| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Cardiac Imaging |
1 From the Departments of Cardiology, Thoraxcenter (R.J.M.v.G., B.J.W.M.R., M.H., P.J.d.F.) and Radiology, Daniel den Hoed Kliniek (P.A.W., H.G.d.B., P.M.A.v.O., M.O.), University Hospital Rotterdam, Groene Hilledijk 301, 3075 EA Rotterdam, the Netherlands. From the 1998 RSNA Scientific Assembly. Received July 28, 1999; revision requested September 1; revision received February 17, 2000; accepted April 4. Address correspondence to R.J.M.v.G. (e-mail: vangeuns@card.azr.nl).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Thirty-eight patients referred for elective coronary angiography were included. The coronary arteries were localized during single-breath-hold, three-dimensional imaging of the entire heart. MR coronary angiography was then performed along the major coronary branches with a double-oblique, three-dimensional, gradient-echo sequence. Conventional coronary angiography was the reference-standard method.
RESULTS: Adequate visualization was achieved with MR coronary angiography in 85%–91% of the proximal coronary arterial branches and in 38%–76% of the middle and distal branches. Overall, 187 (69%) of 272 segments were suitable for comparison between conventional and MR coronary angiography. The diagnostic accuracy of MR coronary angiography for the detection of hemodynamically significant stenoses was 92%; sensitivity, 68%; and specificity, 97%. The sensitivity in individual segments was 50%–77%, whereas the specificity was 94%–100%.
CONCLUSION: Adequate visualization of the major coronary arterial branches was possible in the majority of patients. The observed accuracy of MR coronary angiography for detection of hemodynamically significant coronary arterial stenosis is promising, but it needs to be higher before this modality can be used reliably in a clinical setting. Supplemental material: http://radiology.rsnajnls.org/cgi/content/full/217/1/270/DC1
Index terms: Coronary angiography, 548.1244 • Coronary vessels, MR, 548.121412, 548.121415, 548.121417, 548.12142 • Coronary vessels, stenosis or obstruction, 548.76 • Magnetic resonance (MR), vascular studies, 548.12142
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
MR Imaging
The studies were performed with a 1.5-T whole-body MR imaging system (Magnetom Vision; Siemens, Erlangen, Germany). Patients were placed in a supine position, and a four-channel quadrature phased-array body coil was placed over the thorax. Electrocardiographic electrodes were always set over the anterior part of the thorax and readjusted, if necessary, to obtain reasonable amplitude and clean signal trace in the monitoring unit after the patient was placed inside the magnet bore. After patient positioning, the magnetic field homogeneity over the thorax and heart was assessed to obtain uniform fat suppression in sequences that involved chemical shift fat suppression. This assessment was performed by comparing two three-plane, single-shot spin-echo train imaging (HASTE; Siemens) localizer sequences with and without a chemical shift fat suppression pulse applied. When the reduction of fat signal intensity was deemed inadequate, the shim currents were manipulated and the single-shot spin-echo train imaging localizer sequence with chemical shift fat suppression was repeated until satisfactory results were obtained.
To start the coronary localization procedure, a 3D single-breath-hold, multishot, segmented, echo-planar sequence was used to image the entire heart, with a 120-mm section obtained at end expiration. The data were subjected to multiplanar evaluation to determine the optimal imaging planes for the major coronary arterial branches (Table 1, Fig 1). This localization process has been described in detail previously (12). Seven orientations were obtained with multiplanar reformation before imaging with the breath-hold VCATS protocol proceeded.
|
|
The breath-hold quality was assessed on every attempt by observing the ghost artifacts of the anterior thoracic wall over the heart. In case ghost artifacts appeared, repeated imaging of a particular volume was allowed. If severe ghost artifacts were present and the patient could not hold his or her breath adequately within four attempts, he or she was regarded as nonsuitable for the imaging procedure and thus excluded from the study.
To improve coronary arterial depiction on the volume localizer image, superparamagnetic iron oxide particles (AMI-25 [Endorem]; Laboratoire Guerbet, Roissy, France) were administered as a suspension containing 89.6 mg of iron (11.2 mg/mL) diluted in 100 mL of isotonic glucose solution. The solution was administered in a slow drop infusion for 30 minutes.
Conventional Coronary Angiography
All subjects underwent selective coronary arterial angiography by means of the Judkins technique (13) within 1 month before or after the MR examination. Two experienced cardiologists (B.J.W.M.R., P.J.d.F.) jointly interpreted the angiograms. The coronary tree was divided into proximal, middle, and distal segments according to American Heart Association guidelines (14). Under these guidelines, the left main coronary artery has a single segment. The proximal segment of the left anterior descending artery extends from the bifurcation to the first septal branch. The middle segment of the left anterior descending artery extends from the first to the third septal artery, whereas the distal segment extends from the third septal artery to the apex. The left circumflex coronary artery is divided into three segments by the first and second marginal branches. The proximal segment extends from the bifurcation to the first marginal branch, the middle segment extends from the first to the second marginal branch, and the distal segment extends from the second marginal branch to the posterior lateral branch. The proximal right coronary artery extends from the origin to the first large acute marginal branch, the middle segment extends from the first to the third acute marginal branch, and the distal segment extends from the third acute marginal branch to the posterior descending branch. These segments were visually graded as either having no hemodynamically significant disease—that is, less than 50% diameter stenosis—or having hemodynamically significant disease—that is, greater than 50% diameter stenosis. In cases of disagreement, a third cardiologist made the final decision.
MR Image Interpretation
From each targeted volume, 16 source images in a dynamic loop (Movie 1, http://radiology.rsnajnls.org/cgi/content/full/217/1/270/DC1) were analyzed independently by a cardiologist (R.J.M.v.G.) and a radiologist (H.G.d.B.), who were unaware of the cardiac catheterization results. In cases of disagreement, consensus was achieved in a joint session with a third investigator (M.O.). Of all the coronary segments defined in the American Heart Association guidelines, only eight segments—those of the left main artery; proximal and middle left anterior descending arteries; proximal and middle left circumflex coronary arteries; and proximal, middle, and distal right coronary arteries—were included in this study. This selection resulted in the evaluation of 272 segments in 34 complete patient studies. These segments were regarded as assessable if overlapping structures (ie, veins, pericardium, and unsuppressed fat), image blurring, and/or ghost artifacts could be distinguished from the vessel itself. The segments included in more than one volume were finally evaluated in the volume with the best image quality. The assessable segments were graded as either not having hemodynamically significant diameter stenosis (<50%) or having hemodynamically significant stenosis (>50%).
To investigate the possibilities of integrating the coronary arterial path within a targeted volume into a single image, data sets were reconstructed by using a volume-rendering program (VOXELVIEW; Vital Images, Minneapolis, Minn) that was run on a dedicated graphic workstation (Indigo2; Silicon Graphics, Mountain View, Calif). Segmentation was required to eliminate unwanted structures and view the coronary segment from any viewing angle and thus better assess the presence of any detected stenoses at the initial review. Volume-rendered data were not evaluated for their additional diagnostic value, but rather this technique was used strictly as an exploratory tool to conjecture about the possible likeness of volume-rendered images to the corresponding coronary angiograms.
Statistical Analyses
Conventional coronary angiography served as the standard of reference for determining the diagnostic value of MR coronary angiography. The diagnostic value for the detection of hemodynamically significant (>50% diameter) stenosis in a segment was expressed in sensitivity, specificity, positive predictive, negative predictive, and diagnostic accuracy values.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
The minor technical difficulties included problems performing good electrocardiographic tracing in three (9%) patients and mistriggering from imaging gradient–induced interference in some volume orientations in 30 (8.8%) of 340 measurements. Incomplete fat suppression within the set of volumes collected per patient was always present to a certain degree, and it was prevalent in the distal portion of the left anterior descending artery. Operational problems included input errors in the volume orientation prescription in eight (2.4%) of 340 measurements. Patient-related problems were those due to an inconsistent breath-hold position with respect to the volume localizer in 17 (5.0%) of 340 measurements and incomplete acquisition of all targeted volumes in two patients (five [1.5%] of 340 measurements).
Because of the fast feedback on the data acquired, corrections were directly applied to acquire the desired volume; this resulted in 100% data collection in the segments studied. The mean (± SD) breath-hold time per acquired targeted volume was 23 seconds ± 4. The acquisition of the volume localizer image, selection of the optimal plane location and orientation of the targeted volumes with multiplanar reformation, and acquisition of the targeted-volume images were consistently completed in less than 30 minutes. Of the 272 possible coronary arterial segments in the 34 patients, 187 (69%) were deemed to be assessable by using MR coronary angiography. The range of assessability varied substantially, from 91% for the left main coronary artery to 38% for the middle left circumflex coronary artery. The resultant data on the assessability of individual segments are listed in Table 2.
|
|
|
|
|
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
In this work, we evaluated our initial clinical experience with a breath-hold VCATS protocol that aims to address some of the known difficulties with MR coronary angiography described earlier. Breath-hold acquisitions reduce breathing artifacts, whereas the use of a small imaging volume yields the data consistency required to evaluate MR coronary angiographic data sets for specific coronary segments. In addition, with the 3D nature of the VCATS measurement sequence described, there is the potential to compensate for the SNR loss in faster acquisition scenarios. Nonetheless, all MR coronary angiography techniques must account for the superposition of motion of the coronary vessels during cardiac contraction, which leads to the loss of image detail, and the resolution and SNR deficiencies that are inherent to MR imaging and limit the accurate depiction of the small caliber of the coronary vessels and possible stenoses. Because images are formed from composite data collected during many heartbeats, slow or sudden alterations in the cardiac rhythm during the acquisitions can lead to additional loss of vessel detail.
Breath-hold MR coronary angiography was possible in the majority of patients in this study. Patients with known severe pulmonary disease were excluded in the selection process; this resulted in the acquisition of more reproducible data from cooperative individuals. The free-breathing MR coronary angiography technique with navigator gating may have been a suitable imaging alternative in the excluded patient population. Claustrophobia and inadequate electrocardiographic tracing—the latter of which is associated with small amplitude of the R wave and prone to interference from imaging gradient activity—were among the factors that resulted in the late exclusion of patients and reduced efficiency during imaging. These limitations may improve with the availability of short-bore dedicated cardiac MR imaging systems and optical-transmission electrocardiographic electrodes (15,16).
We believe that in this study with the described MR coronary angiography technique, we achieved a promising level of sensitivity and specificity in the detection of hemodynamically significant coronary arterial stenosis in the symptomatic patient group that was selected. Such patients have a high preexamination likelihood of having coronary arterial stenosis. In general, it is impossible to predict the sensitivity and specificity of a diagnostic examination in patients who are less strictly selected. In addition, the results could have changed substantially with the exclusion of nonassessable segments. This was hinted at by the number of stenoses detected by using VCATS—only 21 (51%) of the 41 stenoses detected in total at conventional angiography.
Additional points regarding the study setup must be clarified. The MR sequence selected for VCATS is not without limitations. First, the SNR that is considered adequate for diagnosis limits the submillimeter resolution that is possible with the described sequence (minimum true voxel size for 21 heartbeats, 1.30 x 0.95 x 2.5 mm3). In practice, an image with lower resolution is feasible, and only vessels with a diameter larger than approximately 2 mm are suitable for evaluation. At present, with conventional coronary angiography, a resolution of 0.1 mm can be obtained and stenoses of high (>70%) and moderate (50%–70%) grades can be differentiated. Therefore, with the current resolution setting in the VCATS protocol, no specific grading of MR angiograms can be attempted.
Second, our MR coronary angiography technique and that of many others are proton density weighted in nature (due to the long magnetization recovery period between data acquisitions, the presence of inflow, and imaging with an incremented flip angle series), despite the additional application of magnetization transfer contrast irradiation to improve myocardial and perhaps plaque signal suppression. Therefore, false-negative results (Figure 5) can be expected, with a normal appearance of the coronary segment at MR coronary angiography and complete occlusion on the corresponding conventional coronary angiogram. The effects of magnetization transfer contrast irradiation on atherosclerotic plaque have been investigated by Pachot-Clouard et al (17). This group suggests that magnetization transfer contrast irradiation–induced signal drop occurs in atherosclerotic plaque components with an effect that is more pronounced for the fibrous cap and media than for the lipid core and adventitia. Nonetheless, with the present setup, in which triggered images are used, it is difficult to define the exact contribution of magnetization transfer contrast irradiation in signal attenuation from plaque for different cardiac rates.
|
|
|
|
Visualization of the left circumflex coronary artery is difficult with all MR coronary angiography techniques, and the present protocol cannot completely solve this problem. One problem arises from the lower SNR caused by the relatively large distance between the left circumflex coronary artery and the surface coils placed around the chest wall. In addition, the close relation between the left circumflex coronary artery, the coronary sinus, and the auricle of the left atrium hampers evaluation of the vessel owing to the lack of resolution and insufficient SNR and contrast-to-noise ratio. Furthermore, the application of magnetization transfer contrast irradiation for an improved contrast-to-noise ratio between the myocardium and blood pools is not without consequences. Magnetization transfer contrast irradiation increases the specific absorption ratio in the patient and reduces the SNR that may be achievable in blood without its application.
The application of volume rendering techniques makes it possible to integrate the 3D course of a coronary segment on a single image. This appears to be useful for delineating the coronary anatomy, as demonstrated in Figures 7–9 (Movies 2,3; http://radiology.rsnajnls.org/cgi/content/full/217/1/270/DC1), and helping to identify coronary lesions from any viewing angle. Nonetheless, postprocessing adds appreciable time to the overall examination. Data transfer to a specialized workstation and the preparation of data by segmenting the unwanted structures before acquiring the final volume-rendered image can be time consuming. Manual segmentation requires 5–15 minutes for each volume; therefore, no attempt was made to render all seven volumes collected in each patient. However, the use of improved software may substantially reduce this manipulation time to make this display utility suitable for routine use.
|
|
|
|
In spite of many measurement difficulties, the visualization on and quality of the images in this initial evaluation were encouraging enough to warrant continued, more extensive clinical trials with the proposed VCATS methodology. The proposed protocol is practical for a clinical setup and provides the means for fast assessment of the coronary arteries with acquisition times of less than 30 minutes. We envision that further improvements in the presently used MR pulse sequence that yield better signal-to-noise and contrast-to-noise ratios and better resolution with a shorter breath-hold time will substantially help increase the sensitivity and specificity of the approach. On the basis of theoretical estimates (18), the addition of intravascular contrast media that provide very short T1 relaxation times in blood (<40 msec) can facilitate some of the improvements (19) needed to consider the routine use of a VCATS protocol for the evaluation of coronary arteries. Furthermore, contrast media may improve the differentiation between the vessel lumen and coronary arterial wall by enabling the acquisition of only a luminogram.
In conclusion, the described VCATS protocol makes it possible to localize the major coronary arterial branches in a short examination time, and the observed degree of accuracy in the detection of hemodynamically significant stenoses within these branches is encouraging. The selected measurement sequence for VCATS needs improved SNR and spatial resolution with shorter measurement times to facilitate a more adequate scenario for coronary arterial assessment on a broader scale. We envision that further refinements in the hardware and software in MR cardiac machines and the introduction of T1-efficient intravascular contrast media will considerably augment the dependability of the proposed methodology.
|
FOOTNOTES |
|---|
Author contributions: Guarantors of integrity of entire study, R.J.M.v.G., P.A.W., H.G.d.B., P.J.d.F., M.O.; study concepts and design, R.J.M.v.G., P.A.W., P.J.d.F., M.O.; definition of intellectual content, R.J.M.v.G., P.A.W., P.J.d.F., M.O.; literature research, R.J.M.v.G., P.A.W.; clinical studies, R.J.M.v.G., P.A.W., B.J.W.M.R.; experimental studies, P.A.W., R.J.M.v.G.; data acquisition, R.J.M.v.G., P.A.W., P.M.A.v.O.; data analysis, R.J.M.v.G., P.A.W., H.G.d.B., P.J.d.F., M.H., B.J.W.M.R.; statistical analysis, R.J.M.v.G., M.H.; manuscript preparation and editing, R.J.M.v.G., P.A.W.; manuscript review, R.J.M.v.G., P.A.W., H.G.d.B., P.J.d.F., M.O.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  D. A. Bluemke, S. Achenbach, M. Budoff, T. C. Gerber, B. Gersh, L. D. Hillis, W. G. Hundley, W. J. Manning, B. F. Printz, M. Stuber, et al. Noninvasive Coronary Artery Imaging: Magnetic Resonance Angiography and Multidetector Computed Tomography Angiography: A Scientific Statement From the American Heart Association Committee on Cardiovascular Imaging and Intervention of the Council on Cardiovascular Radiology and Intervention, and the Councils on Clinical Cardiology and Cardiovascular Disease in the Young Circulation, July 29, 2008; 118(5): 586 - 606. [Full Text] [PDF]
|
|
|
|
 |
|
 P. G. Danias, A. Roussakis, and J. P.A. Ioannidis Diagnostic performance of coronary magnetic resonance angiography as compared against conventional x-ray angiography: A meta-analysis J. Am. Coll. Cardiol., November 2, 2004; 44(9): 1867 - 1876. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Schoenhagen, S. S. Halliburton, A. E. Stillman, S. A. Kuzmiak, S. E. Nissen, E. M. Tuzcu, and R. D. White Noninvasive Imaging of Coronary Arteries: Current and Future Role of Multi-Detector Row CT Radiology, July 1, 2004; 232(1): 7 - 17. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
U. J. Schoepf, C. R. Becker, B. M. Ohnesorge, and E. K. Yucel CT of Coronary Artery Disease Radiology, July 1, 2004; 232(1): 18 - 37. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Reiter, U. Reiter, P. Bergmann, and R. Rienmuller MR imaging-based port placement planning for totally endoscopic coronary artery bypass grafting Interactive CardioVascular and Thoracic Surgery, June 1, 2004; 3(2): 341 - 345. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. J. Budoff, S. Achenbach, and A. Duerinckx Clinical utility of computed tomography and magnetic resonance techniques for noninvasive coronary angiography J. Am. Coll. Cardiol., December 3, 2003; 42(11): 1867 - 1878. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C R Peebles Non-invasive coronary imaging: computed tomography or magnetic resonance imaging? Heart, June 1, 2003; 89(6): 591 - 594. [Full Text] [PDF]
|
|
|
|
|
|
M. S. Dirksen, H. J. Lamb, P. Kunz, P. Robert, C. Corot, and A. de Roos Improved MR Coronary Angiography with Use of a New Rapid Clearance Blood Pool Contrast Agent in Pigs Radiology, June 1, 2003; 227(3): 802 - 808. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. E. Langerak, H. W. Vliegen, J. W. Jukema, P. Kunz, A. H. Zwinderman, H. J. Lamb, E. E. van der Wall, and A. de Roos Value of Magnetic Resonance Imaging for the Noninvasive Detection of Stenosis in Coronary Artery Bypass Grafts and Recipient Coronary Arteries Circulation, March 25, 2003; 107(11): 1502 - 1508. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Bogaert, R. Kuzo, S. Dymarkowski, R. Beckers, J. Piessens, and F. E. Rademakers Coronary Artery Imaging with Real-time Navigator Three-dimensional Turbo-Field-Echo MR Coronary Angiography: Initial Experience Radiology, March 1, 2003; 226(3): 707 - 716. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Z. A. Fayad, V. Fuster, K. Nikolaou, and C. Becker Computed Tomography and Magnetic Resonance Imaging for Noninvasive Coronary Angiography and Plaque Imaging: Current and Potential Future Concepts Circulation, October 8, 2002; 106(15): 2026 - 2034. [Full Text] [PDF]
|
|
|
|
 |
|
 B. Giorgi, S. Dymarkowski, F. Maes, M. Kouwenhoven, and J. Bogaert Improved Visualization of Coronary Arteries Using a New Three-Dimensional Submillimeter MR Coronary Angiography Sequence with Balanced Gradients Am. J. Roentgenol., October 1, 2002; 179(4): 901 - 910. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Y. Kim, P. G. Danias, M. Stuber, S. D. Flamm, S. Plein, E. Nagel, S. E. Langerak, O. M. Weber, E. M. Pedersen, M. Schmidt, et al. Coronary Magnetic Resonance Angiography for the Detection of Coronary Stenoses N. Engl. J. Med., December 27, 2001; 345(26): 1863 - 1869. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Achenbach and W. G. Daniel Noninvasive Coronary Angiography -- An Acceptable Alternative? N. Engl. J. Med., December 27, 2001; 345(26): 1909 - 1910. [Full Text] [PDF]
|
|
|
|
