Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2006 May 1.
Published in final edited form as: Magn Reson Med. 2005 Nov;54(5):1094–1099. doi: 10.1002/mrm.20699

Comparison of X-ray Fluoroscopy and Interventional MRI for the Assessment of Coronary Artery Stenoses in Swine

Jordin D Green 1, Reed A Omary 2,3, Brian E Schirf 2, Richard Tang 2, Biao Lu 2, James A Gehl 2, J Jenny Huang 4, James C Carr 2, F Scott Pereles 2, Debiao Li 2,3,✉
PMCID: PMC1343514  NIHMSID: NIHMS4089  PMID: 16217784

Abstract

The accuracy of a two-step interventional MRI protocol to quantify coronary artery disease was compared to the clinical gold standard, X-ray angiography. Studies were conducted in 9 swine with a surgically induced stenosis in the proximal left circumflex coronary artery. The two-step protocol consisted of catheter-directed magnetic resonance angiography (MRA), which was first used to localize the stenosis, followed by MRI cross-sectional images to quantify the degree of stenosis without use of contrast agent. Line signal intensity profiles were drawn across the vessel diameter at the stenosis site and proximal to the stenosis for each data set to measure percent stenosis for each animal. Catheter-directed MRA successfully detected 8/9 stenoses. Cross-sectional MRI accurately quantified each stenosis, with strong agreement to the measurements made using X-ray fluoroscopy (Intra-class correlation coefficient = 0.955; p < 0.05). This study demonstrates that in the future interventional MRI may be an alternative to X-ray angiography for the detection and quantification of coronary artery disease.

Keywords: Interventional MR, Fluoroscopy, Coronary angiography, comparative studies

INTRODUCTION

Recent studies have proposed catheter-directed magnetic resonance angiography (MRA) as a potential alternative to X-ray fluoroscopy for the detection and treatment of coronary artery stenoses (1,2). MRI has several potential advantages over X-ray fluoroscopy: MRI 1) does not expose the patient or medical team to ionizing radiation or nephrotoxic contrast agents; 2) allows for two-dimensional (2D) and three-dimensional (3D) imaging with arbitrary slice orientation and thickness; and 3) can obtain not only anatomic information but also physiological information, such as myocardial perfusion and flow (3).

Catheter-directed projection coronary MRA is one of the methods proposed for disease diagnosis in an interventional MRI (iMRI) setting. Typically, this is done by injecting diluted contrast media through a catheter that has engaged the coronary ostium or proximal coronary artery, leading to a strong, localized enhancement of the vessel(s) of interest. Because of the constraints on imaging time following an intra-coronary injection (4), most studies to date have used 2D projection imaging to obtain sufficient coverage and in-plane spatial resolution (4–7). Initial studies used a 2D gradient-recalled echo (GRE) sequence with a high flip angle without electrocardiographic (ECG) triggering to rapidly image the left circumflex (LCX) coronary artery following injection of gadolinium (Gd) (1), and more recent studies have demonstrated improved vessel SNR using a steady-state with free precession (SSFP) sequence (4). To date no studies have directly compared the ability of iMRI and X-ray fluoroscopy to detect and quantify coronary artery stenoses. Omary et al. compared catheter-directed MRA to X-ray angiography in swine with surgically-induced bilateral stenoses in the renal arteries and found no difference in detection accuracy between the two modalities (8). However, catheter-directed coronary MRA is more challenging than renal MRA.

It was noted earlier that unlike X-ray fluoroscopy, MRI has the ability to obtain images in any arbitrary slice orientation. With 2D SSFP, it should be possible to acquire coronary artery cross-section images with relatively high spatial resolution and/or SNR without using contrast agent. However, using cross-section images to localize disease over the entire coronary artery is time-consuming and impractical. Therefore, we propose a two-step protocol to increase the detection accuracy of coronary artery disease with iMRI: 1) Stenosis detection: Catheter-directed contrast-enhanced MRA is used for global survey of the artery to identify potential stenoses; and 2) Stenosis quantification: Vessel lumen cross-sections at any “trouble spots” can be used to measure lumen diameter and determine treatment options.

The purpose of this study was to compare the accuracy of X-ray fluoroscopy and iMRI for the assessment of coronary artery stenoses in an animal model with surgically induced coronary artery stenoses. Our proposed iMRI protocol will consist of a combination of 3D catheter-directed MRA and 2D cross-sectional imaging at the stenosis site, with the former being used to detect any potential stenoses and the latter being used to accurately assess the severity of the stenosis. The hypothesis is that this combined MRA/cross-section protocol will have similar accuracy to X-ray fluoroscopy for the detection of coronary artery disease.

MATERIALS AND METHODS

Surgical Preparation

Our institutional animal care and use committee approved all experiments, which were conducted in 9 domestic swine (mean weight = 34 kg).

We surgically induced coronary artery stenoses by placing ameroid constrictors (2.25–2.50 mm inner diameter; Research Instruments SW, Escondido, CA) around the proximal LCX. For the surgical procedure, we first tranquilized the animals with a combination of ketamine, xylazine, and atropine and then administered isoflurane for general anesthesia. A lateral thoracotomy was performed, followed by dissection of the pericardium to expose the LCX. Silk suture was placed around the circumference of the constrictor to prevent movement of the ameroid and reduce outward expansion. This surgical model was chosen because progressive constriction of the ameroid provides a reproducible animal model of chronic coronary artery stenosis (9).

X-Ray Digital Subtraction Angiography

Three weeks later, the animals were transported to the X-ray fluoroscopy suite. This time period was long enough to induce a hemodynamically significant stenosis (> 50% diameter reduction), while minimizing the risk for coronary artery occlusion (10). Swine were tranquilized using an intramuscular injection of a combination of ketamine (10–15 mg/kg), xylazine (1–2 mg/kg), and atropine (0.05 mg/kg). Isoflurane was then administered for general anesthesia.

Using sterile technique, 6- or 7-French conventional vascular sheaths were percutaneously placed in the femoral artery under ultrasound guidance (SiteRite, Bard, Murray Hill, NJ). Intravenous (IV) heparin was administered for anticoagulation, IV lidocaine was administered for prophylaxis of arrhythmias, and transdermal nitropaste was given to prevent vasospasm. Under X-ray fluoroscopic guidance (PowerMobil, Siemens Medical Solutions, Erlangen, Germany), a standard 6-French coronary catheter (Cook, Bloomington, IN) was advanced into the aortic arch and the left main coronary artery was engaged. X-ray imaging of the left coronary artery was performed in frontal, 55° left anterior oblique, and 30° right anterior oblique projections using iodinated contrast agent. Each X-ray angiogram was performed using 10 mL of iohoxel (Omnipaque 300, Amersham, Princeton, NJ). This angiogram was considered the reference standard for subsequent coronary artery stenosis measurements. The catheter was subsequently removed and each pig transferred to the adjacent 1.5 T Sonata MRI scanner (Siemens Medical Solutions, Erlangen, Germany) for the remaining portion of the experiment.

MRI-guided Intervention

After transfer to the MR scanner, the same clinicians used the existing femoral arterial sheath to advance a loopless antenna guidewire coil (Intercept, Surgi-Vision, Inc., North Chelmsford, MA) and the same angiographic catheter used during X-ray fluoroscopy up to the descending aorta and into the left coronary ostium. The guidewire was tracked using a single-shot SSFP sequence with sliding window reconstruction at a frame rate of 9 frames/s using the guidewire coil and one surface coil (Repetition Time (TR)/Echo Time (TE)/flip angle = 2.9 ms/1.45 ms/70°; matrix = 70 x 128 (phase encoding x readout); field of view (FOV) = 206 x 300 mm2 (phase encoding x readout); slice thickness = 30 mm). Catheter tracking was performed using inversion recovery (IR)-GRE at a frame rate of approximately 7 frames/s using the guidewire antenna for signal detection (TR/TE/flip angle = 2.3 ms/1.2 ms/20°; matrix = 74 x 256, FOV = 206 x 300 mm2, slice thickness = 30 mm). Once the catheter engaged the coronary ostium, a small injection of diluted Gd was used to verify positioning using the same IR-GRE sequence. A more detailed description of this procedure has been described previously (11).

With the catheter in position, 3D thick-partition magnetization prepared SSFP (Fig. 1) was performed with intra-coronary injection of Gd contrast agent (Magnevist, Berlex Laboratories, Wayne, NJ) diluted to 8% by volume. 3D slab orientation (usually an approximation of a 55° left anterior or 30° right anterior oblique orientation) was chosen using a combination of a pre-contrast SSFP localizer and low spatial resolution coronary roadmaps. Magnetization preparation consisted of a 90° saturation pulse, followed by a train of four 180° inversion pulses. The saturation pulse was followed by a delay time TS, and the four inversion pulses were each separated by time TI1, TI2, TI3, and TI4. Magnetization preparation was designed to suppress background tissues (such as fat, myocardium) while maintaining high contrast-enhanced blood signal (12).

Figure 1.

Figure 1

Open in a new tab

SSFP with a saturation pulse-four inversion pulse magnetization preparation. After detection of the R-wave, a 90° saturation pulse is applied, followed by saturation time TS. This is followed by a train of 180° inversion pulses, separated by delay times TI1, TI2, TI3, and TI4. Linear flip angle (LFA) preparation pulses are applied at the end of TI4. Note that TS, TI1-TI4 are not necessarily equal.

The purpose for performing intra-coronary MRA was to detect the presence and determine the location of the stenosis for subsequent cross-sectional imaging. A thick-partition 3D sequence was used to maintain coverage and in-plane spatial resolution in a limited acquisition period. With intra-coronary injections of contrast media, imaging time is limited because of the rapid myocardial perfusion of contrast agent (4). For this reason, we only acquired two thick partitions per 3D volume. This sequence has been shown to offer improved contrast-to-noise ratio (CNR) over a similar 2D sequence while keeping acquisition time low and maintaining adequate in-plane spatial resolution (13).

Typical sequence parameters for this catheter-directed MRA protocol were: TR/TE/flip angle = 3.8 ms/1.5 ms/70°; matrix = 174 x 512; FOV = 200 x 400 mm2; 2 partitions interpolated to 4; slab thickness = 4 cm; bandwidth = 610 Hz/Pixel. 25 lines were collected per cardiac cycle using a centric reordering scheme. TS and TI1 – TI4 (Fig. 1) were 15 ms, 115 ms, 15 ms, 50 ms, and 95 ms respectively based on the results of simulations done previously and empirical observations (14). 5 linear flip angle SSFP preparation pulses were applied immediately before image data acquisition during the final TI period (15). Asymmetric sampling was applied in the readout direction to reduce TR. MRA scans were accompanied by an injection of 6–10 mL of Gd (diluted to 8% by volume (16)) over 4–7 s.

If a stenosis was qualitatively evaluated to be > 30%, thin-slice 2D SSFP cross-sectional images were acquired at the stenosis site and proximal to the stenosis. Because the stenosis had already been identified using catheter-directed MRA, only a single slice at each location was required. The slice orientation was always chosen to be perpendicular to the vessel segment. Additionally, contrast injection was unnecessary because the T2/T1-weighted contrast in SSFP allows for sufficient differentiation between the vessel wall and arterial lumen. For cross-sectional SSFP imaging, the following sequence parameters were used: TR/TE/flip angle = 3.8 ms/1.7 ms/65°; matrix = 256 x 512; FOV = 200 x 400 mm2; slice thickness = 2 mm; bandwidth = 610 Hz/Pixel. 15 lines were collected per cardiac cycle using a linear phase-encoding scheme. Each image consisted of 4 signal averages. Asymmetric sampling was used in the readout direction.

Data Analysis

All data sets (X-ray angiography, catheter-directed MRA, cross-sectional MRI) were transferred to a separate workstation for examination. It was hypothesized that MRA could be used to identify potential stenoses, while MRI cross-sections could accurately quantify disease. To test this hypothesis, the full width at half maximum (FWHM) was calculated and used to calculate the percent stenosis for each of the three data sets (X-ray angiography, MRA, and cross-sectional MRI).

For each animal, the X-ray angiography data set which had the highest image quality was processed as follows: using available software (ImageJ), line profiles were drawn perpendicular to the artery at two locations of interest: the closest unaffected LCX segment adjacent to the stenosis (considered to be “normal” artery) and at the stenosis site. Profile data (image intensity vs. pixel number) was then transferred to a spreadsheet and used to calculate the FWHM of the normal and diseased LCX segments. Percent stenosis was then calculated by taking [1 − (FWHMstenosis/FWHMproximal)] x 100%.

MRA and cross-sectional MRI data was analyzed in a similar fashion. Line profiles were drawn on Maximum Intensity Projections (MIP) of the 3D MRA data sets. For the MRI cross-section data sets, each data set consisted of a pair of images; one from a non-diseased proximal LCX segment and one taken at the stenosis site. For elliptical stenoses, care was taken to draw the profile through the center of each artery, typically using the short axis. The same profile orientation was used for both diseased and normal images.

3D MRA was used to screen subjects for the presence of CAD. Clinically, a stenosis is considered hemodynamically significant if the lumen diameter narrowing is greater than 50% (17). The error associated with stenosis measurements made from MRA data is unknown, but it has been previously shown that the error associated with X-ray fluoroscopy stenosis measurements can be as high as 20% when compared to pathology (18). Based on preliminary data and assuming the error for MRA is at least as large as that for X-ray angiography, a significant stenosis may be present if a lumen diameter reduction of 30% or greater is detected in 3D MRA and further cross-sectional MRI examination is required to accurately quantify the degree of stenosis. This more lenient criterion is acceptable for MRA because it is not intended to be used for rigorous quantification of the stenoses.

MRI cross-sections were used to quantitatively measure the degree of stenosis. A statistical package (SPSS; SPSS Inc., Chicago, IL) was used to calculate the intra-class correlation coefficient (ICC) between the percent stenosis values calculated using X-ray data and the MRI cross-section data. Alpha was set to 0.05.

RESULTS

The left coronary ostium was successfully engaged under X-ray fluoroscopy and X-ray angiography was performed in 9/9 pigs. The animals were then transferred to the MR scanner. Under MR-guidance, the left coronary ostium was successfully engaged in 8/9 pigs. In one animal, the catheter was advanced into the sinus portion of the aortic arch, but we were unable to engage the coronary ostium due to difficulties with our internal coil on that particular day.

3D catheter-directed projection coronary MRA was performed in the 8 animals where catheterization was successful (Fig. 2). This data set was then used to plan MR cross-sectional imaging of the LCX at locations proximal to the stenosis and at the stenosis site. In the one animal where the coronary ostium was not engaged, the pre-contrast LCX localizer was used to determine the proximal and stenosis cross-sectional image planes.

Figure 2.

Figure 2

Open in a new tab

Catheter-directed coronary angiography of a swine with a surgically created stenosis in the proximal LCX (arrow) near the aorta (Ao). a) X-ray fluoroscopy. b) Subtracted MIP of 3D catheter-directed MRA. c) Ex vivo digital photo of the heart of the same animal after the experiment, with the LCX and ameroid constrictor exposed. Also included are higher spatial resolution MRI cross-sections (dashed arrows) d) proximal to the stenosed segment and e) at the stenosed segment, obtained without the administration of contrast agent. In this example, the percent stenosis was measured to be 75% using X-ray angiography, 42% using catheter-directed MRA, and 65% using the MRI cross-sections.

Percent stenosis values measured using catheter-directed MRA were compared to values measured using X-ray fluoroscopy in the 8 animals where 3D MRA was successful. Absolute mean difference between the percent stenosis value calculated using X-ray fluoroscopy and those calculated using MRA was 19%. In all animals under MRA, percent stenosis was measured to be greater than 30% (Fig. 3), so all animals which underwent catheter-directed MRA were further evaluated using MRI cross-sections.

Figure 3.

Figure 3

Open in a new tab

Percent stenosis measured using MRA (▪) and X-ray fluoroscopy (•) for the 8 swine where catheter-directed MRA was successful. The absolute difference between X-ray angiography and MRA measurements ranged from a maximum of 34% to a minimum of 3%, with a mean absolute difference of 19%. Under MRA, all eight animals had an estimated stenosis greater than 30% and were further investigated using MRI cross-sections.

Cross-sectional MRI was performed in all 9 animals (Fig. 4). The comparison between percent stenosis measurements using X-ray fluoroscopy and MRI cross-sections are shown as a scatter plot (Fig. 5). The intra-class correlation coefficient was 0.955 (p<0.05).

Figure 4.

Figure 4

Open in a new tab

a) MIP of a 3D catheter-directed MRA data set showing the LAD and LCX. MRA was used to plan cross-sectional MRI in arterial segments proximal to (solid line) and at (dashed line) the stenosis. b) X-ray angiography of the same animal in a similar orientation as a). c) MRI cross-section of the proximal segment, corresponding to the solid line in a). d) MRI cross-section of the diseased segment, corresponding to the dashed line in a). In this example, the stenosis was measured to be 65% using X-ray angiography, 30% using catheter-directed MRA, and 62% using MRI cross-sections.

Figure 5.

Figure 5

Open in a new tab

Correlation between the percent stenosis measured using cross-sectional MRI vs. percent stenosis measured using X-ray angiography. Solid line represents the line of identity. ICC was measured to be 0.955 (p < 0.05).

DISCUSSION

The goal of this study was to compare the accuracy of iMRI and X-ray fluoroscopy to detect and assess coronary artery stenoses. X-ray fluoroscopy, with its ability to image at 0.1 x 0.1 mm2 in-plane spatial resolution and frame rates of up to 30 frames/s, is the clinical gold standard for the detection of coronary artery disease. We hypothesized that a two-step iMRI procedure, where catheter-directed MRA was used to identify areas of potential disease, and MRI cross-sections at the stenosis site were used to quantify stenoses, would have similar accuracy to X-ray fluoroscopy.

As further technical improvements are made, catheter-directed MRA may be sufficient to detect the presence of disease by itself. However, at present time spatial resolution and signal are insufficient to consistently measure coronary artery disease accurately. Quantifying disease is important because different clinical pathways are taken depending on the severity of stenosis. At the same time, MRI cross-sections can take up to 20 heartbeats per scan, and given their small slice thickness (2 mm), it would be exceptionally time consuming to use them for both stenosis localization and evaluation. However, the combined MR-guided two-step process, where catheter-directed MRA is used to localize disease, and MRI cross-sections are used for quantitative analysis, was extremely effective. We were able to successfully localize the LCX stenosis in all animals. The agreement between the X-ray fluoroscopy stenosis quantification values and the cross-sectional MRI values was very convincing (ICC = 0.955; > 0.8 is generally considered a strong correlation), and was also statistically significant (p < 0.05).

On its own, catheter-directed coronary MRA agreed well with X-ray angiography for some animals, but had poor agreement in others. MR-guided coronary catheterization is still under development and is more challenging than under X-ray guidance. This difficulty is due to lack of established MR device tracking protocols, the scarcity of interventional devices specifically designed for MR-guided interventions, and reduced temporal resolution for real-time device tracking under MR guidance. These factors could have affected SNR in some of the MRA data sets.

MRI cross-sectional stenosis measurements agreed remarkably well with X-ray stenosis measurements, and the y-intercept was not statistically different from zero. Cross-sectional images depict the 2D luminal areas instead of projections of the lumen as in MRA and X-ray angiography and therefore are potentially more accurate in defining the lumen diameter and area. Because of the natural contrast properties of SSFP, contrast injections were unnecessary to differentiate between vessel lumen and wall. Imaging time was limited only by breath-hold duration, and could have been extended even further if navigator-echo sequences had been used to compensate for respiratory motion. This allowed for an increase in spatial resolution compared to the MRA sequence and also allowed for the collection of a number of signal averages, which helped improve coronary artery SNR.

Though SSFP has high imaging efficiency, as TR increases (as typically happens as spatial resolution increases), sensitivity to off-resonance also increases, which can result in susceptibility artifacts in the final image. In future studies, it may be beneficial to take advantage of the increase in available imaging time for coronary cross-sectional imaging to raise spatial resolution and use alternative imaging sequences, such as spoiled GRE or spin echo, that do not suffer as severely from off-resonance artifacts at long TRs. Several of these sequences successfully differentiate plaque components in carotid and coronary arteries (19–22), and could be used to provide additional information for diagnosis.

In this study, MRI cross-sections consistently underestimated the degree of stenosis compared to X-ray fluoroscopy (Fig. 6). This suggests systematic errors in either the MRI cross-section measurements or the X-ray angiography measurements. A previous study indicated that X-ray angiography can overestimate and underestimate disease (23). Because X-ray angiography is only the clinical gold standard, a future study should compare MRI and X-ray to another reference standard such as pathology or intravascular ultrasound.

Further experiments are needed to evaluate our two-step protocol in a different animal model of coronary stenosis. Even though the LCX was visibly stenotic in most animals, ameroid-induced low-signal surrounding the artery made it easier to localize the stenosis using MRI. Ideally, it would be more realistic to test our technique in animals with stenoses produced using reverse cable-ties (24) which do not produce a signal void, or in animals with atherosclerosis created through diet and/or wall injury, which better mimics human atherosclerosis. These alternative animal models might also allow for stenosis measurements pre- and post- MR-guided balloon angioplasty (25) or stent placement (2). Ultimately patient studies are required to assess the effectiveness of the technique.

Additionally, the choice of a threshold of 30% to “refer” the subject for evaluation using cross-sectional MRI was somewhat arbitrary. Image review by trained radiologists would have been more appropriate, but was not possible with the chosen animal model because of the signal void around the stenosis which was sometimes created by the ameroid constrictor.

In conclusion, this study has validated a two-step MRI protocol that can potentially detect and quantify coronary artery stenoses in swine. Catheter-directed MRA successfully detected stenoses in all cases where MR-guided coronary catheterization was successful, and cross-sectional MRI has a similar accuracy to X-ray fluoroscopy once the stenosis is localized. This technique may be useful in future MR-guided coronary artery interventions.

Footnotes

Supported in part by National Institute of Health grant no. HL 70859 and HL 38698. RAO was supported in part by NIH K08 DK60020.

References

  • 1.Serfaty JM, Yang X, Aksit P, Quick HH, Solaiyappan M, Atalar E. Toward MRI-guided coronary catheterization: visualization of guiding catheters, guidewires, and anatomy in real time. J Magn Reson Imaging. 2000;12:590–594. doi: 10.1002/1522-2586(200010)12:4<590::aid-jmri11>3.0.co;2-3. [DOI] [PubMed] [Google Scholar]
  • 2.Spuentrup E, Ruebben A, Schaeffter T, Manning W, Gunther R, Buecker A. Magnetic resonance-guided coronary artery stent placement in a swine model. Circulation. 2002;19:874–879. doi: 10.1161/hc0702.104165. [DOI] [PubMed] [Google Scholar]
  • 3.Kim R, Fieno D, Parrish T, Harris K, Chen E, Simonetti O, Bundy J, Finn J, Klocke F, Judd R. Relationship of MRI delayed contrast enhancement to irreversible injury, infarct age, and contractile function. Circulation. 1999;100:1992–2002. doi: 10.1161/01.cir.100.19.1992. [DOI] [PubMed] [Google Scholar]
  • 4.Green JD, Omary RA, Schirf BE, Tang R, Li D. Catheter-directed contrast-enhanced coronary MR angiography in swine using magnetization-prepared True-FISP. Magnetic Resonance in Medicine. 2003;50:1317–1321. doi: 10.1002/mrm.10642. [DOI] [PubMed] [Google Scholar]
  • 5.Serfaty JM, Atalar E, Declerck J, Karmakar P, Quick HH, Shunk KA, Heldman AW, Yang X. Real-time projection MR angiography: feasibility study. Radiology. 2000;217:290–295. doi: 10.1148/radiology.217.1.r00se42290. [DOI] [PubMed] [Google Scholar]
  • 6.Tsekos N, Woodard P, Foster G, Moustakidis P, Sharp T, Herrero P, Gropler R. Dynamic coronary MR angiography and first-pass perfusion with intracoronary administration of contrast agent. J Magn Reson Imaging. 2002;16:311–319. doi: 10.1002/jmri.10161. [DOI] [PubMed] [Google Scholar]
  • 7.Green JD, Omary RA, Finn JP, Chung YC, Tang R, Li YZ, Vasireddy S, Li D. Two- and three-dimensional MR coronary angiography with intraarterial injections of contrast agent in dogs: a feasibility study. Radiology. 2003;226:272–277. doi: 10.1148/radiol.2261011848. [DOI] [PubMed] [Google Scholar]
  • 8.Omary RA, Henseler KP, Unal O, Maciolek LJ, Finn JP, Li D, Nemcek AAJ, Vogelzang RL, Grist TM. Comparison of intraarterial and intravenous gadolinium-enhanced MR angiography with x-ray digital subtraction angiography for the detection of renal artery stenosis in pigs. Am J Roentgenol. 2002;178:119–123. doi: 10.2214/ajr.178.1.1780119. [DOI] [PubMed] [Google Scholar]
  • 9.Inou T, Tomoike H, Watanabe K, Kikuchi Y, Mizukami M, Kurozumi T, Nakamura M. A newly developed X-ray transparent ameroid constrictor for study on progression of gradual coronary stenosis. Basic Res Cardiol. 1980;75:537–543. doi: 10.1007/BF01907835. [DOI] [PubMed] [Google Scholar]
  • 10.Prasad PV, Kim D, Kaiser AM, Chavez D, Gladstone S, Li W, Buxton RB, Edelman RR. Noninvasive comprehensive characterization of renal artery stenosis by combination of STAR angiography and EPISTAR perfusion imaging. Magn Reson Med. 1997;38:776–787. doi: 10.1002/mrm.1910380514. [DOI] [PubMed] [Google Scholar]
  • 11.Omary RA, Green JD, Schirf BE, Li Y, Finn JP, Li D. Real-time magnetic resonance imaging-guided coronary catheterization in swine. Circulation. 2003;107:2656–2659. doi: 10.1161/01.CIR.0000074776.88681.F5. [DOI] [PubMed] [Google Scholar]
  • 12.Mani S, Pauly J, Conolly S, Meyer C, Nishimura D. Background suppression with multiple inversion recovery nulling: applications to projective angiography. Magn Reson Med. 1997;37:898–905. doi: 10.1002/mrm.1910370615. [DOI] [PubMed] [Google Scholar]
  • 13.Green JD, Omary RA, Schirf BE, Tang R, Carr JC, Li D. Three-dimensional contrast-enhanced SSFP for improved catheter-directed coronary MR angiography. J Magn Reson Imaging, in press. [DOI] [PMC free article] [PubMed]
  • 14.Green JD, Omary RA, Schirf BE, Li D. Optimization of a multiple inversion recovery magnetization preparation for thick slice projection imaging in interventional MRI. In: Proceedings of the 11th Annual Meeting of ISMRM, Toronto, Canada, 2003. p. 1191.
  • 15.Deshpande VS, Chung YC, Zhang Q, Shea SM, Li D. Reduction of transient signal oscillations in true-FISP using a linear flip angle series magnetization preparation. Magn Reson Med. 2003;49:151–157. doi: 10.1002/mrm.10337. [DOI] [PubMed] [Google Scholar]
  • 16.Schirf BE, Green JD, Sato K, Larson AC, Shaibani A, Salem R, Li D, Omary RA. Determination of optimal gadolinium concentration using SSFP for catheter-directed contrast-enhanced coronary MR angiography. Acad Radiol, in press. [DOI] [PubMed]
  • 17.Cohen MV, Cohn PF, Herman MV, Gorlin R. Diagnosis and prognosis of main left coronary artery obstruction. Circulation. 1972;45:I57–65. [PubMed] [Google Scholar]
  • 18.Key H, Jackson PC, Thomas EA, Jeans WD, Davies ER. The accuracy of digital subtraction angiography for the quantification of atherosclerosis. Br J Radiol. 1987;60:1083–1088. doi: 10.1259/0007-1285-60-719-1083. [DOI] [PubMed] [Google Scholar]
  • 19.Yuan C, Murakami JW, Hayes CE, Tsuruda JS, Hatsukami TS, Wildy KS, Ferguson MS, Strandness DEJ. Phased-array magnetic resonance imaging of the carotid artery bifurcation: preliminary results in healthy volunteers and a patient with atherosclerotic disease. J Magn Reson Imaging. 1995;5:561–565. doi: 10.1002/jmri.1880050515. [DOI] [PubMed] [Google Scholar]
  • 20.Toussaint JF, LaMuraglia GM, Southern JF, Fuster V, Kantor HL. Magnetic resonance images lipid, fibrous, calcified, hemorrhagic, and thrombotic components of human atherosclerosis in vivo. Circulation. 1996;94:932–938. doi: 10.1161/01.cir.94.5.932. [DOI] [PubMed] [Google Scholar]
  • 21.Botnar R, Stuber M, Kissinger K, Kim W, Spuentrup E, Manning W. Noninvasive coronary vessel wall and plaque imaging with magnetic resonance imaging. Circulation. 2000;102:2582–2587. doi: 10.1161/01.cir.102.21.2582. [DOI] [PubMed] [Google Scholar]
  • 22.Fayad Z, Fuster V, Fallon J, Jayasundera T, Worthley S, Jayasun T, Worthley S, Helft G, Aguinaldo J, Badimon J, Sharma S. Noninvasive in vivo human coronary artery lumen and wall imaging using black-blood magnetic resonance imaging. Circulation. 2000;102:506–510. doi: 10.1161/01.cir.102.5.506. [DOI] [PubMed] [Google Scholar]
  • 23.Isner JM, Kishel J, Kent KM, Ronan JA, Jr, Ross AM, Roberts WC. Accuracy of angiographic determination of left main coronary arterial narrowing. Angiographic--histologic correlative analysis in 28 patients. Circulation. 1981;63:1056–1064. doi: 10.1161/01.cir.63.5.1056. [DOI] [PubMed] [Google Scholar]
  • 24.Yang X, Bolster BJ, Kraitchman D, Atalar E. Intravascular MR-monitored balloon angioplasty: an in vivo feasibility study. J Vasc Interv Radiol. 1998;9:953–959. doi: 10.1016/s1051-0443(98)70429-4. [DOI] [PubMed] [Google Scholar]
  • 25.Omary RA, Frayne R, Unal O, Warner T, Korosec FR, Mistretta CA, Strother CM, Grist TM. MR-guided angioplasty of renal artery stenosis in a pig model: a feasibility study. J Vasc Interv Radiol. 2000;11:373–381. doi: 10.1016/s1051-0443(07)61433-x. [DOI] [PubMed] [Google Scholar]

RESOURCES