| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Pediatric Imaging |
1 From the Departments of Radiology (M.G., N.H., T.K., H.O., T.E., S.V., R.F.) and Pediatric Cardiology (T.B., G.B.), Charité, Campus Virchow-Klinikum, Medizinische Fakultät der Humboldt-Universität, Augustenburger Platz 1, 13353 Berlin, Germany, and the Departments of Congenital Heart Disease and Pediatric Cardiology (M.V.) and Cardiac, Thoracic, and Vascular Surgery (R.H.), German Heart Institute, Berlin. From the 1998 RSNA scientific assembly. Received December 17, 1998; revision requested February 8, 1999; revision received May 4; accepted August 20. Address reprint requests to M.G. (e-mail: matthias.gutberlet@charite.de).
 |
Abstract |
|---|
 TOP Abstract IntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
MATERIALS AND METHODS: Twenty consecutive patients (age range, 2–17 years) with an acoustic window that was insufficient for Doppler transthoracic echocardiography were included in the study. Flow and diameter measurements of the pulmonary arterial trunk and its primary branches were performed with phase-shift velocity mapping and cine MR imaging.
RESULTS: There were good correlations between pressure gradients in the pulmonary arteries estimated with MR imaging and those measured with Doppler echocardiography (r = 0.83, n = 15) and cardiac catheterization (r = 0.90, n = 13). Cine MR imaging revealed that the diameters of the right and left pulmonary arteries decreased with the expansion of the aorta during systole, which increased the peak velocity. This temporary stenosis was more severe in the right than in the left pulmonary artery and was accompanied by a significantly (P < .05) lower volume flow in the right artery.
CONCLUSION: The anatomic situation after arterial switch repair tended to produce temporary stenoses in the primary pulmonary arterial branches, with significant changes in hemodynamics. These changes may affect the long-term outcome and go undetected with other imaging modalities.
Index terms: Aorta, flow dynamics • Aorta, MR, 564.12142, 941.129411, 941.129412 • Aorta, surgery, 56.459, 941.459 • Pulmonary arteries, flow dynamics • Pulmonary arteries, MR, 564.121411, 564.121412, 564.12144 • Transposition of great vessels, 564.1612, 941.149
|
Introduction |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Postoperative obstruction may occur at different levels. In recent years, echocardiography has become the diagnostic method of choice for preoperative evaluation of patients with transposition of the great arteries. Subcostal imaging provides a flexible acoustic window that allows wide angulation and rotation of the transducer beam to optimize simultaneous demonstration of the great arteries (ascending aorta [AA] and main pulmonary artery [PA] with its primary branches) and their respective ventricular connections in the neonate (11). The growth of the child complicates postoperative assessment, however, because it becomes more difficult to find an adequate acoustic window owing to the typical retrosternal location of the pulmonary trunk and its primary branches after the Lecompte maneuver (Fig 1). In such cases, transesophageal echocardiography may by helpful (12–14). Systematic studies of the value of transesophageal echocardiography in such patients have not been performed, to our knowledge, and were not performed in the present study.
|
|
MATERIALS AND METHODS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
The examinations were performed at a mean of 6 years after surgery (Table). All patients had undergone preoperative cardiac catheterization, balloon atrial septostomy, and alprostadil infusion before the primary arterial switch procedure. The Lecompte maneuver (15,16), which causes displacement of the anterior aorta to a position posterior to the bifurcation of the pulmonary artery (Fig 1), had been performed in all 20 patients. In one patient, the pulmonary arterial connection was established by means of an interposed prosthetic tubular conduit.
|
All patients had an insufficient acoustic window at Doppler TTE, which was an indication for the use of MR imaging. Our institutional ethics committee approved the use of MR imaging for flow measurement and ventriculometry in patients with congenital heart disease. Informed consent was obtained from the parents of each patient prior to MR imaging. Before MR imaging, sedation for about 45 minutes was instituted in nine patients by means of oral administration of chloral hydrate (60 mg/kg).
Imaging
Doppler TTE was performed (T.B., M.V.) first. An ultrasonographic (US) unit (128 XP; Acuson, Mountain View, Calif) with a 3.5- or 5.0-MHz transducer was used. MR imaging was then performed (M.G., H.O.) by using a 1.5-T unit (Gyroscan ACS-NT; Philips Medical Systems, Best, the Netherlands). A body or surface coil was used, depending on the chest measurement and heart axis of the patient. For the anatomic evaluation, electrocardiographically gated turbo spin-echo MR images (706/11 [repetition time msec/effective echo time msec]; echo train length of five; section thickness, 3–5 mm; matrix, 256 x 256; trigger delay, 460 msec) were obtained during diastole in the transverse, sagittal, and coronal planes. In one patient (patient 10), a two-dimensional electrocardiographically triggered in-flow MR angiogram (11/5.7 [repetition time msec/echo time msec]; section thickness, 4 mm with 2-mm overlap; matrix, 128 x 256) was obtained for use in three-dimensional reconstruction of the typical nonphysiologic anatomic arrangement of the AA and PA after the arterial switch operation (Fig 1). Quantitative flow measurements were performed with a flow-sensitive gradient-echo sequence by using phase-shift velocity mapping (20/2.4, flip-angle, 30°; section thickness, 3–6 mm; field of view, 150–300 mm; matrix, 96 x 128). Retrospective gating was used to cover the whole cardiac cycle. The temporal resolution was 30–60 msec, depending on the number of signals acquired during the cardiac cycle, which in turn was dependent on the patient's heart rate.
An in-plane measurement was obtained parallel to the course of the vessel to assess the geometry of the main flow vector (Figs 2a, 3a). Next, a through-plane flow measurement was obtained perpendicular to the course of the vessel or to the course of the jet of a stenosis or regurgitation (Figs 2b, 2c, 3b).
|
|
|
|
|
|
Measurements were performed for the PA, AA, RPA, and LPA. In addition to peak velocity, the volume flow rate was measured in milliliters per second for all four vessels. For estimation of the pressure gradients at the site of a stenosis, the modified Bernoulli equation (17,18) was used. For volume flow rate measurements, the cross-sectional area of each vessel was manually outlined on images obtained during the whole cardiac cycle. The MR imager software was used to calculate mean volume flow rates from the mean velocities within the manually outlined cross-sectional areas (19). Furthermore, the maximum and minimum diameters of the PA, AA, RPA, and LPA during the cardiac cycle were measured on modulus images (Fig 3b) derived from the flow-sensitive cine gradient-echo MR images.
All in vivo measurements were performed once. The results of the flow measurements were compared with the results of cardiac catheterization and/or Doppler TTE measurements. Furthermore, the volume flow rates in the PA and AA were assessed to exclude the effects of important left-to-right or right-to-left shunting. To monitor the accuracy of the in vivo flow rate measurements, the ratio of pulmonary flow to systemic (aortic) flow was calculated. The closer this ratio was to 1, the more accurate the measurement was.
Phantom Studies
Phantom studies were performed (M.G., T.K.) with a pulsatile flow phantom (18,20) by using flow-sensitive gradient-echo MR imaging and phase-shift velocity mapping (same sequence parameters as for imaging in patients) to evaluate the accuracy of the volume flow rate measurements obtained with that MR sequence. The echocardiographic trigger was produced by means of an artificial pneumatic heart (Berlin Heart; Mediport Berlin, Germany), which was used to maintain the "circulation" (18,20).
Variable frequencies of 60–100 beats per minute and different volume flow rates of 100–1,500 mL/min were used. Measurements were performed three times with both pulsatile and constant flow. The mean of the three measurements was used for comparison. The volume flow rate in the circulatory model was monitored by means of an integrated (18) flowmeter with a 7.5-MHz probe (model T109R; Transonic Systems, Ithaca, NY) at a constant angle of 30°.
Statistical Analyses
The Wilcoxon rank sum test for unpaired data was used for comparisons. Results were considered to be significantly different for P values of less than .05. Furthermore, regression analyses were performed, and the data were presented by using box plots, regression curves, or both.
|
RESULTS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
|
|
|
|
|
|
Flow Measurements
In vivo peak velocity.—The mean peak velocity was slightly increased in the PA (2.1 m/sec ± 1.1), as compared with normal values (23). The mean peak velocity was almost within the normal range in the AA (1.2 m/sec ± 0.4). The mean peak velocity in the RPA was significantly higher than that in the LPA (1.7 m/sec ± 1.1 vs 1.0 m/sec ± 0.9; P < .05; Table).
Volume flow rate in phantoms.—The volume flow rates measured in the phantom at flow-sensitive gradient-echo MR imaging showed a good correlation with in vivo volume flow rates (r = 0.999, P < .001). The correlation of peak velocities showed a slightly weaker correlation (r = 0.95, P < .001). An in vivo velocity greater than the "encoded velocity" results in aliasing, as occurs at Doppler US. An encoded velocity of up to 100% greater than the in vivo flow velocity did not result in a measurable loss of signal intensity. Therefore, the standard adjustment for the encoded velocity was 2 m/sec. In the case of aliasing or slower flow, the encoded velocity was changed as necessary.
In vivo volume flow rate.—The mean volume flow rate was 52.5 mL/sec ± 19.3 in the PA and 52.1 mL/sec ± 19.4 in the AA, and the sum of the mean volume flow rates in the RPA and LPA was 48.2 mL/sec ± 18.6 (Table). These results demonstrate good internal consistency for the volume flow rate measurements. Almost no difference was observed in the cardiac output of about 3,000 mL/min in the PA and AA, and the mean ratio of pulmonary flow to systemic flow in 19 patients was 1.0 ± 0.1. No patient had a persistent atrial or ventricular septal defect or any other form of shunt. In one patient with high-grade pulmonary valvular incompetence and a stent in the PA, there was not good agreement between the results of the volume flow rates in the AA and PA, so the results in this patient were excluded from this comparison. The mean volume flow rate in the LPA was significantly higher than that in the RPA (27.1 mL/sec ± 16.3 vs 20.8 mL/sec ± 9.9; P < .05; Fig 6).
Vessel Diameter
In all patients, the maximum vessel diameters in the PA and AA were achieved during systole, and the minimum diameters were achieved during diastole, as expected. The mean maximum vessel diameter of the PA was 14.1 mm ± 4.7 (area, 156.1 mm2), and the mean minimum diameter was 13.7 mm ± 4.0 (area, 147.4 mm2). During systole, therefore, the mean increase in diameter was 3%, and the mean increase in area was 6%. The mean maximum diameter of the AA was 19.0 mm ± 5.6 (area, 283.5 mm2), and the mean minimum diameter was 16.2 mm ± 5.1 (area, 206 mm2). During systole, therefore, the mean increase in diameter was 17.3%, and the mean increase in area was 37.6%.
The AA bulges into the pulmonary arteries during systole (Fig 3a). Therefore, the maximum diameter in the RPA at the region closest to the AA was achieved in all patients during diastole, and the minimum diameter was achieved during systole. In the LPA, however, the maximum diameter was achieved during systole in some patients (Fig 3b) and during diastole in others. The mean maximum diameter of the RPA was 9.0 mm ± 2.8 (area, 63.6 mm2), and the minimum diameter was 6.5 mm ± 2.3 (area, 33.2 mm2). During systole, therefore, the mean reduction in diameter was 31.4%, and the mean reduction in area was 54%.
In the LPA, the mean maximum diameter was 8.8 mm ± 3.8 (area, 61 mm2), and the mean minimum diameter was 8.0 mm ± 3.9 (area, 50.3 mm2). During systole, therefore, the mean reduction in diameter was 9.1%, and the mean reduction in area was 17.6%. The difference between the systolic diameter in the RPA and that in the LPA was not significant (P = .18) (Fig 7).
Stenosis and Valvular Incompetence
Morphology.—In the AA and the descending aorta, no vessel or valvular stenosis could be detected on turbo spin-echo or flow-sensitive cine gradient-echo MR images in any patient. This was confirmed in all patients at Doppler TTE, cardiac catheterization, or both, for the aortic valve and the AA, which was visualized at TTE in all patients. The Doppler TTE and cardiac catheterization criterion for no vessel or valvular stenosis in the AA or descending aorta was a peak flow rate of less than 1.6 m/sec or a peak pressure gradient of less than 10 mm Hg. None of the patients examined had a "neocoarctation" such as has been described by others (23) after use of the Lecompte maneuver (in which the anterior aspect of the AA manifests posterior displacement caused by the stretched, anteriorly placed PA and bifurcation).
Pressure gradients.—Pressure gradients were calculated from MR images only for the PA, LPA, and RPA and were compared with gradients estimated with Doppler TTE or measured at cardiac catheterization, if available. The pressure gradients calculated with results from phase-shift velocity mapping showed a good correlation with the invasive (catheterization) data, which were available for 14 patients. In one patient with a severe stenosis of the pulmonary valve and an additional stenosis of the LPA, quantification of the pressure gradient with phase-shift velocity mapping was impossible, owing to the presence of turbulence; therefore, results in this patient were excluded from the comparison (Fig 5c). Cardiac catheterization revealed a pressure gradient of 115 mm Hg at the pulmonary valve, but Doppler TTE resulted in underestimation of the gradient (Table). Despite the inability to quantify, a high-grade stenosis was suspected on the basis of findings from turbo spin-echo MR imaging and the decreased signal intensity in the area of the pulmonary valve and the LPA on the modulus image derived from flow-sensitive gradient-echo MR images.
Pressure gradients estimated at Doppler TTE in 11 patients also showed a good correlation (r = 0.75) with those measured at cardiac catheterization (Fig 5a), but the tendency toward overestimation of pressure gradients with TTE was greater than that with MR imaging. In three patients in whom complete catheterization and MR imaging data were obtained, we could not use Doppler TTE to quantify a gradient in the PA owing to an inadequate acoustic window. Therefore, complete data for comparison among the three modalities were available in 10 patients (Fig 4).
|
DISCUSSION |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Flow Measurements
Peak velocity.—Simple depiction of anatomic structures during a single phase of the cardiac cycle often is not sufficient for evaluation of the complex dynamics involved. In addition to demonstration of a stenosis, therefore, the hemodynamics of the stenosis must be evaluated by means of flow measurements. As in other studies (18,27), a good correlation between MR imaging data and cardiac catheterization data was obtained in patients with a mild or moderate stenosis (Fig 5c).
Use of the modified Bernoulli equation provides an estimate of the instantaneous gradient, which differs from that measured at cardiac catheterization. This difference increases with the severity of stenosis (28–30). In the present study, however, estimation with the Bernoulli equation was shown to be adequate for clinical practice. The criteria for further invasive diagnostic procedures and/or repeated surgery at our institution are a pressure gradient of more than 50 mm Hg, even in the absence of clinically detectable symptoms, or a gradient of more than 40 mm Hg in the presence of symptoms and/or changes in ventricular function or morphology. No further invasive diagnostic procedure is necessary in patients with a gradient of less than 40 mm Hg. In all patients, we were able to distinguish among categories by using MR imaging.
Volume flow rate in phantoms.—Our phantom study results demonstrated the accuracy of the flow-sensitive gradient-echo MR sequence with phase-shift velocity mapping for flow quantification. The influence of breathing on the accuracy of flow measurements has previously been described (31) and was not evaluated in the present study. The influence of pulsatile or continuous flow was investigated, and there was no significant difference between the two.
In vivo volume flow rate.—Mild or moderate dynamic obstruction of the LPA and RPA is caused by bulging during systole of the neoaorta into the pulmonary arterial bifurcation. This bulging is due to the special nonphysiologic anatomic situation of the pulmonary arterial branches surrounding the AA after performance of the Lecompte maneuver (Fig 1) (15,16). This situation causes changes in the hemodynamics in the RPA and LPA, whereas the larger vessels (the PA and AA) are less affected and show only a slight increase in mean peak velocity. In the more affected RPA, we found a significantly elevated peak velocity caused by dynamic or temporary obstruction, whereas the mean volume flow rate was significantly less in the RPA than in the LPA (Fig 6).
Volume flow measurements were obtained not only in the LPA and RPA but also in the PA and AA to serve as internal standards for the in vivo flow measurements. No significant differences between these arteries were found, which indicated the internal consistency of the in vivo measurements. There was only one patient who had high-grade incompetence of the pulmonary valve: The ratio of pulmonary flow to systemic flow in this patient was substantially greater than 1.0 (Table); this high ratio was due to turbulence, which prevented adequate measurements of flow rates. In all other patients, the results showed a good agreement, with a pulmonary flow–to–systemic flow ratio of 1.0 ± 1.0.
Diameter Measurements
Changes in vessel diameter during the cardiac cycle were examined. By using cine gradient-echo MR imaging, we found a reduction in vessel diameter and vessel area during systole in the RPA at the site of greatest contact with the AA (Fig 3a), whereas in the LPA some patients had a decrease in diameter and others had an increase, as is normal for the vessel (Fig 3b). In all patients, expansion of the AA during systole was the reason for this transitory stenosis. In most of the patients, the great arteries were not positioned one behind the other. The neo-PA was usually positioned somewhat to the side of the neoaorta (32). In most patients who have undergone the Lecompte maneuver, therefore, the course of the LPA from the PA is at a steep angle and is less close to the AA (33), as compared with the course of the RPA, which, in most cases, is at a flat angle close to the AA. This caused the RPA to be more affected by the expanding AA during systole, resulting in a transitory stenosis of the RPA. As a result, an increase in the mean peak velocity and a decrease in the volume flow rate in the RPA were observed. The complex dynamics of the AA, LPA, and RPA in this particular anatomic situation could not be assessed at cardiac catheterization.
Cardiac catheterization usually cannot simultaneously demonstrate the AA and PA. Simultaneous depiction of the right and left ventricular systems is one of the advantages of cine MR imaging. Cine gradient-echo MR imaging in the present study demonstrated that the proximity of the right ventricular outflow tract, PA, RPA, and LPA to the high-pressure left ventricular system may be one reason for the development of stenosis in patients such as those in our study. Depending on the particular anatomic situation present before surgery, the PA, RPA, or LPA could be affected. Paul and Wernovsky (34) reported that the LPA is more often stenosed, but our results, like those of other investigators (15,25), do not support this claim. The only occlusion observed in the present study was in the LPA, 0.5 cm distal to the PA bifurcation (Fig 2).
Clinical Importance
The mechanism for obstruction of pulmonary blood flow is multifactorial. Mild to moderate obstruction of the LPA and RPA immediately distal to the bifurcation may be caused by systolic bulging of the AA into the bifurcation, which, in addition to kinking of the RPA and/or LPA due to cranial displacement of the branches, may lead to a permanent obstruction. Other mechanisms of permanent supravalvular obstruction of the pulmonary outflow tract include inadequate growth of the pulmonary arteries during somatic growth (35,36), circumferential narrowing at the suture line, or a combination of anatomic mechanisms, resulting in multiple obstructions at various levels.
The different lesions respond to various treatments, including balloon angioplasty, which can successfully help relieve discrete branch stenosis or circumferential narrowing in the PA. It is clinically important to determine prospectively which patients may be candidates for an attempt at nonsurgical treatment of an obstruction. MR imaging in concert with phase-shift velocity mapping may be the noninvasive method of choice to help select patients for balloon angioplasty and for follow-up after the intervention.
Under normal circumstances, slightly more than 50% of the cardiac output flows through the RPA (37–40). Therefore, a marked change in the distribution of blood flow may be of clinical importance and could influence the long-term outcome in patients. Thus, progressive reduction in blood flow in the RPA, as well as in the LPA, will necessitate a second surgical or interventional procedure.
Limitations of the Present Study
The small sample size of the study group and the wide range of patient ages and sizes may have been among the major limitations of this study. Another limitation was the fact that cardiac catheterization data were available for only 14 of 20 patients. Although the arterial switch operation has been in use for more than 2 decades, there still is not a large population being followed up and who are, therefore, possible candidates for MR imaging. Despite the small sample size and heterogeneity of the group in our study, however, MR imaging demonstrated temporary obstructions in pulmonary arterial branches that caused important changes in the pulmonary blood distribution after the arterial switch operation in almost all patients; such changes can go undetected with other imaging modalities, including cardiac catheterization.
A temporary or systolic obstruction may affect the long-term outcome in these patients and should be studied further, because such an obstruction may progress and finally lead to right ventricular overload, which must be treated with repeated surgery. Although such stenoses are not permanent, they are, nonetheless, hemodynamically significant because they occur during systole.
The accuracy of flow measurements can be affected by breathing. This effect was not evaluated in our study, but it has been examined by other investigators. In one study (31), MR imaging resulted in overestimation of actual flow during systole and underestimation during diastole. The errors in estimation decrease with measurements obtained during increasing numbers of phases of the cardiac cycle. Peak flow rate was more affected than was volume flow rate. Breathing should not have substantially altered our findings, because the origin of the pulmonary arteries is not much affected by thoracic movement during respiration. Furthermore, the RPA and LPA are affected by this respiratory motion in the same way. Therefore, the significant difference in blood distribution of the mean volume flow rates between the RPA and the LPA should not have been altered. In addition, we acquired measurements during at least 12 phases per cardiac cycle; in the aforementioned experimental study (31), measurements were obtained during one to six phases.
In general, there is a tendency toward overestimation of pressure gradients with phase-shift velocity mapping (18,19), similar to the results with Doppler TTE, especially in the setting of severe stenosis. This may not be clinically relevant for decision making with regard to intervention. Finally, noninvasive MR imaging may not be tolerated by all patients because of claustrophobia, noise sensitivity, or other contraindications (eg, presence of a pacemaker or metal implant). Most vascular stents are MR compatible. One patient in our group had a pulmonary arterial stent and was examined without any complications.
|
Footnotes |
|---|
Author contributions: Guarantor of integrity of entire study, M.G.; study concepts and design, M.G., T.B., N.H.; definition of intellectual content, M.G., T.B., N.H., M.V., R.H., G.B., R.F.; literature research, M.G., T.K., T.E., H.O., S.V.; clinical studies, M.G., T.B., M.V.; experimental studies, M.G., T.K.; data acquisition, M.G., T.B., H.O., T.K.; data analysis, M.G., T.K., T.E.; statistical analysis, M.G., T.K., T.E.; manuscript preparation, M.G.; manuscript editing, M.G., T.E.; manuscript review, T.B., M.V., N.H., S.V., R.H., G.B., R.F.
|
References |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  H. B. Grotenhuis, A. de Roos, J. Ottenkamp, P. H. Schoof, H. W. Vliegen, and L. J. M. Kroft MR Imaging of Right Ventricular Function after the Ross Procedure for Aortic Valve Replacement: Initial Experience Radiology, December 4, 2007; (2007) 2462070198. [Abstract] [Full Text]
|
|
|
|
 |
|
 E. Rodriguez, R. Soler, R. Fernandez, and I. Raposo Postoperative Imaging in Cyanotic Congenital Heart Diseases: Part 1, Normal Findings Am. J. Roentgenol., December 1, 2007; 189(6): 1353 - 1360. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Soler, E. Rodriguez, M. Alvarez, and I. Raposo Postoperative Imaging in Cyanotic Congenital Heart Diseases: Part 2, Complications Am. J. Roentgenol., December 1, 2007; 189(6): 1361 - 1369. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H B Grotenhuis, L J M Kroft, S G C van Elderen, J J M Westenberg, J Doornbos, M G Hazekamp, H W Vliegen, J Ottenkamp, and A de Roos Right ventricular hypertrophy and diastolic dysfunction in arterial switch patients without pulmonary artery stenosis Heart, December 1, 2007; 93(12): 1604 - 1608. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I.-S. Chiu, J.-K. Wang, and M.-H. Wu Spiral arterial switch operation in transposition of the great arteries J. Thorac. Cardiovasc. Surg., November 1, 2002; 124(5): 1050 - 1052. [Full Text] [PDF]
|
|
|
|
 |
|
 T. Kuehne, M. Saeed, G. Reddy, H. Akbari, K. Gleason, D. Turner, D. Teitel, P. Moore, and C. B. Higgins Sequential Magnetic Resonance Monitoring of Pulmonary Flow With Endovascular Stents Placed Across the Pulmonary Valve in Growing Swine Circulation, November 6, 2001; 104(19): 2363 - 2368. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Kuehne, M. Saeed, P. Moore, K. Gleason, G. Reddy, D. Teitel, and C. B. Higgins Influence of Blood-Pool Contrast Media on MR Imaging and Flow Measurements in the Presence of Pulmonary Arterial Stents in Swine Radiology, May 1, 2002; 223(2): 439 - 445. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||
