| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Vascular and Interventional Radiology |
1 From the Departments of Radiology (C.L.F., L.E.M.D, A.V.T., A.B.D.v.R., H.C.M.v.d.B), Nephrology (P.D.D.), and Vascular Surgery (P.W.M.C., J.B.), Catharina Hospital, Michelangelolaan 2, 5623 EJ, Eindhoven, the Netherlands; and Department of Epidemiology and Biostatistics and Department of Radiology, Erasmus MC-University Medical Center Rotterdam, the Netherlands (Y.S.L.). Received November 19, 2003; revision requested February 6, 2004; revision received March 11; accepted April 8. Address correspondence to L.E.M.D. (e-mail: lemduijm@hotmail.com).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Ethical review board approval and written informed consent were obtained. MR angiography and DSA were performed in 51 dysfunctional hemodialysis fistulas and grafts in 48 consecutive patients. Vascular tree of accesses was divided into between three and eight segments depending on access type (AVF or AVG) and length of venous outflow. Images obtained with MR and DSA were interpreted by two MR radiologists and two interventional radiologists, respectively, who were blinded to information from each other and other studies. DSA was reference standard for stenosis detection. Sensitivity, specificity, and predictive values with 95% confidence intervals (CIs) of contrast-enhanced MR in detection of vascular segments containing hemodynamically significant (
50%) stenosis were calculated. Linear-weighted 
statistic was calculated for contrast-enhanced MR and DSA to determine interobserver agreement regarding stenosis detection.
RESULTS: A total of 282 vascular segments were evaluated. Contrast-enhanced MR depicted three false-positive stenoses and all but two of 70 significant stenoses depicted with DSA. Sensitivity, specificity, and positive and negative predictive values of MR in detection of vessel segments with significant stenoses were 97% (95% CI: 90%, 99%), 99% (95% CI: 96%, 100%), 96% (95% CI: 88%, 99%), and 99% (95% CI: 97%, 100%), respectively. MR demonstrated significant stenosis in four of five nondiagnostic DSA segments, whereas DSA showed no significant stenosis in four nondiagnostic MR segments. Linear-weighted statistic for interobserver agreement regarding stenosis detection was 0.92 (95% CI: 0.89, 0.95) for MR and 0.95 (95% CI: 0.92, 0.97) for DSA.
CONCLUSION: MR angiography depicts stenoses in dysfunctional hemodialysis accesses but has limited clinical value as result of current inability to perform MR-guided access interventions after stenosis detection. MR of dysfunctional access should be considered only if nondiagnostic vascular segment is present at DSA.
© RSNA, 2005
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Several surveillance techniques may be used in the timely detection of access stenosis development. Venous pressure measurements can be used to trace more distally located stenoses in the venous outflow, but the development of arterial inflow resistance is not reflected in these measurements (7). Access flow measurements are helpful in selecting patients at risk for graft thrombosis, and screening with color Doppler ultrasonography (US) at regular intervals in an asymptomatic population is another method used to detect most hemodialysis access stenoses (8,9). Drawbacks of color Doppler US are overestimation of stenoses at the arterial anastomoses, inaccuracy in the detection of nonthrombosed obstruction of the upper extremity venous outflow, and absence of an angiographic map necessary for surgery or percutaneous therapy (10). In addition, the quality of color Doppler US images depends on the skill of the operator (11).
Digital subtraction angiography (DSA) is the current method of choice for anatomic hemodialysis access evaluation. DSA is no more invasive than needle puncture for dialysis, but it exposes the patient to ionizing radiation and iodinated contrast material. The latter may cause allergic reactions (12). In contrast, magnetic resonance (MR) angiography is noninvasive and lacks any ionizing radiation. On the other hand, MR-guided access intervention is not yet clinically applicable, although the feasibility of MR-guided balloon angioplasty has been demonstrated with experimental studies (13,14). Also, to the present day, MR angiography has little diagnostic value in the evaluation of dysfunctional hemodialysis access grafts and fistulas, as single-injection single-volume MR angiography protocols were used in recent studies. In these studies, the limited field of view is usually confined to the perianastomotic region and does not allow for the detection of central venous stenoses (15,16).
In a recent feasibility study of 10 patients, we used a new three-dimensional contrast material–enhanced MR angiography technique to depict the complete vascular tree of hemodialysis accesses and access stenoses (17). Thus, the purpose of the current study was to prospectively assess three-dimensional contrast-enhanced MR angiography for stenosis depiction in patients with malfunctioning hemodialysis fistulas and grafts, as compared with DSA.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
MR Angiography
For all acquisitions, we used a commercially available 1.5-T MR imager (Gyroscan Intera with software release 8.1 and 9.1; Philips Medical Systems, Best, the Netherlands) with master gradients (maximum amplitude, 30 mT/m). Patients were placed in a semioblique supine position on the side of the shunt arm with the upper extremities lying next to the body. An 18-gauge intravenous cannula (Venflon; Ohmeda, Helsingborg, Sweden) was inserted into an antecubital vein of the upper extremity contralateral to the access. Time-of-flight surveys of the inflow-outflow tract (subclavian, axillary, and upper arm arteries and veins) and shunt region were performed after marking the hemodialysis access with a lipid marker to accurately plan high-spatial-resolution contrast-enhanced MR angiography.
A test bolus of 2 mL of gadoteridol (Prohance; Bracco, Milan, Italy) and a subsequent flush of 10 mL saline (injection rate, 0.6 mL/sec) were injected through the intravenous cannula by using an MR-compatible injector (Spectris MR injector; Medrad, Indianola, Pa). Test-bolus injection and a timed examination were started simultaneously. A thick coronal slab was imaged with a real-time mode (acquisition time for each image, 1.5 seconds). The time interval between the start of injection and the maximum of the time-intensity curve, in a region of interest placed in the hemodialysis fistula or loop graft, was used to determine the contrast bolus arrival time at the level of the access. The latter was used as the time delay between the start of contrast material injection and the start of high-spatial-resolution acquisition. The region of interest was placed by one of the authors (H.C.M.v.d.B.) and did not have a fixed size, as it was adjusted to the diameter of the access in which the contrast bolus arrival time was determined.
A total of 39 ml of gadoteridol was injected in two subsequent volumes, with a flow rate of 0.6 mL/sec: 20 mL were injected into the shunt region, and 19 mL were injected into the inflow-outflow tract. The total amount (41 mL) of contrast material did not exceed the maximally allowed dose of gadoteridol (0.6 mL per kilogram of body weight), as the body weight of patients in our study was 70 kg or more.
Two overlapping three-dimensional volumes of the shunt region and inflow-outflow tract were acquired with the contrast-enhanced MR angiography technique. An optimized elliptic centric profile order was used for k-space filling (contrast-enhanced timed robust angiography) of both volumes.
Parameters used in the three-dimensional T1-weighted fast field-echo sequence to image the shunt region were as follows: repetition time msec/echo time msec, 4.1/1.34; flip angle, 20°; acquisition matrix, 432 x 346; number of sections, 120; overcontiguous section thickness, 0.55 mm; acquisition voxel size, 1.00 x 1.00 x 1.10 mm; interpolated voxel size, 0.84 x 0.84 x 0.55 mm; imaging time, 32 seconds.
Parameters used to obtain the three-dimensional volume for the inflow-outflow tract were as follows: 3.6/1.23; flip angle, 20°; acquisition matrix, 368 x 294; number of sections, 95; overcontiguous section thickness, 0.9 mm; acquisition voxel size, 1.17 x 1.17 x 1.80 mm; interpolated voxel size, 0.84 x 0.84 x 0.90 mm; imaging time, 32 seconds. Maximum intensity projections were calculated from the three-dimensional data sets at a remote dedicated workstation (Easy Vision; Philips Medical Systems). The mean examination time, including preparation, timing bolus test, and postprocessing, was 35 minutes (range, 28–55 minutes). All MR angiography examinations were performed without any complications.
DSA Technique
All angiograms were obtained with a DSA system (Multistar Time Operation Performance; Siemens Medical Engineering, Forchheim, Germany) by using the nonionic contrast material iomeprol (Iomeron 350; Bracco). A 4-F dilator was inserted after retrograde puncture of the venous part of the AVF or AVG. Images of the complete venous outflow were obtained with repeated manual injection of 5–10 mL of contrast material. Images of the access and its inflow tract were obtained either with manual compression or with flow interruption through a cuff of the venous outflow (18). In cases where the arterial inflow could not be depicted adequately, an attempt was made to advance a guidewire from the venous puncture site into the feeding artery. The dilator was exchanged for a 4-F straight catheter, whose tip was positioned distally in the arterial inflow. Magnification images and angled views of suspected stenoses were obtained. Imaging parameters included a matrix of 1024 x 1024 and a field of view of 14–40 cm. DSA examinations were performed by experienced interventional radiologists (L.E.M.D. and A.V.T.), who had access to findings of previous DSA studies. Both interventional radiologists had more than 10 years of experience with DSA of AVFs and AVGs, and they were not aware of the results of contrast-enhanced MR angiography when they performed the DSA examination and interpreted the results. The mean examination time, which included patient preparation and access puncture but did not include time needed for any interventional procedure, was 23 minutes (range, 15–35 minutes). In connection with DSA, percutaneous transluminal angioplasty was performed of those stenoses and showed a luminal reduction of 50% or more at DSA.
Image Analysis
For image analysis, the vascular tree was divided into the following segments (Fig 1): distal arterial inflow (segment I); arterial anastomosis, including 1 cm of vessel length on both sides of the anastomosis (segment II, in case of an AVG); bridge graft (segment III); venous anastomosis, including 1 cm of vessel length on both sides of the anastomosis (segment IV, in case of an AVG); and arteriovenous anastomosis, including 1 cm of vessel length on both sides of the anastomosis (segment V, in case of an AVF). Venous outflow was subdivided into four segments from the wrist up to the superior caval vein (segment VI1, forearm; segment VI2, distal half upper arm; segment VI3, proximal half of upper arm, including cephalic vein; segment VI4, central venous outflow comprising subclavian, bracheocephalic, and superior caval veins). Thus, the vascular tree of an AVG and AVF was divided into at least five and three segments, respectively.
|
Reasons for nondiagnostic images were mentioned by a reading radiologist, in case one or several segments were considered nondiagnostic. For analysis purposes, only the most severe stenosis per vessel segment was taken into account. Source images were available to all reviewers; the MR angiogram reviewers could use maximum intensity projections and multiplanar reformations. Stenoses with a luminal narrowing that exceeded 50% were considered hemodynamically significant. Reading of images obtained with contrast-enhanced MR angiography and DSA, including postprocessing and measuring stenoses with electronic calipers, was performed by two MR radiologists (H.C.M.v.d.B. and A.B.D.v.R.) and two interventional radiologists (L.E.M.D. and A.V.T.), respectively. Each MR radiologist had read more than 200 MR angiograms of various vascular territories annually for the past 5 years but had very limited experience with the recently introduced technique of MR angiography of dialysis accesses. MR radiologists assessed MR angiograms independently from each other and were blinded to information from the DSA study. In case of disagreement, consensus was reached after mutual consultation. This assessment technique was also used by the two interventional radiologists, who were unaware of the results of MR angiography.
After comparing images obtained with contrast-enhanced MR angiography with those obtained with DSA, percutaneous transluminal angioplasty was scheduled for cases where contrast-enhanced MR angiography depicted a stenosis of 50% or more in nondiagnostic DSA segments. For the latter lesions, US-dilution flow measurements were obtained shortly before percutaneous transluminal angioplasty and at the time of the first dialysis session after percutaneous transluminal angioplasty to determine flow improvement.
Statistical Analysis
DSA was considered to be the standard of reference. Sensitivity, specificity, predictive values, and accuracy with 95% confidence intervals [CIs] of MR angiography in the detection of vascular segments containing a significant (50%) stenosis were calculated (19). In addition, we calculated previously mentioned test characteristics when dividing the vascular tree into the following anatomically distinct regions: the arterial region, shunt region, and venous region. We considered stenoses in the different regions to be independent of one another. To divide the vascular tree into the three regions, we considered the arterial region to be region I. The shunt region was similar to segment V in an AVF. For AVGs, we randomly selected segment II, III, or IV for analysis. Similarly, for the venous region, we randomly selected a segment from segment VI1–4. To determine interobserver agreement with regard to detection of stenoses between the two radiologists, the linear-weighted statistic was calculated on the segment level for both contrast-enhanced MR angiography and DSA (20). The analyses were performed with SAS Systems for Windows, release 6.12 (SAS Institute, Cary, NC).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
In total, 282 vascular segments could be evaluated with both contrast-enhanced MR angiography and DSA (AVF: segment I, n = 32; segment V, n = 29; segment VI1, n = 9; segment VI2, n = 28; segment VI3, n = 32; segment VI4, n = 33) (AVG: segment I, n = 18; segment II, n = 18; segment III, n = 18; segment IV, n = 17; segment VI1, n = 0; segment VI2, n = 14; segment VI3, n = 16; segment VI4, n = 18). After consensus review, 70 hemodynamically significant stenoses were identified at DSA (AVF: segment V, n = 19; segment VI1, n = 1; segment VI2, n = 9; segment VI3, n = 10; segment VI4, n = 4) (AVG: segment II, n = 1; segment III, n = 9; segment IV, n = 14; segment VI2, n = 3) (Figs 2, 3). Contrast-enhanced MR angiography depicted 68 of these stenoses and three false-positive stenoses (Table 1). Both stenoses that were missed at MR angiography were located in the outflow tract of an AVF, whereas the false-positive stenoses were located in the outflow tract of an AVF and the arterial anastomosis and loop proper of two different AVGs. Overall, the sensitivity and specificity of MR angiography in the detection of vessel segments with a stenosis of 50% or more was 97% (95% CI: 90%, 99%) and 99% (95% CI: 96%, 100%), respectively. We found a positive predictive value of 96% (95% CI: 88%, 99%); a negative predictive value of 99% (95% CI: 97%, 100%); and an accuracy of 98% (95% CI: 96%, 99%). Contrast-enhanced MR angiography demonstrated a significant stenosis in four vascular segments that were considered nondiagnostic at DSA (Fig 4), whereas DSA showed no significant stenoses in nondiagnostic segments that were assessed with contrast-enhanced MR angiography (Fig 5). Access flow improved from 317 mL/min ± 20 to 689 mL/min ± 31 after percutaneous transluminal angioplasty of the four stenoses detected with MR angiography.
|
|
|
|
|
Interobserver disagreement about the degree of stenosis existed in 20 segments assessed with contrast-enhanced MR angiography and 13 segments assessed with DSA (Table 2). The linear-weighted value regarding the detection of stenosis was 0.92 (95% CI: 0.89, 0.95) for contrast-enhanced MR angiography and 0.95 (95% CI: 0.92, 0.97) for DSA, indicating almost perfect agreement for both diagnostic modalities.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
We computed test characteristics by using segment level data and not patient level data. An argument for analysis at patient level would be that the decision to treat a patient is usually made at the patient level. In this study, however, hemodialysis shunts that were already clinically dysfunctional were assessed. Thus, we did not want to know whether or not to treat patients; rather, we wanted to know where the lesion that was causing dysfunction was located and if MR angiography could depict this lesion reliably. Analyses that are based on segment level, however, may overestimate the accuracy because stenoses in adjacent segments may be dependent on one another. Furthermore, it can be argued that considering segments artificially increases statistical power. For these reasons, we also divided the vascular tree into three regions, in which stenoses can be assumed to be independent of one another. To determine the grade of stenosis in a certain region, we randomly selected one segment of this region if it consisted of multiple segments. Selecting one segment per region ensured the comparison of the same lesion at DSA and MR angiography. Test characteristics computed when dividing the vascular tree into regions were not very different from those computed at the segment level. Although this indicates that correlation of stenoses in adjacent segments was negligible, we acknowledge that our approach may have affected the results because the existence of dependency cannot be ruled out entirely.
In a series of 15 dysfunctional accesses, Planken et al (16) mentioned that the sensitivity and specificity of contrast-enhanced MR angiography were 100% and 10%, respectively, for detection of significant stenoses. Their large number of false-positive lesions detected with contrast-enhanced MR angiography was caused by a limited spatial resolution, which resulted in stenosis overestimation. Moreover, because of the use of a rectangular surface coil, no information could be obtained about the venous outflow of the upper arm. With use of a different MR imaging protocol, we were able to achieve a 58% reduction in voxel size, which enabled us to obtain a better spatial resolution. Smits et al (15) obtained high-quality MR images of accesses by injecting gadopentetate dimeglumine directly into the access. This approach may be preferable to intravenous administration of contrast material, as puncture of peripheral veins should be avoided whenever possible under the doctrine of venous preservation (21). Unfortunately, a major drawback of the approach of Smits et al (15) was the inability to depict complete venous outflow. Apart from our series, we are not aware of other reports that depict the complete vascular tree of hemodialysis access fistulas and grafts. Imaging of the veins central to the cubital fossa is of major importance; we found that 37% of all stenoses were located in this part of the outflow tract, which is in line with other results (6).
Nondiagnostic vascular segments may occasionally be present at DSA. Planken et al (16) found that one-quarter of vessel segments could not be evaluated properly at DSA, while Smits et al (15) reported that 8% of the segments were not adequately depicted. In our series, less than 2% of the segments were considered nondiagnostic at DSA. This small number of nondiagnostic segments may be explained by the fact that we advanced a catheter from the venous puncture site into the feeding artery when arterial inflow could not be depicted adequately, whereas Smits et al (15) and Planken et al (16) just positioned a catheter in the venous part of the access. Although two significant stenoses that were detected with DSA were missed with contrast-enhanced MR angiography, contrast-enhanced MR angiography did depict four hemodynamically significant stenoses in segments that were considered nondiagnostic at DSA. In these cases, anastomotic regions could not be assessed with DSA because of overprojection of a large pseudoaneurysm, inability to define the anastomosis, severe vascular kinking, or overprojection of vein branches. Contrast-enhanced MR angiography offers the ability to acquire three-dimensional data sets, which may potentially solve the problem of vessel overlap. Also, the images can be reformatted to show the access from arbitrary projection angles. The percutaneous transluminal angioplasty interventions of the four stenotic anastomoses, which are based on images obtained with contrast-enhanced MR angiography, turned out to be successful in terms of recovered flow in the first dialysis session after the intervention. Thus, it is likely that these lesions represented hemodynamically significant stenoses that could have resulted in acute access thrombosis if percutaneous transluminal angioplasty had not been performed. This finding is in line with the findings of other studies; for example, van der Linden et al (4) mention that not every significant stenosis may be identified with DSA.
During MR angiography, patients were placed in the semioblique supine position on the side of the shunt arm, with the upper extremities lying next to the body. This position caused outflow tract compression in two vessel segments. This artifact could be prevented by performing the examination with the patient in the prone or lateral position, with the shunt arm stretched overhead (22). This position, however, is much more uncomfortable and may lead to substantial motion artifacts (23). With the introduction of surface coils, positioning of the shunt arm in slight abduction can be achieved, and this may reduce the chance of outflow tract compression.
Proper positioning of the three-dimensional imaging volume with contrast-enhanced MR angiography is of utmost importance. In one case, the three-dimensional volume did not cover the complete vascular tree of the access. Because of contrast material dose limitations, contrast-enhanced MR angiography cannot be repeated on the same occasion. The implementation of targeted unenhanced MR sequences, performed directly after the contrast-enhanced series, may compensate for those situations in which the vascular tree is not depicted completely (24).
The time span between the MR examination and a possible intervention, either endovascular or surgical, should be as short as possible to prevent an increased risk for access thrombosis. In more than one-third of the cases, there was no interventional delay whatsoever, as the MR examination could be scheduled immediately before an already fixed DSA appointment. Although we observed no access thrombosis in the remaining patients, we acknowledge that MR imaging should not cause a delay in the treatment of patients in whom a high-grade stenosis is suspected.
Some authors advocate the use of diagnostic US in the assessment of clinically suspected hemodialysis access stenosis (10). However, the draining or central venous vasculature can be difficult to evaluate accurately for focal stenosis in the absence of thrombosis or other obvious signs of obstruction. A subclavian vascular access stenosis may be present, especially in patients with a history of subclavian vein cannulation (25). Also, we did not include diagnostic US, as the aim of this study was to compare angiographic-like modalities.
The combination of contrast-enhanced MR angiography and MR velocity mapping may be especially useful for follow-up over time, as both the anatomy and the function of dialysis fistulas and grafts can be assessed (26,27). When a low value of access flow is measured and the underlying stenosis is depicted by MR angiography, an interventional procedure can then be scheduled.
A major limitation of our study is the absence of MR-guided access interventions. A clear advantage of DSA over contrast-enhanced MR angiography is the possibility of performing the corrective intervention immediately, should a hemodynamically significant stenosis be detected. Techniques for MR-guided endovascular interventions still have to be developed further to provide a clinically attractive alternative (28). The presence of vascular metallic stents, especially those made of stainless steel, may be another limitation of contrast-enhanced MR angiography, as it is well known that these objects can hamper postinterventional MR imaging (29). In our series, stents located in the upper arm and shoulder region caused substantial imaging artifacts in one segment. The use of MR angiography on a routine basis may also be hampered by the presence of contraindications for this examination. In our series, MR-related contraindications precluded MR angiography in one of every nine study candidates.
In conclusion, high-spatial-resolution three-dimensional contrast-enhanced MR angiography offers the opportunity to reliably evaluate dysfunctional hemodialysis accesses. However, we discourage routine MR angiography evaluation of these accesses because of the current absence of viable MR-guided interventions. We advise that MR angiography be restricted to the evaluation of a dysfunctional access if the results of a DSA examination are inconclusive.
|
FOOTNOTES |
|---|
Authors stated no financial relationship to disclose.
Author contributions: Guarantors of integrity of entire study, C.L.F., L.E.M.D.; study concepts and design, L.E.M.D.; literature research, C.L.F., L.E.M.D.; clinical studies, all authors; data acquisition, C.L.F., L.E.M.D., A.V.T., A.B.D., H.C.M.v.d.B.; data analysis/interpretation, all authors; statistical analysis, L.E.M.D., Y.S.L.; manuscript preparation, Y.S.L.; manuscript definition of intellectual content, all authors; manuscript editing, L.E.M.D., Y.S.L., H.C.M.v.d.B.; manuscript revision/review and final version approval, all authors
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. Nael, M. Krishnam, S. G. Ruehm, H. J. Michaely, G. Laub, and J. P. Finn Time-Resolved MR Angiography in the Evaluation of Central Thoracic Venous Occlusive Disease Am. J. Roentgenol., June 1, 2009; 192(6): 1731 - 1738. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Heye, G. Maleux, K. Claes, D. Kuypers, and R. Oyen Stenosis Detection in Native Hemodialysis Fistulas with MDCT Angiography Am. J. Roentgenol., April 1, 2009; 192(4): 1079 - 1084. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. E. M. Duijm, E. H. Overbosch, Y. S. Liem, R. N. Planken, J. H. M. Tordoir, P. W. M. Cuypers, P. Douwes-Draaijer, and M. W. de Haan Retrograde catheterization of haemodialysis fistulae and grafts: angiographic depiction of the entire vascular access tree and stenosis treatment Nephrol. Dial. Transplant., February 1, 2009; 24(2): 539 - 547. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Y. Kim, R. A. Mirza, J. A. Bryant, E. D. Whiting, D. M. Delong, C. E. Spritzer, and E. M. Merkle Central Veins of the Chest: Evaluation with Time-resolved MR Angiography Radiology, May 1, 2008; 247(2): 558 - 566. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Zhang, E. M. Hecht, T. Maldonado, and V. S. Lee Time-Resolved 3D MR Angiography with Parallel Imaging for Evaluation of Hemodialysis Fistulas and Grafts: Initial Experience. Am. J. Roentgenol., May 1, 2006; 186(5): 1436 - 1442. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
