HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Michaely, H. J. Articles by Reiser, M. F. Search for Related Content PubMed PubMed Citation Articles by Michaely, H. J. Articles by Reiser, M. F.

Renal Artery Stenosis: Functional Assessment with Dynamic MR Perfusion Measurements—Feasibility Study¹

¹ From the Institute of Clinical Radiology, University Hospitals-Grosshadern (H.J.M., S.O.S., D.F., C.B., A.S., J.R., M.F.R.), Department of Internal Medicine, Division of Nephrology (W.S.), and Department of Pathology (M.W.), Ludwig-Maximilians-University, Marchioninistrasse 15, 81377 Munich, Germany; Siemens Medical Solutions, Erlangen, Germany (N.O.); Central Unit Biostatistics, German Cancer Research Center, Heidelberg, Germany (C.I.); and Research and Development, GE Healthcare Bio-Sciences, Ismaning, Germany (C.R.). From the 2004 RSNA Annual Meeting. Received September 8, 2004; revision requested November 12; revision received February 1, 2005; accepted February 25; final version accepted April 5. Address correspondence to H.J.M. (e-mail: henrik.michaely{at}med.uni-muenchen.de).

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION References

 
Purpose: To prospectively assess feasibility of renal magnetic resonance (MR) perfusion measurement method based on turbo fast low-angle shot sequences for grading effect of renal artery stenosis (RAS) on parenchymal perfusion.

Materials and Methods: Institutional review board approved this study, and patients gave written consent. Seventy-three patients (34 male, 39 female; age range, 17–84 years) who were clinically suspected of having RAS underwent contrast material–enhanced (gadodiamide) saturation-recovery turbo fast low-angle shot imaging for measurement of renal perfusion and high-spatial-resolution MR angiography for RAS detection and grading. Degree of stenosis was evaluated as high grade (75% stenosis), low to intermediate grade (>0% to <75% stenosis), or absent. High temporal resolution of the turbo fast low-angle shot sequence allowed acquisition of an exact first-pass tracing of the contrast agent bolus from which a signal intensity (SI)–time curve was derived. On the basis of this curve, mean transit time (MTT) of the contrast agent bolus, maximal upslope (MUS) of the curve, maximum SI, and time to SI peak (TTP) were calculated with a gamma variate fit. Wilcoxon rank sum test, Pearson product moment correlation, and paired t test were used for statistical analysis.

Results: Twenty-four renal arteries had high-grade RAS, 12 renal arteries had low- to intermediate-grade RAS, and 104 renal arteries had no RAS. Significant differences between patients without stenoses or with low- to intermediate-grade stenoses and patients with high-grade stenoses were found for MTT, MUS, and TTP (P < .001). Perfusion parameters were correlated with patients' serum creatinine levels, and significant correlations were found for MTT (r = 0.41), MUS (r = 0.48), and TTP (r = 0.4), with P < .001.

Conclusion: MR perfusion parameters can be used to assess effect of RAS on parenchymal perfusion. Perfusion measurements reflect renal function as measured with serum creatinine levels.

© RSNA, 2006

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION References

 
Magnetic resonance (MR) imaging of the kidneys, which includes the visualization of renal morphologic features and the grading of renal artery stenosis (RAS) by using MR angiography (1–4), has been widely used in the past years. In addition, MR flow measurements are now routinely performed by several groups to evaluate the hemodynamic significance of a detected stenosis (5–7). These methods, however, are not used to assess the effect of RAS on the parenchyma. This is why innovations in MR perfusion measurement (8–12) have been proposed. Renal perfusion measurements may complement the current examination protocols for evaluation of patients who are suspected of having RAS when findings at examination with conventional sequences and MR angiography cannot be used to establish the diagnosis. This application includes the detection of segmental RAS (13), the assessment of in-stent restenosis (2,14,15) after stent placement, or the detection of parenchymal disease in the absence of RAS. The clinical relevance of perfusion measurement is underlined by studies that show that reduced renal perfusion is associated with increased rates of morbidity and mortality in different settings (16,17).

We considered it possible to develop a method to derive perfusion parameters from bolus contrast enhancement measurements of gadolinium chelates in the kidney. Since the contrast agent can be assumed not to leave the vessel bed in the first pass, the enhancement can be assumed to reflect renal perfusion. Therefore, our method will be referred to as perfusion measurement in the following text. The aim of our study, therefore, was to prospectively assess the feasibility of a method of renal MR perfusion measurement based on turbo fast low-angle shot sequences (18) for grading of the effect of RAS on parenchymal perfusion.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION References

 
For this study, free contrast agent was provided by GE Healthcare Bio-Sciences, Ismaning, Germany. The authors who were not employees of GE Healthcare Bio-Sciences at every time had full control of inclusion of any data and information that might have presented a conflict of interest for those authors who were employees of GE Healthcare Bio-Sciences.

Patients
Seventy-three consecutive patients (34 male and 39 female patients; mean age, 54.7 years; range, 17–84 years) who were referred because they were suspected of having RAS were examined between June 2003 and April 2004. Patients were included when there was a strong clinical suspicion of RAS that was determined by the presence or the new onset of noncontrollable hypertension, positive or nonconclusive renal Doppler ultrasonographic (US) results, and/or clinical findings suggestive of RAS, as described by Rundback et al (19). Because five patients had undergone nephrectomy and one additional patient had a single kidney, a total of 140 kidneys were examined. Serum creatinine levels were recorded in all patients. The mean interval between the MR imaging examination and the creatinine sampling was 1.8 days (range, same day as examination to 12 days before or after examination). The degree of stenosis was determined, with high-spatial-resolution MR angiography as the reference standard, to be either high grade (75% stenosis), low to intermediate grade (>0% to <75% stenosis), or absent. The study was approved by the institutional review board, and oral and written consent were obtained from all patients.

MR Imaging
All experiments were performed with a 1.5-T imager (Magnetom Sonata; Siemens Medical Systems, Erlangen, Germany) with a maximum gradient strength of 40 mT/m, a minimum rise time of 200 µsec, and eight receiver channels. For signal reception, a recently introduced dedicated 12-element phased-array coil system was used, and this system consisted of one anterior and one posterior flexible coil element and each coil element had a set of six receiver elements. Outside elements on each side were combined to fit the limit of eight receiver channels. For the administration of the contrast agent, an MR imaging–compatible power injector (Spectris Solaris; Medrad, Indianola, Pa) was used. The MR imaging examination included morphologic true fast imaging with steady-state precession sequences in all orientations, to allow proper positioning for the subsequent MR perfusion measurements, and MR angiography.

MR perfusion imaging.—A turbo fast low-angle shot sequence with magnetization preparation was used for the perfusion measurement. This type of MR imaging sequence grants a linear relationship between contrast medium concentration and signal intensity (SI) (20,21). The linearity between contrast medium concentration and SI has been proved in previous phantom measurements for a wide range of different concentrations up to 1.1 mmol/L. The peak tissue concentrations and the peak blood concentrations were within the linear area. To obtain a complete and homogeneous saturation over the complete selected imaging volume, a modified single-shot fast low-angle shot sequence with saturation-recovery preparation and linear reordering was employed. For magnetization preparation, a pulse train of three short /2 pulses with constant amplitude and phase cycling in a phase angle of /2 was applied. The pulse train was followed by a 10 mT/m spoiler gradient over 1 msec. Use of more than three pulses did not improve the saturation effect substantially but made the preparation phase longer. Different intervals between the pulses and the various spoiler moments were employed to improve the global spatial saturation of magnetization prior to the fast low-angle shot readout kernel. The time between the preparation pulses and the acquisition of the central image line was 131 msec (inversion time).

The bandwidth of the pixels was set to 980 Hz to realize a short echo time of 1.04 msec. The image matrix was 256 x 110, which provided an image interval of 254 msec per image and allowed us to acquire four images per second. The pixels were interpolated to a rectangular pixel size. The remaining parameters of the sequence were as follows: repetition time, 245 msec; field of view in the read direction, 350 mm; field of view in the phase direction, 100%; matrix, 256 x 110; flip angle, 12°; number of measurements, 250; number of signals acquired, one; total measurement time, 254 seconds. A temporal resolution of 1 second per group of four sections was achieved. A pixel spacing of 1.56 mm was calculated. The section thickness was 8 mm, resulting in a voxel size of 20 mm³. The kidneys were imaged consecutively with four sections in an oblique coronal orientation, slightly tilted parallel to fit the long axis of the organ. These four sections were positioned to cover the entire kidney.

Because of the inherently low signal-to-noise ratio of turbo fast low-angle shot sequences, a standard bolus of 0.1 mmol per kilogram body weight of gadodiamide (Omniscan; Amersham Health, Little Chalfont, England) was injected into an antecubital vein at a flow rate of 4 mL/sec for the perfusion measurements. Injection of the contrast agent was followed by the injection of 30 mL of saline at a flow rate of 4 mL/sec. Patients were instructed to hold their breath as long as possible and to breathe regularly after the breath-hold period. The total acquisition time for the perfusion sequence was 4 minutes 14 seconds. The examination was paused after the perfusion measurement for another 2 minutes to allow the contrast agent to leave the vessels.

MR angiography.—MR angiography was performed with parallel imaging, and such imaging allowed an isotropic spatial resolution of 0.9 x 0.8 x 0.9 mm³ in a 23-second breath hold. The parameters were as follows: repetition time msec/echo time msec, 3.79/1.39; field of view in the read direction, 400 mm; field of view in the phase direction, 87%; number of sections per slab, 80; section thickness, 0.9 mm; number of frequency encoding steps, 512; flip angle, 25°; and bandwidth, 350 Hz/pixel. Parallel imaging was performed with an automatic simultaneous acquisition of spatial harmonics–based generalized autocalibrating partially parallel acquisition algorithm (22). Because of the integrated acquisition of reference lines, a linear k-space acquisition mode had to be chosen for the three-dimensional MR angiographic sequence with gadolinium-based contrast agent. Appropriate timing of the administration of the contrast agent bolus was ensured with the application of a test-bolus technique (23). The perfusion measurement was not used as a test bolus because of its different rate of gadolinium-based contrast agent injection of 4 mL/sec. Instead, a separate test bolus of 2 mL gadodiamide at 2 mL/sec was used. For MR angiography, a bolus of 0.2 mmol/kg gadodiamide was injected into an antecubital vein at a flow rate of 2 mL/sec. Injection of contrast agent was followed by the injection of 30 mL of normal saline at a flow rate of 2 mL/sec.

SI-Time Curve Analysis
The SI-time curve obtained from a region of interest over the renal cortex consisted of two overlying parts: One curve represents the first pass of the bolus of contrast agent in the cortical vessels; and the other, the filtration of the contrast agent into the glomeruli. Figure 1 displays images that are examples of the perfusion measurement of a single section. Figure 2 demonstrates the subsequently derived SI-time curve with its two distinct components. The analysis of the SI-time curves focused on the separation of the first-pass effects from the filtration and excretion parts of the curve. It was therefore necessary to perform imaging with a high temporal resolution. The model aims to reflect the SI-time slope as closely and robustly as possible within the scope of an automatic parametric calculation. The gamma variate for the first-pass perfusion (Eq [1]) and an exponential function for the filtration (Eq [2]) can be mathematically combined to yield the resulting fit (Eq [3]). Equation (1) is as follows:

(1)

where g(t) is the function that describes the first-pass part of the perfusion measurement, and t is the time after the start of the sequence. Equation (2) is presented thus:

(2)

where c(t) is the function that describes the filtration part of the perfusion measurement, and is the factor that determines the beginning of the filtration part in relation to the first-pass part. Equation (3) was calculated thus:

(3)

where s(t) is the overall function composed of g(t) and c(t) after correction to baseline, S(t) is the overall function before baseline correction, and T₀ is the appearance time of the contrast agent. The parameters p₀ through p₈ are explained in Table 1.

View larger version (138K): [in this window] [in a new window] [Download PPT slide]   Figure 1a: Coronal saturation-recovery turbo fast low-angle shot MR images (245/1.04; flip angle, 12°) show perfusion measurement in a normal kidney of a 22-year-old woman. (a) Native image before arrival of contrast agent bolus. (b) Image shows early corticomedullary differentiation 7 seconds after arrival of contrast agent in kidneys. (c) Image in medullary phase with decreased cortical enhancement and increased medullary contrast agent enhancement 120 seconds after initial arrival of contrast agent in kidneys. (d) Excretory phase image obtained 200 seconds after contrast agent arrival with clearly defined caliceal system and proximal ureter. Note marked difference in cortical enhancement compared with that on b.

View larger version (162K): [in this window] [in a new window] [Download PPT slide]   Figure 1b: Coronal saturation-recovery turbo fast low-angle shot MR images (245/1.04; flip angle, 12°) show perfusion measurement in a normal kidney of a 22-year-old woman. (a) Native image before arrival of contrast agent bolus. (b) Image shows early corticomedullary differentiation 7 seconds after arrival of contrast agent in kidneys. (c) Image in medullary phase with decreased cortical enhancement and increased medullary contrast agent enhancement 120 seconds after initial arrival of contrast agent in kidneys. (d) Excretory phase image obtained 200 seconds after contrast agent arrival with clearly defined caliceal system and proximal ureter. Note marked difference in cortical enhancement compared with that on b.

View larger version (144K): [in this window] [in a new window] [Download PPT slide]   Figure 1c: Coronal saturation-recovery turbo fast low-angle shot MR images (245/1.04; flip angle, 12°) show perfusion measurement in a normal kidney of a 22-year-old woman. (a) Native image before arrival of contrast agent bolus. (b) Image shows early corticomedullary differentiation 7 seconds after arrival of contrast agent in kidneys. (c) Image in medullary phase with decreased cortical enhancement and increased medullary contrast agent enhancement 120 seconds after initial arrival of contrast agent in kidneys. (d) Excretory phase image obtained 200 seconds after contrast agent arrival with clearly defined caliceal system and proximal ureter. Note marked difference in cortical enhancement compared with that on b.

View larger version (143K): [in this window] [in a new window] [Download PPT slide]   Figure 1d: Coronal saturation-recovery turbo fast low-angle shot MR images (245/1.04; flip angle, 12°) show perfusion measurement in a normal kidney of a 22-year-old woman. (a) Native image before arrival of contrast agent bolus. (b) Image shows early corticomedullary differentiation 7 seconds after arrival of contrast agent in kidneys. (c) Image in medullary phase with decreased cortical enhancement and increased medullary contrast agent enhancement 120 seconds after initial arrival of contrast agent in kidneys. (d) Excretory phase image obtained 200 seconds after contrast agent arrival with clearly defined caliceal system and proximal ureter. Note marked difference in cortical enhancement compared with that on b.

View larger version (13K): [in this window] [in a new window] [Download PPT slide]   Figure 2a: SI-time curves of renal cortex of healthy kidney in 22-year-old woman. (a) Bold curve signifies the first pass, as well as the beginning filtration, of the contrast agent. This bold curve consists of (b) a separate first-pass curve and (c) a filtration curve, which are shown in a with the fine line and the dotted-dashed line, respectively. The first-pass curve is described with a gamma variate fit, whereas the filtration curve is characterized by an exponential function.

View larger version (12K): [in this window] [in a new window] [Download PPT slide]   Figure 2b: SI-time curves of renal cortex of healthy kidney in 22-year-old woman. (a) Bold curve signifies the first pass, as well as the beginning filtration, of the contrast agent. This bold curve consists of (b) a separate first-pass curve and (c) a filtration curve, which are shown in a with the fine line and the dotted-dashed line, respectively. The first-pass curve is described with a gamma variate fit, whereas the filtration curve is characterized by an exponential function.

View larger version (10K): [in this window] [in a new window] [Download PPT slide]   Figure 2c: SI-time curves of renal cortex of healthy kidney in 22-year-old woman. (a) Bold curve signifies the first pass, as well as the beginning filtration, of the contrast agent. This bold curve consists of (b) a separate first-pass curve and (c) a filtration curve, which are shown in a with the fine line and the dotted-dashed line, respectively. The first-pass curve is described with a gamma variate fit, whereas the filtration curve is characterized by an exponential function.

View larger version (25K): [in this window] [in a new window] [Download PPT slide]   Figure 3: Graphic depiction of perfusion parameters. Perfusion parameters are derived from the gamma variate fit of the first-pass perfusion (dashed line). The maximum SI is marked as 1; MUS, as 2; MTT, as 3; and TTP, as 4.

All parameters were fitted automatically by using software (Merz, version 0.92; Siemens Medical Systems). If desired, it is possible to interact with the fitting process by confining the intervals and parameters for finding reasonable start values. The fitting was performed with a slightly modified Marquardt-Levenberg least squares algorithm (28). For this work, only the initial part of the curve that represented the first pass was analyzed.

Data Analysis
The three-dimensional MR angiographic data sets obtained with gadolinium-based contrast agent were independently evaluated by two board-certified attending radiologists with special training in vascular and interventional radiology (S.O.S., with 8 years of training, and J.R., with 9 years of training). All data sets were reviewed electronically at a dedicated three-dimensional postprocessing workstation to allow electronic measurements, windowing, and three-dimensional reformatting of the data sets. The degree of stenosis was measured by using calipers in the in-plane view of thin maximum intensity projections of 10-mm thickness. For further analysis, kidneys were classified into three groups on the basis of the degree of in-plane RAS according to the North American Symptomatic Carotid Endarterectomy Trial criteria (29): ((a) no RAS, (b) low- to intermediate-grade RAS (<75% in-plane stenosis), and (c) high-grade RAS (75% in-plane stenosis). In addition, the location of RAS was determined to be proximal, distal, or segmental.

MR perfusion measurements were analyzed (H.J.M., with 4 years of experience with MR imaging of the kidneys, and C.B., with 2 years of the same type of experience) at a separate workstation by using the Merz software. This software allows a region-of-interest–based evaluation of contrast agent enhancement on a pixel-by-pixel basis. To correct for breathing-related motion, the Merz software contains an automated pixel-by-pixel motion correction. With Merz software, the mean SI of the region of interest was plotted over time. This SI of a pixel was considered to originate from the intravascular contrast agent during the first pass only. As the measurement continued and the intravascular contrast agent concentration decreased, the SI mainly resulted from the filtered contrast agent. Derived from the parameters of the SI curve, the software calculated four perfusion parameters that are based on a gamma variate fit, which was explained previously as follows: TTP, MTT, MUS, and maximal SI. The Merz software also can represent these kidney perfusion parameters as color-coded maps, which permit visual assessment of regional perfusion.

For the analysis, two sections with the best coverage of the kidneys were chosen. To compensate for intraobserver variability, two regions of interest were placed in each kidney by one of the investigators (H.J.M. or C.B.) in the two sections over the renal cortex of each kidney but sparing the surrounding fat, as well as the renal medulla. The regions of interest were at least 400 pixels and were positioned over the entire renal cortex. The mean of each parameter derived from these two regions of interest was used for further evaluations. At the time of perfusion analysis, the investigators were not aware of the presence of RAS. All data sets were shown in an anonymous fashion and were identified only with a four-digit number.

To evaluate interobserver variability in the perfusion measurement analysis, data from the first 40 kidneys were independently analyzed by both readers (H.J.M., C.B.), who were blinded to the presence of an RAS. In addition, a qualitative assessment of the time-SI curve was performed, with a focus on the steepness of the upslope and the time of the peak SI.

Statistical Analysis
The Wilcoxon rank sum test was used to compare all four perfusion parameters (MTT, MUS, TTP, MSI) for kidneys without RAS, kidneys with low- to intermediate-grade RAS, and kidneys with high-grade RAS. P values were adjusted according to the Bonferroni-Holm method to control the familywise error rate at a level of = 5%. Paired t tests were performed to assess the age-related differences among the three groups. Normal distribution of the data for the perfusion parameters and age was assessed with the Shapiro-Wilk test, and parametric and nonparametric tests were performed in accordance with the results. Correlations between the parameters were estimated with the Pearson product moment correlation coefficient together with the 95% confidence interval (CI).

Because unilateral RAS could possibly lead to altered blood flow even in the nonaffected kidney with activation of the renin-angiotensin system, a random sample (n = 73) was used for the statistical analysis and included only one kidney per patient. All analyses were performed by using the entire sample and the random sample. Only those results were deemed significant that showed significance in the full sample, as well as in the random sample. Adjusted P values in the Results section refer to the results for the full sample. Results also could be displayed in color-coded parametric maps for visualization of areas with impaired renal perfusion. For the comparison between the perfusion parameters with the serum creatinine levels and the patients' age, correlation analyses were performed. Pearson product moment correlation coefficients, together with the 95% CIs, were calculated as well. For the estimation of interobserver variability, the results of both investigators were correlated. The statistical analysis was performed by using software (R, version 1.8.1; R-Foundation for Statistical Computing, Vienna, Austria).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION References

 
No kidney had to be excluded from the analysis because the signal-to-noise ratio was too low. In all 140 kidneys, the parameters could be determined to fit the curve adequately. In patients without RAS, the SI-time curves showed a typical pattern characterized by three distinct features: (a) a steep upslope and (b) an early SI peak with (c) a subsequent slow decline of the signal (Fig 4a, 4b). In kidneys with high-grade RAS, a slowed upslope, a flattened peak, and a delayed signal decay in the renal cortex were found (Fig 4c, 4d). In low- to intermediate-grade RAS, the characteristic SI-time curve was not observed. This finding also can be seen in Figure 5, which demonstrates the maximum intensity projection views and perfusion measurements in the same patient before and after percutaneous angioplasty for high-grade RAS. The previously decreased perfusion recovered after successful therapy.

View larger version (158K): [in this window] [in a new window] [Download PPT slide]   Figure 4a: (a) Coronal 30-mm thin maximum intensity projection obtained with contrast material–enhanced gradient-echo sequence (3.79/1.39; flip angle, 25°) shows healthy kidney without RAS in 42-year-old man. (b) SI-time curve for renal cortex in healthy kidney shows steep upslope, early peak, and subsequent decay of the signal in the cortex. (c) Coronal 30-mm thin maximum intensity projection obtained with contrast-enhanced gradient-echo sequence (3.79/1.39; flip angle, 25°) shows kidney with RAS (arrowhead) in 56-year-old man. (d) SI-time curve for kidney with RAS shows slowed upslope, flattened peak, and delayed signal decay in renal cortex.

View larger version (11K): [in this window] [in a new window] [Download PPT slide]   Figure 4b: (a) Coronal 30-mm thin maximum intensity projection obtained with contrast material–enhanced gradient-echo sequence (3.79/1.39; flip angle, 25°) shows healthy kidney without RAS in 42-year-old man. (b) SI-time curve for renal cortex in healthy kidney shows steep upslope, early peak, and subsequent decay of the signal in the cortex. (c) Coronal 30-mm thin maximum intensity projection obtained with contrast-enhanced gradient-echo sequence (3.79/1.39; flip angle, 25°) shows kidney with RAS (arrowhead) in 56-year-old man. (d) SI-time curve for kidney with RAS shows slowed upslope, flattened peak, and delayed signal decay in renal cortex.

View larger version (109K): [in this window] [in a new window] [Download PPT slide]   Figure 4c: (a) Coronal 30-mm thin maximum intensity projection obtained with contrast material–enhanced gradient-echo sequence (3.79/1.39; flip angle, 25°) shows healthy kidney without RAS in 42-year-old man. (b) SI-time curve for renal cortex in healthy kidney shows steep upslope, early peak, and subsequent decay of the signal in the cortex. (c) Coronal 30-mm thin maximum intensity projection obtained with contrast-enhanced gradient-echo sequence (3.79/1.39; flip angle, 25°) shows kidney with RAS (arrowhead) in 56-year-old man. (d) SI-time curve for kidney with RAS shows slowed upslope, flattened peak, and delayed signal decay in renal cortex.

View larger version (11K): [in this window] [in a new window] [Download PPT slide]   Figure 4d: (a) Coronal 30-mm thin maximum intensity projection obtained with contrast material–enhanced gradient-echo sequence (3.79/1.39; flip angle, 25°) shows healthy kidney without RAS in 42-year-old man. (b) SI-time curve for renal cortex in healthy kidney shows steep upslope, early peak, and subsequent decay of the signal in the cortex. (c) Coronal 30-mm thin maximum intensity projection obtained with contrast-enhanced gradient-echo sequence (3.79/1.39; flip angle, 25°) shows kidney with RAS (arrowhead) in 56-year-old man. (d) SI-time curve for kidney with RAS shows slowed upslope, flattened peak, and delayed signal decay in renal cortex.

View larger version (152K): [in this window] [in a new window] [Download PPT slide]   Figure 5a: (a) Coronal 30-mm thin maximum intensity projection obtained with contrast-enhanced gradient-echo sequence (3.79/1.39; flip angle, 25°) in 76-year-old woman with high-grade RAS (arrowhead) of left renal artery. (b) Coronal 30-mm thin maximum intensity projection obtained with contrast-enhanced gradient-echo sequence (3.79/1.39; flip angle, 25°) in same patient after successful angioplasty with stent placement. Because of susceptibility artifacts (arrowhead) from the stent, vessel lumen could not be seen after stent placement. (c) SI-time curves for the affected kidney returned to normal after therapy. Pretherapy cortical SI-time curve is marked by red line, which reveals typical slowed upslope and flattened peak. Postintervention cortical SI-time curve (black line) returns to normal with increased upslope, pronounced peak, and slow decay of signal after the peak. A.U. = arbitrary units.

View larger version (160K): [in this window] [in a new window] [Download PPT slide]   Figure 5b: (a) Coronal 30-mm thin maximum intensity projection obtained with contrast-enhanced gradient-echo sequence (3.79/1.39; flip angle, 25°) in 76-year-old woman with high-grade RAS (arrowhead) of left renal artery. (b) Coronal 30-mm thin maximum intensity projection obtained with contrast-enhanced gradient-echo sequence (3.79/1.39; flip angle, 25°) in same patient after successful angioplasty with stent placement. Because of susceptibility artifacts (arrowhead) from the stent, vessel lumen could not be seen after stent placement. (c) SI-time curves for the affected kidney returned to normal after therapy. Pretherapy cortical SI-time curve is marked by red line, which reveals typical slowed upslope and flattened peak. Postintervention cortical SI-time curve (black line) returns to normal with increased upslope, pronounced peak, and slow decay of signal after the peak. A.U. = arbitrary units.

View larger version (17K): [in this window] [in a new window] [Download PPT slide]   Figure 5c: (a) Coronal 30-mm thin maximum intensity projection obtained with contrast-enhanced gradient-echo sequence (3.79/1.39; flip angle, 25°) in 76-year-old woman with high-grade RAS (arrowhead) of left renal artery. (b) Coronal 30-mm thin maximum intensity projection obtained with contrast-enhanced gradient-echo sequence (3.79/1.39; flip angle, 25°) in same patient after successful angioplasty with stent placement. Because of susceptibility artifacts (arrowhead) from the stent, vessel lumen could not be seen after stent placement. (c) SI-time curves for the affected kidney returned to normal after therapy. Pretherapy cortical SI-time curve is marked by red line, which reveals typical slowed upslope and flattened peak. Postintervention cortical SI-time curve (black line) returns to normal with increased upslope, pronounced peak, and slow decay of signal after the peak. A.U. = arbitrary units.

Differences between Grades of Stenoses
All four parameters differed substantially for kidneys without stenosis, those with low- to intermediate-grade stenosis, and those with high-grade stenosis. Significant differences in the parameters MUS, MTT, and TTP (adjusted P values of <.001) were found between kidneys without RAS or kidneys with low- to intermediate-grade RAS and those with high-grade RAS. Although significant differences in MUS were observed between healthy kidneys and those with high-grade RAS, MUS is variable in healthy kidneys with a low MTT. No significant differences in maximal SI were observed. There were no significant differences in any parameter between kidneys without RAS and those with low- to intermediate-grade RAS (Fig 6, Table 2). An excellent correlation between MTT and TTP (r = 0.96; 95% CI: 0.94, 0.98; P < .001) and a good correlation between TTP and MUS (r = –0.60; 95% CI: –0.73, –0.43; P < .001) were observed. No significant relationship between sex and any of the perfusion parameters could be demonstrated.

View larger version (17K): [in this window] [in a new window] [Download PPT slide]   Figure 6: Box-and-whiskers plots for MTT in groups without RAS (degree of stenosis, 0), low- to intermediate-grade RAS (degree of stenosis, 1), and high-grade RAS (degree of stenosis, 2). MTT was significantly (P < .001) lower in kidneys without high-grade RAS regardless of whether low- to intermediate-grade RAS was present.

Segmental Perfusion Deficits
In three kidneys, segmental perfusion deficits were detected. In two of these kidneys, segmental RAS was revealed; in the third kidney, biopsy-proved regional ischemia was found. Color-coded parametric maps proved to be especially helpful in the detection of segmental perfusion deficits. In the patient whose images are shown in Figure 7 , no pathologic findings were revealed at MR angiography; however, segmental RAS consistent with fibromuscular dysplasia was demonstrated at intraarterial digital subtraction angiography. The color-coded parametric map of this patient's kidney showed decreased perfusion at the affected lower pole. In the patient whose images are shown in Figure 8, ischemic changes were observed in the upper pole of the kidney in the absence of RAS; these were also demonstrated with color-coded parametric maps and were proved with biopsy.

View larger version (81K): [in this window] [in a new window] [Download PPT slide]   Figure 7a: Segmental perfusion deficit in 35-year-old man with segmental fibromuscular dysplasia of lower pole of renal artery. (a–c) Coronal perfusion images (saturation-recovery turbo fast low-angle shot, 245/1.04; flip angle, 12°) show decreased enhancement at affected lower pole compared with rest of kidney. (d) Color-coded parametric map of MUS shows decreased or impaired SI. Affected pole (open arrowhead) shows lower SI than rest of the kidney, consistent with slowed flow in this area. Note lack of SI at upper pole (solid arrowhead), which represents kidney infarction after attempted stent placement in segmental artery at another hospital.

View larger version (98K): [in this window] [in a new window] [Download PPT slide]   Figure 7b: Segmental perfusion deficit in 35-year-old man with segmental fibromuscular dysplasia of lower pole of renal artery. (a–c) Coronal perfusion images (saturation-recovery turbo fast low-angle shot, 245/1.04; flip angle, 12°) show decreased enhancement at affected lower pole compared with rest of kidney. (d) Color-coded parametric map of MUS shows decreased or impaired SI. Affected pole (open arrowhead) shows lower SI than rest of the kidney, consistent with slowed flow in this area. Note lack of SI at upper pole (solid arrowhead), which represents kidney infarction after attempted stent placement in segmental artery at another hospital.

View larger version (109K): [in this window] [in a new window] [Download PPT slide]   Figure 7c: Segmental perfusion deficit in 35-year-old man with segmental fibromuscular dysplasia of lower pole of renal artery. (a–c) Coronal perfusion images (saturation-recovery turbo fast low-angle shot, 245/1.04; flip angle, 12°) show decreased enhancement at affected lower pole compared with rest of kidney. (d) Color-coded parametric map of MUS shows decreased or impaired SI. Affected pole (open arrowhead) shows lower SI than rest of the kidney, consistent with slowed flow in this area. Note lack of SI at upper pole (solid arrowhead), which represents kidney infarction after attempted stent placement in segmental artery at another hospital.

View larger version (77K): [in this window] [in a new window] [Download PPT slide]   Figure 7d: Segmental perfusion deficit in 35-year-old man with segmental fibromuscular dysplasia of lower pole of renal artery. (a–c) Coronal perfusion images (saturation-recovery turbo fast low-angle shot, 245/1.04; flip angle, 12°) show decreased enhancement at affected lower pole compared with rest of kidney. (d) Color-coded parametric map of MUS shows decreased or impaired SI. Affected pole (open arrowhead) shows lower SI than rest of the kidney, consistent with slowed flow in this area. Note lack of SI at upper pole (solid arrowhead), which represents kidney infarction after attempted stent placement in segmental artery at another hospital.

View larger version (19K): [in this window] [in a new window] [Download PPT slide]   Figure 8a: Segmental perfusion deficit in 60-year-old man 7 years after renal transplantation who was suspected of having RAS, with rising serum creatinine level. No RAS was found at MR angiography. (a) SI-time curve shows that results of perfusion measurement were conspicuous: SI-time curve for lower pole (black line) was normal, whereas that for upper pole (gray line) revealed decreased perfusion. (b) Color-coded parametric map graphically depicts reduced perfusion at the upper pole, with reduced blood flow in the entire upper pole compared with that in the lower pole. Bright green indicates high MUS, which is equivalent to better perfusion. (c) Biopsy specimen from upper pole showed chronic ischemic reaction with dilated and partly atrophic tubules and interstitial fibrosis. (Hematoxylin-eosin stain; original magnification, x50.)

View larger version (103K): [in this window] [in a new window] [Download PPT slide]   Figure 8b: Segmental perfusion deficit in 60-year-old man 7 years after renal transplantation who was suspected of having RAS, with rising serum creatinine level. No RAS was found at MR angiography. (a) SI-time curve shows that results of perfusion measurement were conspicuous: SI-time curve for lower pole (black line) was normal, whereas that for upper pole (gray line) revealed decreased perfusion. (b) Color-coded parametric map graphically depicts reduced perfusion at the upper pole, with reduced blood flow in the entire upper pole compared with that in the lower pole. Bright green indicates high MUS, which is equivalent to better perfusion. (c) Biopsy specimen from upper pole showed chronic ischemic reaction with dilated and partly atrophic tubules and interstitial fibrosis. (Hematoxylin-eosin stain; original magnification, x50.)

View larger version (104K): [in this window] [in a new window] [Download PPT slide]   Figure 8c: Segmental perfusion deficit in 60-year-old man 7 years after renal transplantation who was suspected of having RAS, with rising serum creatinine level. No RAS was found at MR angiography. (a) SI-time curve shows that results of perfusion measurement were conspicuous: SI-time curve for lower pole (black line) was normal, whereas that for upper pole (gray line) revealed decreased perfusion. (b) Color-coded parametric map graphically depicts reduced perfusion at the upper pole, with reduced blood flow in the entire upper pole compared with that in the lower pole. Bright green indicates high MUS, which is equivalent to better perfusion. (c) Biopsy specimen from upper pole showed chronic ischemic reaction with dilated and partly atrophic tubules and interstitial fibrosis. (Hematoxylin-eosin stain; original magnification, x50.)

View larger version (17K): [in this window] [in a new window] [Download PPT slide]   Figure 9: Scatterplot shows strong correlation between reduced MUS and elevated serum creatinine levels. A. U. = arbitrary units.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION References

In our study, patients who ranged in age from 17 to 84 years were examined. Although MTT and MUS were more often impaired (elevated MTT and decreased MUS) in older patients than in younger patients—most likely caused by the increased incidence of RAS in older patients—no significant correlation between age and any of the perfusion parameters could be identified. Therefore, these parameters seem to reflect a disease-associated change, rather than an age-related change, in renal perfusion. MTT showed the largest statistical difference between healthy renal arteries and highly stenotic renal arteries. The kidneys with intermediate-grade RAS showed elevated, but not statistically significant, differences in MTT. It was observed, however, that MTT rises with an increasing grade of RAS. This finding is in accordance with findings in previous studies (27) in which the MTT, which rises with decreasing renal function, was found to be a marker for parenchymal disease.

The MUS is a parameter, known from cardiac first-pass perfusion measurements, in which an upslope that is based on a few data points is used routinely (32). Although MUS, which has been proposed as an important parameter of renal perfusion (10), is significantly (P < .001) reduced in high-grade RAS as well, it is variable in healthy kidneys with a low MTT. Therefore, it appears that MTT better reflects the true first-pass kinetics. Therefore, it seems appropriate not to use MUS as a single parameter but to use it in concert with MTT. TTP has been introduced as another characteristic parameter of parenchymal blood flow (9). We found a good correlation between TTP and MTT. With regard to TTP, no significant differences were demonstrated between healthy kidneys or kidneys with low-grade RAS and kidneys with high-grade RAS. Further studies are indicated to confirm this result.

Compared with other contrast-enhanced techniques for measuring renal perfusion that have been presented in the literature (8–10), our approach has two intrinsic advantages. First, we achieved a very high temporal resolution of one acquisition per second, as well as a high spatial resolution with four sections, by using a dedicated MR imager with high gradient strengths. This high temporal resolution is particularly important when we monitor the first pass of the contrast agent through the kidneys. A lower temporal resolution may lead to insufficient data acquisition and obscure the initial upslope, the maximum SI, or the beginning of the signal decay. First, underestimation of renal perfusion can result from use of SI-time curves that are based on a data set with a low temporal resolution. Second, acquisition of four coronal sections through the kidneys provides a reasonable coverage of the renal parenchyma and allows the detection of segmental perfusion deficits.

With other renal perfusion methods reported in the literature, only one transverse section is acquired through both kidneys, and detection of segmental perfusion deficits is not permitted (8,9). In the study of Gandy et al (9), the delay time between aortic and renal SI peaks was considered to show characteristic changes with increasing RAS; however, the characteristics of the renal blood flow were not analyzed. The application of fast spoiled gradient-echo three-dimensional sequences without magnetization preparation, such as the volume-interpolated breath-hold examination sequence, has been proposed as well. The main advantage of the volume-interpolated breath-hold examination sequence is the acquisition of an entire three-dimensional data set of the kidneys in 3 seconds (11). This acquisition is at the cost of a lower temporal resolution, which may compromise the estimation of perfusion parameters. This cost of a lower temporal resolution is probably why the volume-interpolated breath-hold examination technique is aimed more at obtaining MR renograms than at assessing renal perfusion. Moreover, the volume-interpolated breath-hold examination technique does not grant linearity between SI and contrast agent concentration because of the lack of preparation pulses; therefore, a semiquantitative analysis is not possible.

Although the use of low-dose (2 mL) gadopentetate dimeglumine at a low flow rate has been favored in the literature (8), we preferred another approach with a higher dose of gadodiamide dimeglumine at a high flow rate. This granted us a higher contrast-to-noise ratio and prevented poor contrast enhancement from disqualifying any kidneys from analysis. The high flow rate also provides a sharper bolus geometry and leads to a more well-defined pharmacologic assessment (33). The higher flow rate of the bolus injection is a prerequisite for a valid and reliable model configuration of the renal compartments, and the model configuration serves as the basis for the SI-time curve analysis. Our study was focused on the assessment of the first-pass perfusion, which was based on a gamma variate fit, in the renal cortex. The next important steps would be the absolute quantification of the perfusion and the analysis of the filtration part of the SI-time curve. So far, absolute quantification has only been performed by using intravascular contrast agents (34–36). It has been difficult for extracellular contrast agents to fill this role. Analysis of the filtration curve of the contrast agent also is challenging, because no established model for the exchange processes of the contrast agent in the kidney exists to date, as far as we know. These steps, however, would enhance our assessment of renal function and renoparenchymal disease.

Renal scintigraphy is still considered the reference standard for renal perfusion. This test, however, is rarely ordered by the referring physicians at our institution. The reasons are mainly the long examination time, the use of ionizing radiation, the low spatial resolution, and the lack of absolute quantification. Because of these limitations, scintigraphy has been regarded mostly as a functional imaging modality. In contrast, the main indication for MR imaging has been the depiction of morphologic features. Now, MR flow and perfusion measurements also introduce functional aspects. MR flow measurements have been tested and validated extensively (7,29,37) in patients with RAS. Among the first published data for MR perfusion measurements were data in regard to changes in patients with RAS (9). Our results confirm these data. Our data also agree with previously published CT perfusion data in the study of Paul et al (38). In that previous study, the investigators found changes in the attenuation-time curves for patients with RAS that were similar to those we found in the SI-time curves.

In addition, we also found a significant correlation between the serum creatinine level and MTT, MUS, and TTP. MR perfusion measurements also reflected changes in the serum creatinine level. These results were confirmed by using the random sample, which included a single kidney per patient, to diminish the problem of secondary renal function impairment of a healthy kidney in the case of a contralateral kidney with RAS. Whether the perfusion parameters were altered because of the RAS-induced ischemic nephropathy or they reflect the overall renal function needs to be shown in further studies.

For further investigations, a more homogeneous patient population is desired. The somewhat heterogeneous patient collective—which included patients without abnormalities at MR imaging, patients with unilateral RAS, patients with bilateral RAS, and patients with additional parenchymal disease—was a limitation of our study. An intraindividual comparison between a healthy kidney and a kidney with RAS was not performed because of the small number of patients with only one affected kidney. This limitation may be overcome by including a larger number of patients in future studies. Another drawback of our study was the focus on only the cortical first-pass perfusion. With this method, contrast enhancement after injection of a gadolinium chelate was used as a surrogate for parenchymal perfusion, and only an indirect assessment of the renal perfusion could be achieved.

In summary, MR perfusion measurements with high temporal and spatial resolution are technically feasible and reliable. MR perfusion measurements allow a semiquantitative evaluation of parameters derived from the measurements, and results of this evaluation indicate that there are significant differences in these parameters between healthy kidneys and those with high-grade RAS.

	FOOTNOTES

 

Abbreviations: CI = confidence interval • MTT = mean transit time • MUS = maximal upslope • RAS = renal artery stenosis • SI = signal intensity • TTP = time to peak SI

See Materials and Methods for pertinent disclosures.

Author contributions: Guarantors of integrity of entire study, H.J.M., S.O.S., M.F.R.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manuscript, all authors; literature research, H.J.M., S.O.S., N.O., C.I., C.R., W.S., M.W., M.F.R.; clinical studies, H.J.M., S.O.S., N.O., D.F., C.B., A.S., J.R., W.S., M.W., M.F.R.; statistical analysis, H.J.M., S.O.S., C.I., C.R.; and manuscript editing, all authors

	References

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION References

 

1. Schoenberg SO, Rieger J, Nittka M, Dietrich O, Johannson LO, Reiser MF. Renal MR angiography: current debates and developments in imaging of renal artery stenosis. Semin Ultrasound CT MR 2003;24:255–267.[CrossRef][Medline]
2. Shetty AN, Bis KG, Kirsch M, Weintraub J, Laub G. Contrast-enhanced breath-hold three-dimensional magnetic resonance angiography in the evaluation of renal arteries: optimization of technique and pitfalls. J Magn Reson Imaging 2000;12:912–923.[CrossRef][Medline]
3. Tan KT, van Beek EJ, Brown PW, van Delden OM, Tijssen J, Ramsay LE. Magnetic resonance angiography for the diagnosis of renal artery stenosis: a meta-analysis. Clin Radiol 2002;57:617–624.[CrossRef][Medline]
4. Prince MR. Renal MR angiography: a comprehensive approach. J Magn Reson Imaging 1998;8:511–516.[Medline]
5. de Haan MW, Kouwenhoven M, Kessels AG, van Engelshoven JM. Renal artery blood flow: quantification with breath-hold or respiratory triggered phase-contrast MR imaging. Eur Radiol 2000;10:1133–1137.[CrossRef][Medline]
6. de Haan MW, van Engelshoven JM, Houben AJ, et al. Phase-contrast magnetic resonance flow quantification in renal arteries: comparison with 133Xenon washout measurements. Hypertension 2003;41:114–118.[Abstract/Free Full Text]
7. Schoenberg SO, Bock M, Kallinowski F, Just A. Correlation of hemodynamic impact and morphologic degree of renal artery stenosis in a canine model. J Am Soc Nephrol 2000;11:2190–2198.[Abstract/Free Full Text]
8. Lee VS, Rusinek H, Johnson G, Rofsky NM, Krinsky GA, Weinreb JC. MR renography with low-dose gadopentetate dimeglumine: feasibility. Radiology 2001;221:371–379.[Abstract/Free Full Text]
9. Gandy SJ, Sudarshan TA, Sheppard DG, Allan LC, McLeay TB, Houston JG. Dynamic MRI contrast enhancement of renal cortex: a functional assessment of renovascular disease in patients with renal artery stenosis. J Magn Reson Imaging 2003;18:461–466.[CrossRef][Medline]
10. Montet X, Ivancevic MK, Belenger J, et al. Noninvasive measurement of absolute renal perfusion by contrast medium-enhanced magnetic resonance imaging. Invest Radiol 2003;38:584–592.[Medline]
11. Lee VS, Rusinek H, Noz ME, Lee P, Raghavan M, Kramer EL. Dynamic three-dimensional MR renography for the measurement of single kidney function: initial experience. Radiology 2003;227:289–294.[Abstract/Free Full Text]
12. Karger N, Biederer J, Lusse S, et al. Quantitation of renal perfusion using arterial spin labeling with FAIR-UFLARE. Magn Reson Imaging 2000;18:641–647.[CrossRef][Medline]
13. Vasbinder GB, Maki JH, Nijenhuis RJ, et al. Motion of the distal renal artery during three-dimensional contrast-enhanced breath-hold MRA. J Magn Reson Imaging 2002;16:685–696.[CrossRef][Medline]
14. Spuentrup E, Ruebben A, Stuber M, Gunther RW, Buecker A. Metallic renal artery MR imaging stent: artifact-free lumen visualization with projection and standard renal MR angiography. Radiology 2003;227:897–902.[Abstract/Free Full Text]
15. Bartels LW, Bakker CJ, Viergever MA. Improved lumen visualization in metallic vascular implants by reducing RF artifacts. Magn Reson Med 2002;47:171–180.[CrossRef][Medline]
16. Fann JI, Sarris GE, Mitchell RS, et al. Treatment of patients with aortic dissection presenting with peripheral vascular complications. Ann Surg 1990;212:705–713.[Medline]
17. Hillege HL, Girbes AR, de Kam PJ, et al. Renal function, neurohormonal activation, and survival in patients with chronic heart failure. Circulation 2000;102:203–210.[Abstract/Free Full Text]
18. Laissy JP, Faraggi M, Lebtahi R, et al. Functional evaluation of normal and ischemic kidney by means of gadolinium-DOTA enhanced TurboFLASH MR imaging: a preliminary comparison with 99Tc-MAG3 dynamic scintigraphy. Magn Reson Imaging 1994;12:413–419.[CrossRef][Medline]
19. Rundback JH, Sacks D, Kent KC, et al. Guidelines for the reporting of renal artery revascularization in clinical trials. American Heart Association. Circulation 2002;106:1572–1585.
20. Hoffmann U, Brix G, Knopp MV, Hess T, Lorenz WJ. Pharmacokinetic mapping of the breast: a new method for dynamic MR mammography. Magn Reson Med 1995;33:506–514.[Medline]
21. Hawighorst H, Knapstein PG, Weikel W, et al. Cervical carcinoma: comparison of standard and pharmacokinetic MR imaging. Radiology 1996;201:531–539.[Abstract/Free Full Text]
22. Griswold MA, Jakob PM, Heidemann RM, et al. Generalized autocalibrating partially parallel acquisitions (GRAPPA). Magn Reson Med 2002;47:1202–1210.[CrossRef][Medline]
23. Schoenberg SO, Knopp MV, Prince MR, Londy F, Knopp MA. Arterial-phase three-dimensional gadolinium magnetic resonance angiography of the renal arteries: strategies for timing and contrast media injection—original investigation. Invest Radiol 1998;33:506–514.[CrossRef][Medline]
24. Thompson HK, Starmer CF, Whalen RE, McIntosh HD. Indicator transit time considered as a gamma variate. Circ Res 1964;14:502–515.[Free Full Text]
25. Romero JC, Lerman LO. Novel noninvasive techniques for studying renal function in man. Semin Nephrol 2000;20:456–462.[Medline]
26. Krier JD, Ritman EL, Bajzer Z, Romero JC, Lerman A, Lerman LO. Noninvasive measurement of concurrent single-kidney perfusion, glomerular filtration, and tubular function. Am J Physiol Renal Physiol 2001;281:F630–F638.[Abstract/Free Full Text]
27. Tofts PS. Modeling tracer kinetics in dynamic Gd-DTPA MR imaging. J Magn Reson Imaging 1997;7:91–101.[Medline]
28. Press WH, Teukolsky SA, Vetterling WT, Flannery BP. Modeling of data. In: Numerical recipes in C: the art of scientific computing. 2nd ed. Cambridge, England: Cambridge University Press, 1992; 521–529.
29. Schoenberg SO, Knopp MV, Londy F, et al. Morphologic and functional magnetic resonance imaging of renal artery stenosis: a multireader tricenter study. J Am Soc Nephrol 2002;13:158–169.[Abstract/Free Full Text]
30. Kirchheim HR, Ehmke H, Hackenthal E, Lowe W, Persson P. Autoregulation of renal blood flow, glomerular filtration rate and renin release in conscious dogs. Pflugers Arch 1987;410:441–449.[CrossRef][Medline]
31. Schoenberg SO, Just A, Bock M, Knopp MV, Persson PB, Kirchheim HR. Noninvasive analysis of renal artery blood flow dynamics with MR cine phase-contrast flow measurements. Am J Physiol 1997;272(5 pt 2):H2477–H2484.