzkrankheit entwickelt. Trotz ihrer technischen Grenzen istdie koronare MRA in der Darstellung von Koronar-arterienanomalien und der Beurteilung der Bypass-Durch-gängigkeit zweifellos von Nutzen. Der Forschungsschwer-punkt liegt derzeit auf der Entwicklung optimaler re-

ry compensation strategies, improved spatial and temporal res-olution and faster acquisition of image data. The accurate de-tection of stenoses and assessment of the severity of coronaryatherosclerosis is presently being evaluated with large multi-center studies. With further technique enhancements and moreclinical experience, coronary MRA is likely to become the dom-inant noninvasive modality in clinical cardiology.

spiratorischer Kompensationsstrategien, einer verbessertenräumlichen und zeitlichen Auflösung und einer schnellerenSammlung von Bilddaten. Die exakte Erfassung von Stenosenund die Einschätzung der Schwere einer Atherosklerosewerden gegenwärtig in großen Multicenterstudien evaluiert.Mit technischen Verbesserungen und zunehmender klinischerErfahrung verspricht die koronare MRA zum überragendennichtinvasiven Verfahren in der klinischen Kardiologie zuwerden.

Key Words: Magnetic resonance angiography · Coronary artery imaging

Schlüsselwörter: Magnetresonanzangiographie · Darstellung von Koronararterien

1 Charles A. Dana Research Institute and the Harvard-ThorndikeLaboratory, Department of Medicine, Cardiovascular Division, and

2Department of Radiology, Beth Israel Deaconess Medical Centerand Harvard Medical School, Boston, MA, USA.

IntroductionDespite improvements in both prevention and treat-ment, coronary artery disease remains the leading causeof morbidity and mortality in industrialized nations [1].While numerous noninvasive tests are available to de-tect regional ischemia, direct assessment of coronary ar-tery integrity would represent a major advance in clini-cal care. Over the past decade, clinical magnetic resonance angiography (MRA) of large vessels such asthe aorta and carotid arteries has been widely accepted.MRA of the coronary arteries, however, offers particu-lar challenges. Early attempts to image the proximal

coronary arteries with conventional ECG-gated spin-echo sequences were only occasionally able to identifythe coronary ostia [30, 42], but did not succeed in visual-izing longer coronary artery segments or focal stenoses.The main obstacles to coronary MRA include: a) thesmall diameter (3 to 4 mm) of coronary vessels; b) mo-tion artifacts from cardiac contraction and respiration-related bulk cardiac motion; c) the presence of high sig-nal from the surrounding epicardial fat; and d) thetortuous course of the vessels, which results in luminaldiscontinuities on tomographic imaging due to vessel de-viation outside the image plane. Despite these limita-

tions, the potential to obtain images with high spatial and temporal resolution, the absence of associated ex-posure to ionizing radiation and the ability to image inany orientation have been attractive advantages ofMRA, in particular for coronary artery imaging. Othernoninvasive modalities, including echocardiography and computerized tomography, have additional limita-tions, such as the poor penetration and resolution (forechocardiography) and the need for X-ray exposure,iodinated contrast administration and prolonged breath-holding (for computed tomography). Accord-ingly, coronary MRA has emerged as the most promis-ing noninvasive test for assessment of coronary artery integrity. Over the last 10 years, the field of coronaryMRA has progressed rapidly, with research advances inseveral parallel areas, as described in the following dis-cussion.

Methods for Respiratory CompensationIn addition to ECG-gating, all current coronary MRAapproaches utilize some form of respiratory compensa-tion, including breath-hold and free-breathing tech-niques. Although breath-holding is attractive because itoffers the potential to quickly obtain image data, it alsocarries significant limitations due to breath-hold dura-tion and frequent need for breath-hold repetition.Breath-holding requires detailed patient instruc-tion/coaching and considerable motivation and cooper-ation. The breath-hold duration also limits spatial andtemporal resolution as well as the use of signal enhance-ments (such as averaging). Many patients, especiallythose with cardiac or pulmonary disease, have difficultysustaining adequate breath-holds. Although breath-holding ability can be augmented with the use of respir-atory maneuvers such as supplemental oxygen and hy-perventilation [9, 34], such maneuvers still requiresignificant patient involvement and are not practical forthe general population. Furthermore, there is signifi-cant variability of the breath-hold level even in motivat-ed subjects, which may account for varying cardiac posi-tion among serial breath-holds and poor slice registration. The result may be gaps between the coro-nary segments, which could be misinterpreted as signalvoids from coronary stenoses. Finally, during prolongedbreath-holds, the diaphragm (and heart) drifts towards a more cephallad position [9], thereby contributing toblurring and image degradation. Although alternativesuspended respiration techniques, including multiplebrief breath-holds [14] and coached breath-holding with

visual or audible feedback [31, 64], have been used tominimize respiratory motion artifacts and patient in-convenience, these approaches are only appropriate forhighly motivated subjects. Thus, while multiple breath-hold strategies are often successful with motivatedhealthy volunteers, their applicability to the broad range of patients with cardiovascular disease is morelimited.

To overcome the limitations of breath-hold imag-ing, free-breathing, respiratory-compensated coronaryMRA methods have been developed. The respiratorycycle may be easily monitored with the use of respirato-ry belts to track chest wall expansion. Image acquisitioncan thus be gated to a certain part of the respiratory cy-cle (commonly to the end-expiratory – minimal expan-sion position). Coronary MRA using such devices hasbeen shown to be both feasible and practical [6, 23, 36,40, 64]. However, external respiratory bellows-gating isnot robust, possibly due to temporal dissociationbetween chest wall expansion and cardiac motion.

A more direct and elegant method of monitoring di-aphragmatic and resultant cardiac displacement duringfree breathing is with the use of MR navigators, a tech-nique that provides real-time positional data about moving structures. If, for example, the navigator beam ispositioned at the base of the left ventricle, in close prox-imity to the coronary artery ostia, the respiration-relat-ed basal cardiac motion can be monitored, and imageacquisition can be gated to effectively “freeze” the res-piratory motion of the proximal epicardial coronary vessels. Because of limitations of the navigator imple-mentation on certain vendor hardware platforms (relat-ed to local signal loss – spin-echo navigators) and be-cause of relatively complicated navigator positioning atthe cardiac base, a vertical navigator is commonly posi-tioned on the dome of the right hemidiaphragm. As-suming a good correlation between coronary and dia-phragmatic position during the respiratory cycle, diaphragmatic positional information can be used to gate coronary MRA acquisitions. The use of prone im-aging appears to be particularly beneficial (Stuber M, unpublished data). Although there is significant individ-ual variability in the respiratory kinematics of the heart(coronaries) and the diaphragm, recent approaches have been shown to consistently provide sufficient im-age quality for major coronary artery visualization. Sev-eral navigator implementations have been described

Danias PG, et al. Current Status of Coronary MR Angiography

432 Herz 25 · 2000 · Nr. 4 © Urban & Vogel



and shown to be practical for coronary MRA in volun-teers and selected patients with coronary disease. Theseinclude retrospective gating, prospective gating, andprospective gating with slice-tracking [8, 39, 57]. Very sophisticated approaches to compensate for respiratorymotion and improve scan efficiency, including the di-minishing variance algorithm [49, 50], phase reordering[25, 26], and linear phase shift [17], have also been re-cently described.

2-D versus 3-D ImagingInitial coronary MRA approaches used two-dimension-al (2-D) strategies. K-space segmentation was intro-duced for coronary MRA in the early 1990s, allowingthe acquisition of multiple phase-encoding steps duringeach cardiac cycle [5, 16]. Accordingly, with 8 phase-en-coded acquisitions during each RR interval, 16 heart-beats were sufficient to fill 128 lines of k-space (in thephase direction). This enabled the use of breath-holdingto minimize respiratory motion artifacts. Additional se-quence components included a fat saturation pre-pulseto null the signal from surrounding epicardial fat. Withthis approach and use of a 260-mm field-of-view, a 1.9 �0.9-mm2 in-plane resolution could be obtained, allowingthe visualization of the proximal segments of coronaryarteries [16, 22, 32, 40, 44].

In general, 2-D approaches take advantage of theenhanced contrast between laminar coronary bloodflow and stationary tissues, related to inflow of unsatu-rated blood. 2-D approaches allow the relatively easyimplementation of breath-holding, as previously de-scribed. However, the coronary vessel tortuosity is animportant impediment for 2-D coronary MRA. Focalsignal loss due to the deviation of the coronary arteryout of the image plane may be misinterpreted as a focalstenosis. Additional limitations of the 2-D approach in-clude relatively low signal-to-noise ratio (SNR) and thepotential for slice registration errors related to inconsis-tency between serial breath-holds. Technical hardwareand software improvements have allowed the develop-ment of 3-D coronary MRA techniques, which havebeen implemented with various types of respiratorycompensation (see above). In comparison to 2-D tech-niques, 3-D coronary MRA offers enhanced SNR andimproved slice registration. Image data can be post-pro-cessed in any orientation (particularly with isotropic im-aging). The consideration of lower contrast-to-noise ra-tio (CNR) due to relative attenuation of the blood

inflow-related contrast effect, has been dealt with theuse of contrast agents, or manipulation of the myocar-dial tissue magnetization (T2 preparation pre-pulses [3,4, 60], spin-locking [12], or magnetization transfer [29,41, 68]). With state of the art-imaging, submillimeter in-plane resolution has been achieved in various hardwareplatforms [37, 57] (Figures 1 and 2).

Contrast-Enhanced Coronary MRAThe use of intravenous contrast agents for coronaryMRA has several theoretical advantages. Currently ap-proved gadolinium-(Gd-)based intravenous contrastagents decrease the T1 relaxation time of blood, thus im-proving the CNR for blood [21, 56] and have been wide-ly embraced for aortic and renal artery MRA. Accord-ingly, contrast enhancement has potential advantagesfor 3-D approaches, in which the inflow contrast effect isattenuated. Extracellular contrast agents, such as gado-pentate diglumine, gadodiamide and gadotetriiol, quick-ly diffuse in the extravascular space, resulting in a rapiddecrease of the contrast between blood pool and myo-cardium. Therefore, for coronary MRA using extracellu-lar agents, first pass imaging (and consequently pro-longed breath-holding) is required. Repeat imaging inshort intervals is impractical. The newly developed intra-vascular agents, including iron oxide, MS-325 andNC100150 [21, 27, 55, 56, 58], have the advantage of per-sisting in the blood pool at high concentrations for longtime periods, thus allowing for more lengthy approaches(such as use of free-breathing navigator-assisted imag-ing) and repeat imaging (Figure 3). The use of theseintravascular agents for coronary MRA is still consid-ered investigational and phase-II clinical trials of the ef-ficacy and safety of these agents are currently under way.

Fast Imaging ApproachesIn order to decrease the data acquisition and scan dura-tion, faster acquisition schemes have been developedover the last several years. Conventional spin-echo tech-niques were replaced by gradient-echo acquisitions, andfollowed by the introduction of echoplanar and hybridsequences. Most recently, parallel imaging has evolvedand promises to markedly decrease imaging times to afraction of current applications. The use of such fasttechniques has recently been reported for imaging ofthe coronary arteries [54] and hold great potential. Par-allel acquisition utilizes the variable sensitivities of com-ponent coils, which depend on the individual coil posi-tion, to render spatial image information [53, 65].



Figure 1Reformatted high-resolution image (0.5 � 0.5 mm2) of the right coro-nary artery (arrows) of a normal volunteer. A long portion of the vesselincluding its proximal and mid portions. Part of the left circumflex cor-onary artery is also visualized (open arrow). (Modified from [7]. Re-printed with permission.)

Abbildung 1Reformatiertes MR-Bild der rechten Kranzarterie (Pfeile) einerNormalperson. Das Bild ist mit hoher Ortsauflösung angefertigtworden (0,5 � 0,5 mm2). Zu sehen ist eine lange Strecke des Gefäßeseinschließlich der proximalen und mittleren Abschnitte. Ein Teil derArteria circumflexa ist ebenfalls sichtbar (offener Pfeil). (Modifiziertvon [ 7] . Wiedergegeben mit Genehmigung.)

Figure 3The use of the investigational intravascularagent MS-325 (Epix Medical, Inc, Cam-bridge, MA, USA; left panel) increased thecontrast-to-noise ratio (CNR) during a 3-Dhigh-resolution free-breathing double-oblique scan of the right coronary artery(arrows), compared to the non-contrastscan (right panel). Representative contigu-ous 3 mm thick slices are shown.

Abbildung 3Der Einsatz des noch nicht klinisch zugelas-senen intravaskulären Kon-trastmittelsMS-325 (Epix Medical, Ing, Cambridge, MA,USA, linkes Bild) erhöht das Kontast-Rausch-Verhältnis (CNR) während eines 3-D doppelt angulierten Scans (hochauflösend, normales Atmen) der rechtenKranzarterie (Pfeile), verglichen mit demScan ohne Kontrastmittel (rechtes Bild). Eswerden repräsentative Zusammenhängemit 3-mm-Schichten gezeigt.

Figure 2High-resolution (0.64 � 0.64 mm2) coronary MRA of a normal volun-teer, acquired in a single 14-heartbeat breath-hold with spiral imag-ing. Long segments of the first order diagonal branches (white ar-rows) are visualized. The LAD (black arrow), ascending aorta (AAo) andright ventricular outflow tract (RVOT) are also labeled. (From [10]. Re-printed with permission.)

Abbildung 2MR-Koronarangiogramm in hoher Auflösung (0.64 � 0.64 mm2) einerNormalperson, das innerhalb eines Atemanhaltemanövers über 14Herzschläge in Spiraltechnik angefertigt wurde. Man sieht lange Ab-schnitte der großen Diagonaläste (weiße Pfeile). Die LAD (schwarzerPfeil), Aorta ascendens (Aao) und der rechtsventrikuläre Ausflusstrakt(RVOT) sind ebenfalls bezeichnet. (Aus [10] wiedergegeben mitGenehmigung).

CNR = 18.3 CNR = 9.3



Visualization of Normal Coronary ArteriesThe first goal in the evolution of coronary MRA was thereliable visualization of the proximal coronary arteriesand several of the initial technique validation studieswere targeted towards this direction. The reported visu-alized vessel length varies greatly in different reports;however, most recent studies report complete visualiza-tion of the left main coronary artery, at least 4 to 5 cm ofthe proximal left anterior descending and right coro-nary arteries, and approximately 3 cm of the left circum-flex coronary artery [22, 28, 32, 41, 44, 47]. It was quick-ly demonstrated that in normal volunteers coronaryMRA provides vessel diameter measures that are simi-lar to pathology-derived values [13, 20, 32] but also cor-related well with contrast angiography measurements[51]. Quantitation of coronary artery diameter stenosisin areas of focal disease has not been performed. Aswith peripheral MRA, it is likely that coronary MRAmay also overestimate the degree of stenoses.

Current Clinical ApplicationsPresently, the best clinical application of coronaryMRA is for evaluation of the origin and proximal courseof known or suspected anomalous coronary arteries. Al-though relatively uncommon (occurring in < 1% of car-diac catheterization procedures), coronary anomalies inwhich the anomalous vessel travels between the aortaand main pulmonary artery are associated with in-creased mortality. Coronary MRA has advantages overconventional X-ray angiography, because MR imagingis inherently a tomographic technique that allows sur-veying any image plane, allowing for optimal definitionof the anatomic relationships of the coronary vesselswith surrounding structures (Figure 4). Presently, therehave been 4 studies reported [35, 45, 59, 62] collectivelyinvolving 65 patients in whom coronary MRA was com-pared to conventional contrast angiography for theevaluation of the origin and course of anomalous coro-nary arteries. These studies have demonstrated that cor-onary MRA was superior to X-ray angiography, sug-gesting that coronary MRA may be the new “goldstandard” for assessment of this condition.

Evaluation of Coronary Artery DiseaseThe accurate detection of native vessel coronary ste-noses is the long-sought application of coronary MRA.However, current spatial resolution of coronary MRAremains inferior to that available by conventional X-ray angiography. Despite the present limitations, vari-

ous coronary MRA techniques have been shown tocorrectly identify proximal coronary stenoses in sever-al clinical series [15, 33, 38, 43, 46, 47]. With the use ofgradient-echo sequences, focal stenoses present as sig-nal voids from spin-dephasing related to turbulentflow at the site of stenosis. Signal voids are not pathog-nomonic for the atherostenotic disease. Vessel tortuos-ity with deviation of the vessel outside the image plane,susceptibility artifacts (e. g., from focal calcifications),motion and other artifacts may also cause focal signalloss, which may be difficult to differentiate from a focalstenosis.

The first study comparing coronary MRA with con-ventional angiography was reported by our group in

Figure 4Anomalous right coronary artery (open arrow), originating from theleft coronary cusp (arrowhead). Reformatted image to show the originand proximal course of the left and right coronary arteries (AAo = as-cending aorta, LM = left main, LAD = left anterior descending coronaryartery, LCX = left circumflex coronary artery).

Abbildung 4Anomaler Verlauf einer rechten Koronararterie (offener Pfeil), die ausdem links koronaren Sinus entspringt (Pfeilspitze). ReformatiertesBild, das den Ursprung und den proximalen Verlauf der linken undrechten Koronararterien zeigt (AAo = Aorta ascendens, LM =Hauptstamm der linken Kranzarterie, LAD = Ramus interventricularisanterior, LCX = Ramus circumflexus).

1993 [33], using a segmented k-space breath-hold 2-Dapproach. Coronary MRA was shown to have high sen-sitivity and specificity for detecting significant (> 50%diameter stenosis) coronary disease. Subsequent stud-ies [15, 38, 43, 46, 47] have reported variable sensitivityand specificity values for the detection of coronary ar-tery disease. The apparent discrepancy is likely due todifferences in technique (scanner platform, imaging se-quence), irregular rhythms and different patient inclu-sion criteria. Recently, coronary MRA was shown tocorrectly identify patency of the infarct-related arteryand presence of collateral blood flow, in patients whohave suffered a recent myocardial infarction [23]. New-er non-breath-hold and breath-hold approaches for 3-Dcoronary MRA have demonstrated the ability of thistechnique to detect coronary stenoses [3, 57, 68] (Fig-

ures 5 and 6). A navigator-based approach has been im-plemented at multiple sites worldwide that are partici-pating in an international multicenter study, which isnow under way (Figure 7). This study will evaluate theaccuracy of high-resolution free-breathing 3-D coronaryMRA in patients referred for X-ray coronary angiogra-phy. The results of this and similar studies will likely de-fine the near-term clinical role of coronary MRA.

Coronary MRA Following RevascularizationThe presence of metallic objects in the imaging field re-sults in local susceptibility artifacts. Imaging of patientswith intracoronary stents has been shown to be safe, asneither motion [52] nor local tissue warming [18] is ofparticular concern. However, the local image artifactfrom intracoronary stents makes the visual assessment



Figure 5Coronary MRA (panels A and B) and corre-sponding X-ray angiogram (panel C) in a pa-tient with RCA disease. Focal stenosespresent in the proximal and mid portion ofthe vessel (arrows) are readily detected assignal loss on the coronary MRA. (From [57].Reprinted with permission from Elsevier Sci-ence Inc.)

Abbildung 5MR-Koronarangiographie (Abbildungen Aund B) und dazu gehöriges konventionellesRöntgenangiogramm (Abbildung C) beieinem Patienten mit erkrankter rechterKranzarterie. Fokale Stenosen sind in denproximalen und mittleren Abschnitten desGefäßes zu sehen (Pfeile) und stellen sich alsZonen mit Signalverlust im MR-Koronaran-giogramm dar (aus [57], wiedergegeben mitGenehmigung von Elsevier Science Inc.)

Figure 6Sequential focal stenoses of the right coro-nary artery (arrows), shown on the X-ray an-giogram (left panel) and the correspondingcoronary MRA (right panel). The methodo-logy implemented is a 3-D breath-hold se-quence (VCATS) [68]. (By courtesy of PiotrWielopolski, PhD.)

Abbildung 6Sequenzielle fokale Stenosen der rechtenKranzarterie (Pfeile) im konventionellen Röntgenangiogramm (linke Abbildung) unddas dazugehörige MR-Koronarangiogramm(rechte Abbildung). Methodisch wurde eine 3-D-Sequenz in Atemanhaltetechnik verwen-det (VCATS) [ 68] . (Mit freundlicher Genehmi-gung von Piotr Wielopolski, PhD).



of in-stent stenoses impossible and until MR-“friendly”stents are available, the presence of coronary stents willremain an impediment for evaluation of coronary ste-noses in this population.

Aortocoronary bypass grafts, including both saph-enous vein and internal mammary conduits, are rela-tively easy to image due to their relatively stationaryposition and larger vessel diameter, compared to na-tive coronary arteries. Furthermore, their predictableand less convoluted course has allowed MR imagingeven with more conventional techniques for over adecade. Both ECG-gated spin-echo [19, 24, 48, 66] andgradient-echo [2, 19, 67] techniques have been used toassess bypass graft patency. By visualizing rapid bloodflow in at least 2 locations along its expected course,one can conclude that there is flow present throughthe bypass graft, suggestive of patency. If the graft can

only be visualized at 1 level, the graft is considered“indeterminate”, and if the graft cannot be visualizedat all, it is considered to be occluded. The use of con-trast-enhanced coronary MRA may further enhancethe accuracy for assessment of graft patency [11, 19,61, 63].

The clinically relevant question, however, is oftennot merely whether a bypass graft is patent or not. Dis-tal stenoses of the native coronary arteries beyond thegraft touchdown and focal graft stenoses could both ac-count for clinically significant ischemia and may requirespecific therapies, even if the graft is deemed “patent”.Additional limitations of bypass graft imaging includethe local signal loss/artifact associated with implantedmetallic objects such as hemostatic clips, graft-locatingrings, sternal wires, coexistent prosthetic valves and sup-porting struts or rings.

Figure 7Representative coronary artery images obtained from volunteers at multiple MR centers participating in an international multicenter study eval-uating the accuracy of coronary MRA (compared against X-ray angiography), using a free-breathing T2-pre-pulse navigator approach [3, 57].

Abbildung 7Repräsentative Koronarbilder, die von Normalpersonen an verschiedenen MR-Centern angefertigt wurden. Diese MR-Center nehmen an einerinternationalen Multicenterstudie teil, die die Genauigkeit der MR-Koronarangiographie evaluiert im Vergleich mit der konventionellenRöntgenangiographie. In allen Fällen wurde als Sequenz eine Navigatorsequenz mit T2-Präpuls eingesetzt, die normales Atmen erlaubte [3, 57].

Future PerspectivesWith the development of better hardware (gradients,coils) and software, it is expected that coronary MRAwill allow for imaging of the proximal and mid segmentsof the major coronary vessels in the vast majority of sub-jects. Carefully designed multicenter trials testing thenewer methodologies are likely to dominate over thecoming years thereby defining the clinical role of coro-nary MRA.

SummaryOver the last decade, remarkable advances have oc-curred in the field of coronary MRA. Current strategiesreliably visualize the origin and proximal course of thecoronary arteries in the vast majority of patients and al-low the identification and characterization of anoma-lous coronary arteries. Aortocoronary bypass graft pa-tency can also be readily assessed in many patients. Atpresent, the clinical utility of coronary MRA for evalua-tion of native coronary disease remains unproven. Sev-eral competing approaches are currently being evaluat-ed in multicenter trials. As a clinical tool, coronaryMRA will have to prove that it is fast, accurate and eas-ily accepted by both patients and physicians.

Dr. Danias is supported in part by the Clinical Investigator TrainingProgram, Beth Israel Deaconess Medical Center – Harvard/MITHealth Sciences and Technology, Boston, MA, in collaboration withPfizer Inc.Dr. Manning is an Established Investigator of the American HeartAssociation, Dallas, TX, USA (9740003N).

References1. American Heart Association. Statistical Facts. Dallas, TX: Ameri-

can Heart Association, 1996.2. Aurigemma GP, Reichek N, Axel L, et al. Noninvasive determina-

tion of coronary artery bypass graft patency by cine magnetic res-onance imaging. Circulation 1989; 80:1595–602.

3. Botnar RM, Stuber M, Danias PG, et al. Improved coronary arterydefinition with T2-weighted free-breathing 3D-coronary MRA.Circulation 1999;99:3139–48.

4. Brittain JH, Hu BS, Wright GA, et al. Coronary angiography withmagnetization-prepared T2 contrast. Magn Reson Med 1995;33:689–96.

5. Burstein D. MR imaging of coronary artery flow in isolated and invivo hearts. J Magn Reson Imaging 1991;1:337–46.

6. Cho ZH, Mun CW, Friedenberg RM. NMR angiography of coronaryvessels with 2-D planar image scanning. Magn Reson Med1991;20:134–43.

7. Danias PG, Edelman RR, Manning WJ. Magnetic resonance angio-graphy of great vessels and the coronary arteries. In: Pohost G,O’Rourke R, Shah P, Berman D, eds. Imaging in cardiovascular dis-ease. Philadelphia: Lippincott-Raven, 1999.

8. Danias PG, McConnell MV, Khasgiwala VC, et al. Prospective navi-gator correction of image position for coronary MR angiography.Radiology 1997;203:733–6.

9. Danias PG, Stuber M, Botnar RM, et al. Navigator assessment ofbreath-hold duration: impact of supplemental oxygen and hyper-ventilation. AJR Am J Roentgenol 1998;171:395–7.

10. Danias PG, Stuber M, Edelman RR, Manning WJ. Coronary MRA: Aclinical experience in USA. J Magn Reson Imaging 1999;10:713–20.

11. Davis CP, Hany TF, Kaufmann P, et al. Coronary artery bypassgrafts 3D MR angiography. Proc Int Soc Magn Reson Med 1997;2:834.abstract.

12. Dixon WT, Oshinski JN, Trudeau JD, et al. Myocardial suppressionin vivo by spin locking with composite pulses. Magn Reson Med1996;36:90–4.

13. Dodge JT Jr, Brown BG, Bolson EL, et al. Lumen diameter of normalhuman coronary arteries. Influence of age, sex, anatomic varia-tion, and left ventricular hypertrophy or dilation. Circulation 1992;86:232–46.

14. Doyle M, Scheidegger MB, de Graaf RG, et al. Coronary artery imag-ing in multiple 1-sec breath holds. Magn Reson Imaging 1993;11:3–6.

15. Duerinckx AJ, Urman MK. Two-dimensional coronary MR angio-graphy: analysis of initial clinical results. Radiology 1994;193:731–8.

16. Edelman RR, Manning WJ, Burstein D, et al. Coronary arteries:breath-hold MR angiography. Radiology 1991;181:641–3.

17. Foo TKF, Ho VB, King KF. Three-dimensional double-oblique coro-nary artery MR imaging using real-time respiratory navigator andlinear phase shift processing. Proc Int Soc Magn Reson Med1998;1:323.abstract.

18. Friedrich MG, Kivelitz D, Strohm O, et al. Intracoronary stents aresafe during MR imaging of the heart. J Cardiovasc Magn Reson1999;1:258.abstract.

19. Galjee MA, van Rossum AC, Doesburg T, et al. Value of magneticresonance imaging in assessing patency and function of coronaryartery bypass grafts. An angiographically controlled study. Circu-lation 1996;93:660–6.

20. Hoffman MBM, Kerwin TP, Kovacs SJ, et al. Coronary artery diam-eter determination from contrast-enhanced MR angiography;comparison with selective X-ray contrast angiography. Proc IntSoc Magn Reson Med 1998;2:845.abstract.

21. Hofman MB, Henson RE, Kovacs SJ, et al. Blood pool agent strong-ly improves 3D magnetic resonance coronary angiography usingan inversion pre-pulse. Magn Reson Med 1999;41:360–7.

22. Hofman MB, Paschal CB, Li D, et al. MRI of coronary arteries: 2Dbreath-hold vs 3D respiratory-gated acquisition. J Comput AssistTomogr 1995;19:56–62.

23. Hundley WG, Clarke GD, Landau C, et al. Noninvasive determina-tion of infarct artery patency by cine magnetic resonance angio-graphy. Circulation 1995;91:1347–53.

24. Jenkins JP, Love HG, Foster CJ, et al. Detection of coronary arterybypass graft patency as assessed by magnetic resonance imaging.Br J Radiol 1988;61:2–4.

25. Jhooti P, Keegan J, Gatehouse PD, et al. 3D coronary artery imag-ing with phase reordering for improved scan efficiency. Magn Re-son Med 1999;41:555–62.

26. Jhooti P, Wiesmann F, Taylor AM, et al. Hybrid ordered phase en-coding (HOPE): an improved approach for respiratory artifact re-duction. J Magn Reson Imaging 1998;8:968–80.

27. Li D, Dolan RP, Walovitch RC, et al. Three-dimensional MRI of coro-nary arteries using an intravascular contrast agent. Magn ResonMed 1998;39:1014–8.

28. Li D, Kaushikkar S, Haacke EM, et al. Coronary arteries: three-di-mensional MR imaging with retrospective respiratory gating. Ra-diology 1996;201:857–63.

29. Li D, Paschal CB, Haacke EM, et al. Coronary arteries: three-dimen-sional MR imaging with fat saturation and magnetization trans-fer contrast. Radiology 1993;187:401–6.





30. Lieberman JM, Botti RE, Nelson AD. Magnetic resonance imagingof the heart. Radiol Clin North Am 1984;22:847–58.

31. Liu YL, Riederer SJ, Rossman PJ, et al. A monitoring, feedback, andtriggering system for reproducible breath-hold MR imaging.Magn Reson Med 1993;30:507–11.

32. Fat-suppressed breath-hold magnetic resonance coronary angio-graphy. Circulation 1993;87:94–104.

33. Manning WJ, Li W, Edelman RR. A preliminary report comparingmagnetic resonance coronary angiography with conventional an-giography. N Engl J Med 1993;328:828–32.

34. Marks B, Mitchell DG, Simelaro JP. Breath-holding in healthy andpulmonary-compromised populations: effects of hyperventila-tion and oxygen inspiration. J Magn Reson Imaging 1997;7:595–7.

35. McConnell MV, Ganz P, Selwyn AP, et al. Identification of anoma-lous coronary arteries and their anatomic course by magnetic res-onance coronary angiography. Circulation 1995;92:3158–62.

36. McConnell MV, Khasgiwala VC, Savord BJ, et al. Comparison ofrespiratory suppression methods and navigator locations for MRcoronary angiography. AJR Am J Roentgenol 1997;168:1369–75.

37. Meyer CH, Hu BS, Kerr AB, et al. High-resolution multislice spiralcoronary angiography with real-time interactive localization. ProcSoc Int Magn Reson Med 1996;1:439.abstract.

38. Muller MF, Fleisch M, Kroeker R, et al. Proximal coronary arterystenosis: three-dimensional MRI with fat saturation and naviga-tor echo. J Magn Reson Imaging 1997;7:644–51.

39. Oshinski JN, Hofland L, Dixon WT, et al. Magnetic resonance coro-nary angiography using navigator echo gated real-time slice fol-lowing. Int J Card Imaging 1998;14:191–9.

40. Oshinski JN, Hofland L, Mukundan S Jr, et al. Two-dimensional cor-onary MR angiography without breath holding. Radiology 1996;201:737–43.

41. Paschal CB, Haacke EM, Adler LP. Three-dimensional MR imagingof the coronary arteries: preliminary clinical experience. J MagnReson Imaging 1993;3:491–500.

42. Paulin S, von Schulthess GK, Fossel E, et al. MR imaging of the aortic root and proximal coronary arteries. AJR Am J Roentgenol1987;148:665–70.

43. Pennell DJ, Bogren HG, Keegan J, et al. Assessment of coronary ar-tery stenosis by magnetic resonance imaging. Heart 1996;75:127–33.

44. Pennell DJ, Keegan J, et al. Magnetic resonance imaging of coro-nary arteries: technique and preliminary results. Br Heart J 1993;70:315–26.

45. Post JC, van Rossum AC, Bronzwaer JG, et al. Magnetic resonanceangiography of anomalous coronary arteries. A new gold stan-dard for delineating the proximal course? Circulation 1995;92:3163–71.

46. Post JC, van Rossum AC, Hofman MB, det al. Clinical utility of two-dimensional magnetic resonance angiography in detecting coro-nary artery disease. Eur Heart J 1997;18: 426–33.

47. Post JC, van Rossum AC, Hofman MB, et al. Three-dimensional respiratory-gated MR angiography of coronary arteries: compari-son with conventional coronary angiography. AJR Am J Roentge-nol 1996;166:1399–404.

48. Rubinstein RI, Askenase AD, Thickman D, et al. Magnetic reso-nance imaging to evaluate patency of aortocoronary bypassgrafts. Circulation 1987;76:786–91.

49. Sachs TS, Meyer CH, Irarrazabal P, et al. The diminishing variancealgorithm for real-time reduction of motion artifacts in MRI.Magn Reson Med 1995;34:412-22.

50. Sachs TS, Meyer CH, Pauly JM, et al. Coronary angiography usingthe real-time interactive 3D DVA. Proc Int Soc Magn Reson Med1998;1:20.abstract.

51. Scheidegger MB, Vassalli G, Hess OM, et al. Validation of coronaryartery MR angiography: comparison of measured vessel diame-

ters with quantitative contrast angiography. Proc Soc Magn Re-son Med 1994;1:497.abstract.

52. Scott NA, Pettigrew RI. Absence of movement of coronary stentsafter placement in a magnetic resonance imaging field. Am J Car-diol 1994;73:900–1.

53. Sodickson DK, Manning WJ. Simultaneous acquisition of spatialharmonics (SMASH): fast imaging with radiofrequency coil arrays.Magn Reson Med 1997;38:591–603.

54. Sodickson DK, Stuber M, Botnar RM, et al. Accelerated coronaryMR angiography in volunteers and patients using double-oblique3D acquisitions combined with SMASH. J Cardiovasc Magn Reson1999;1:260–1.abstract.

55. Stillman AE, Wilke N, Li D, et al. Ultrasmall superparamagneticiron oxide to enhance MRA of the renal and coronary arteries:studies in human patients. J Comput Assist Tomogr 1996;20:51–5.

56. Stuber M, Botnar R, Danias PG, et al. Contrast agent-enhancedfree-breathing, three-dimensional coronary magnetic resonanceangiography. J Magn Reson Imaging 1999;10:790–9.

57. Stuber M, Botnar RM, Danias PG, et al. Double-oblique free-breathinghigh resolution 3D coronary MRA. J Am Coll Cardiol 1999;34:524–31.

58. Taylor AM, Panting JR, Keegan J, et al. Safety and preliminary find-ings with the intravascular contrast agent NC100150 injection forMR coronary angiography. J Magn Reson Imaging 1999;9:220–7.

59. Taylor AM, Thorne SA, Rubens MB, et al. Coronary artery imagingin congenital heart disease: complementary role of magnetic res-onance and x-ray coronary angiography. J Cardiovasc Magn Reson1999;1:284–5.abstract.

60. Thedens DR, Irarrazaval P, Meyer CH, et al. Three-dimensional cor-onary angiography using partial k-space acquisition and T2-prepcontrast. Proc Int Soc Magn Reson Med 1998;1:321.abstract.

61. Van Rossum AC, Galjee MA, Post JC, et al. A practical approach toMRI of coronary artery bypass graft patency and flow. Int J CardImaging 1997;13:199–204.

62. Vliegen HW, Doornbos J, de Roos A, et al. Value of fast gradientecho magnetic resonance angiography as an adjunct to coronaryarteriography in detecting and confirming the course of clinicallysignificant coronary artery anomalies. Am J Cardiol 1997;79:773–6.

63. Vrachliotis TG, Bis KG, Aliabadi D, et al. Contrast-enhancedbreath-hold MR angiography for evaluating patency of coronaryartery bypass grafts. AJR Am J Roentgenol 1997;168:1073–80.

64. Wang Y, Christy PS, Korosec FR, et al. Coronary MRI with a respira-tory feedback monitor: the 2D imaging case. Magn Reson Med1995;33:116–21.

65. Weiger M, Scheidegger MB, Pruessmann KP, et al. Cardiac real-time acquisition using coil sensitivity encoding. Proc Int Soc MagnReson Med 1998;2:803.abstract.

66. White RD, Caputo GR, Mark AS, et al. Coronary artery bypass graftpatency: noninvasive evaluation with MR imaging. Radiology1987;164:681–6.

67. White RD, Pflugfelder PW, Lipton MJ, et al. Coronary artery bypassgrafts: evaluation of patency with cine MR imaging. AJR Am JRoentgenol 1988;150:1271–4.

68. Wielopolski PA, van Geuns RJ, de Feyter PJ, et al. Breath-hold coro-nary MR angiography with volume-targeted imaging. Radiology1998;209:209–19.

Address for Correspondence:Peter G. Danias, MD, PhD,Beth Israel Deaconess Medical Center,330 Brookline Avenue,Boston, MA 02215, USA,Phone (+1/617) 667-4842, Fax -4833,e-mail: [email protected]

Documents

Coronary MR Angiography: Current Status