Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online before print January 26, 2006, 10.1148/radiol.2382041623
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2382041623v1
238/3/1013    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Fenchel, M.
Right arrow Articles by Schick, F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Fenchel, M.
Right arrow Articles by Schick, F.
(Radiology 2006;238:1013-1021.)
© RSNA, 2006

Technical Developments

Perfusion MR Imaging with FAIR True FISP Spin Labeling in Patients with and without Renal Artery Stenosis: Initial Experience1

Michael Fenchel, MD, Petros Martirosian, PhD, Juergen Langanke, MD, Jenny Giersch, BSc, Stephan Miller, MD, Norbert I. Stauder, MD, Ulrich Kramer, MD, Claus D. Claussen, MD and Fritz Schick, MD, PhD

1 From the Department of Diagnostic Radiology (M.F., S.M., N.I.S., U.K., C.D.C.); Department of Diagnostic Radiology, Section of Experimental Radiology (P.M., F.S.); and Department of Internal Medicine, Division of Nephrology (J.L., J.G.), Eberhard-Karls-University Tuebingen, Hoppe-Seyler-Str 3, 72076 Tuebingen, Germany. From the 2003 and 2004 RSNA Annual Meetings. Received September 20, 2004; revision requested November 24; revision received February 9, 2005; accepted March 3; final version accepted May 3. Address correspondence to M.F. (e-mail: michael.fenchel{at}med.uni-tuebingen.de).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
The purpose of this study was to prospectively evaluate an arterial spin-labeling technique, flow-sensitive alternating inversion-recovery (FAIR) true fast imaging with steady-state precession (FISP), for noninvasive quantification of renal perfusion in patients without a history of renal artery stenosis (RAS) and in patients with proved RAS. The study was approved by the local ethics committee, and all participants provided written informed consent. Six patients with hypertension but no history of renal artery disease and 12 patients with RAS underwent FAIR true FISP magnetic resonance (MR) imaging in a whole-body 1.5-T unit. RAS grade and scintigraphic perfusion data served as the reference standards. On the FAIR true FISP perfusion images, severe RAS (>70% luminal narrowing) could be clearly distinguished from no or mild RAS and moderate RAS (≤70% luminal narrowing) (P < .005). Significant correlations between FAIR perfusion data and stenosis grade (r = –0.76) and between FAIR and single photon emission computed tomographic perfusion values (r = 0.83) were observed. FAIR true FISP was found to be suitable for quantitative perfusion imaging of the kidneys in patients with RAS.

© RSNA, 2006


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
Quantitative changes in the tissue perfusion rate can yield relevant information about the function of organs in abnormal conditions. In the kidneys, functional impairment occurs in association with a broad range of systemic diseases, including atherosclerosis, hypertension, diabetes, and autoimmune disease. In addition, spatially resolved perfusion data are of particular interest for postoperative monitoring of patients with kidney transplants.

Functional imaging modalities have been used for assessment of renal artery stenosis (RAS) (1–3), evaluation of renal masses (4), and early detection of transplant rejection (5,6). Many of the currently applied functional renal imaging methods are based on the introduction of exogenous radioactive tracer compounds (1,7). However, there are potential risks associated with the use of such tracers, including anaphylaxis and radiation exposure. For this reason, a completely noninvasive, repeatable and easily applied method to detect and quantify renal perfusion alterations with high spatial resolution imaging of the kidneys is desirable (8,9).

Tissue perfusion can be studied by using arterial spin-labeling (ASL) techniques, with which arterial presaturation pulses are applied to invert or saturate the magnetization of arterial blood flowing into a recorded section. Subsequently, the experiment is repeated without labeling the arterial blood (10). Image subtraction yields a map of inflowing spins with a signal intensity proportional to the level of tissue perfusion. This principle allows quantification of regional perfusion in tissue, which is expressed in milliliters per 100 g per minute. In several previous studies, ASL techniques have been applied, mainly to brain perfusion, by using an echo-planar (11–14) or spoiled gradient-echo (15,16) readout. However, echo-planar techniques are susceptible to static field inhomogeneity and signal loss in tissue with short T2. Use of the recently introduced flow-sensitive alternating inversion-recovery (FAIR) true fast imaging with steady-state precession (FISP) perfusion technique (17,18) eliminates both susceptibility artifacts and image distortions, which generally appear on echo-planar images, especially those of organs with inherent field inhomogeneities. The FAIR true FISP sequence was shown to be suitable for quantification of kidney perfusion in healthy volunteers (18), yielding high signal-to-noise ratios and excellent image quality.

Examinations of kidney perfusion with conventional radiographic techniques or magnetic resonance (MR) imaging require the administration of contrast material, which might be nephrotoxic, especially in patients with chronic renal insufficiency (19,20). Accordingly, patients with impaired renal function in particular would benefit from imaging examinations that do not require contrast agents. Therefore, the purpose of our study was to prospectively evaluate the ASL technique FAIR true FISP for the noninvasive quantification of renal perfusion in volunteer patients without a history of RAS and in patients with proved RAS.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 References
 
Patients
Patients (outpatients) were recruited from our nephrology department. Between September 2002 and November 2003, renal MR imaging was performed in 12 consecutive patients (seven men, five women; mean age, 60.8 years ± 9.1 [standard deviation]; age range, 46–71 years) with proved RAS and in six consecutive patients (four men, two women; mean age, 54.2 years ± 13.4; age range, 29–63 years) with arterial hypertension of unknown origin but no RAS who were scheduled to undergo MR imaging to rule out RAS. Patients with contraindications to MR imaging (ie, pacemaker and/or claustrophobia) were excluded from the study. The study was approved by the local ethics committee, and all participants provided informed written consent.

FAIR True FISP Sequence
The FAIR ASL technique (13,14) was combined with a true FISP (21) readout pulse sequence and implemented on a 1.5-T clinical MR unit (Magnetom Sonata; Siemens, Erlangen, Germany) that operated with a maximum gradient strength of 40 mT/m and a slew rate of 200 mT/m/msec along all three axes. The body coil was used for radiofrequency transmission, whereas combined spine- and body-array coils from the manufacturer were applied for signal detection.

The FAIR true FISP approach, as well as the technical and theoretic backgrounds of the sequence, has been previously described in detail (18). In brief, the FAIR true FISP sequence parameters used were 4.6/2.3 (repetition time msec/echo time msec), a 70° flip angle, a 128 x 128 matrix, a 340–360-mm field of view, and a 651 Hz/pixel bandwidth. Centric reordered phase-encoding data acquisition was applied to achieve high sensitivity to perfusion (22). The effective inversion time (TI) to the k-space center was 1000–1200 msec. This TI was chosen, according to the protocol of Kwong et al (14), because the maximum perfusion curve is expected at 1000–1200 msec, when the TI is equal to the T1. Given a TI of 1200 msec, the total time for the acquisition of one image was approximately 1.8 seconds.

For all examinations, the section thickness of the images was 8 mm and the slab thickness for section-selective inversion in the FAIR preparation was 20 mm. One section was acquired per measurement. Breath-hold examinations, which yielded two images with selective and nonselective inversion, and normal-respiration examinations, which yielded 36 images with selective and nonselective inversion, were performed in all subjects. Breath-hold images were acquired within 18 seconds, and non–breath-hold images were acquired within approximately 4 minutes.

ASL Perfusion Examinations
FAIR true FISP perfusion images were tilted in a coronal orientation to match the longitudinal axis of the kidneys. Care was taken to not include the renal arteries or the descending aorta above the level of the renal arteries in the region of the selective inversion.

In addition, the T1 was measured by using an inversion-recovery preparation with variable TIs (100–10 000 msec) that were applied before the true FISP sequence was performed. A set of 13 inversion-recovery images were acquired within 2 minutes by using a multiple-breath-hold technique and the same coronal orientation. After these images were registered, estimates of the T1 were obtained by using a nonlinear least square fit of the signal intensity measured for each TI. Pixel-by-pixel calculations yielded T1 relaxation maps for subsequent image analysis. For normalization, additional images were acquired by using a true FISP acquisition without inversion for assessment of density-related signal intensities.

Quantitative Analysis of MR Perfusion Data
A quantitative model for analysis of renal blood flow with use of FAIR techniques has been previously described by using the extended Bloch equations (8). Use of this model enables one to simulate the difference in longitudinal magnetization, {Delta}M(TI), between FAIR experiments with global preparation and those with section-selective preparation (13,14):

Formula 1(1)
where Msel is the longitudinal magnetization with section-selective preparation and Mnon is the longitudinal magnetization without section-selective preparation. M0 represents the tissue equilibrium magnetization per unit mass of tissue, T1 is the longitudinal relaxation time of tissue, f is the perfusion rate (usually expressed in milliliters per 100 g per minute), and {lambda} is the blood-tissue water partition coefficient, which is thought to be nearly constant at 80 mL/100 g (9,23). Perfusion maps can be calculated pixel-by-pixel by analyzing the difference in magnetization ({Delta}M) at a given TI, the undisturbed magnetization (ie, M0), and the T1 by using the following equation:

Formula 2(2)
In contrast to the signal intensity difference measured on echo-planar images, the signal intensity difference measured on true FISP images does not exactly reflect the longitudinal magnetization after the end of the FAIR preparation (18). To minimize this effect, centric reordered phase encoding was applied in the FAIR true FISP sequence to generate reliable quantitative perfusion maps.

MR Perfusion Image Processing
Magnitude images with section-selective and global inversion were collected in two sets. The magnitude images in each set were averaged, and the final images were subtracted at a stand-alone personal computer by using Matlab software (MathWorks, Natick, Mass). Images acquired at non–breath-hold examinations were registered by using rigid transformation before data evaluation. The final perfusion-weighted images were interpolated to a 256 x 256 matrix. In all subjects, the MR data were analyzed and postprocessed by the same radiologist (M.F., with 3 years experience in MR perfusion techniques). Care was exercised to ensure that after registration there would be optimal matching of the renal contours on all images obtained at the non–breath-hold examinations. Rigid registration was performed separately for the right and left kidneys.

Pixels with non–physiologically high perfusion of more than 250 mL/(100 g · min–1) in the renal medulla or more than 600 mL/(100 g · min–1) in the renal cortex were excluded from the evaluation. We constructed T1 relaxation maps pixel by pixel from inversion-recovery images with variable TIs by fitting the signal intensity values to a monoexponential curve with use of the Matlab program. These T1 maps were used solely for tissue classification: Pixels were assigned to different tissue categories (ie, macroscopic vessels, renal medulla, or renal cortex) according to their value on the relaxation maps. The renal medulla was separated from the renal cortex by using a value between the peak of the T1 data for the medulla and the peak of the T1 data for the cortex, which was between 1200 and 1300 msec. The T1 segmentation results were reviewed by an observer (M.F.) for anatomic correctness. Perfusion values were determined separately for each tissue category. The T1 values for the cortex and the medulla were set at 966 and 1412 msec, respectively, according to values reported in the literature (24). The quantitative perfusion images were constructed on a pixel-by-pixel basis by using Equation (2). Color-encoded perfusion maps were constructed for perfusion values ranging from 0 to 500 mL/(100 g · min–1). Relative perfusion values for the right and left kidneys were calculated according to the following equation:

Formula 3(3)
where prel is the relative kidney perfusion, pabs is the absolute kidney perfusion, pabs/l is the absolute perfusion in the left kidney, and pabs/r is the absolute perfusion in the right kidney.

Perfusion Scintigraphy
After the patients were hydrated (with 10 mL of saline per kilogram of body weight), we performed baseline scintigraphy by administering 100 MBq of technetium Tc 99m mertiatide (TechneScan MAG3; Mallinckrodt, Hazelwood, Mo), with an effective dose of 1.2 mSv. A 20-minute dynamic acquisition was simultaneously started by using a 64 x 64 matrix. Twenty images were obtained in 6 seconds, and then 36 images were obtained in 30 seconds. Two regions of interest that were automatically drawn around the kidneys enabled automatic identification of two perirenal background ROIs. After background subtraction, a temporal renal activity curve was plotted. Effective renal plasma flow was calculated separately for each kidney according to the following formula: Fren = CTc-mer/0.55 x (1 – H) (25), where Fren is the renal plasma flow (in milliliters per minute), CTc-mer is the clearance of technetium Tc 99m mertiatide (in milliliters per minute) from the kidney, and H is the hematocrit level.

Stenosis Grading
Conventional (x-ray) angiography, computed tomographic angiography, or MR angiography of the renal arteries was performed according to the clinical indication 1–12 weeks before the MR examination. No treatments for RAS were performed between the angiographic and MR examinations.

The angiographic images were assessed by consensus between two experienced radiologists (N.I.S. and U.K., both with 3 years experience acquiring and interpreting angiographic images). After selecting the projection that showed maximal RAS severity, we measured the luminal diameters of the stenotic arterial segment and the adjacent segments without luminal narrowing. The severity of the stenosis was expressed as a percentage reduction in internal diameter in relation to the estimated diameter interpolated from the diameters at the proximal and distal boundaries of the stenosis. RAS grade was expressed as a stenosis percentage as well as by using a three-grade scale: Grade A indicated no or mild RAS, or 0%–30% luminal narrowing; grade B indicated moderate stenosis, or 31%–70% luminal narrowing; and grade C indicated severe stenosis, or 71%–100% luminal narrowing. Grade A and grade B stenoses were not considered to be hemodynamically significant.

Statistical Analyses
The perfusion data acquired at breath-hold and non–breath-hold examinations of the entire kidney and of the cortex and the medulla were tested for normal distribution by using the Kolmogorov-Smirnov test. Subsequently, perfusion values acquired at breath-hold and non–breath-hold examinations were compared by using the paired two-tailed Student t test. Perfusion values for the medulla and the cortex also were compared by using the paired two-tailed Student t test.

The Kruskal-Wallis test was used to determine whether the renal perfusion values grouped according to RAS grade (no or mild, moderate, or severe RAS) differed significantly. After determining that differences existed among the groups, we used the Bonferroni multiple-comparison post hoc test to determine where the differences occurred.

Quantitative perfusion values were correlated with stenosis grade (ie, stenosis percentage) and scintigraphic perfusion data by using Pearson correlation coefficients. All statistical evaluations were performed by using JMP, version 4 (SAS Institute, Cary, NC), statistical software. P ≤ .05 was considered to indicate a significant difference.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 References
 
All patients were examined by using the described protocol, including the breath-hold techniques, without any critical problems. The kidneys were outlined clearly on all images, and reliable perfusion values could be calculated. In all patients, FAIR true FISP perfusion images were free of artifacts and were of diagnostic quality. The perfusion images showed higher perfusion in the cortex than in the medulla and enabled reliable differentiation between the two compartments. Inspection of the T1 segmentations by the observer yielded anatomically correct results in all cases.

Patients without RAS
The mean kidney perfusion values for the patients without RAS were similar between the breath-hold and non–breath-hold examinations (Table 1). Specifically, no significant differences in total kidney (P = .84), medulla (P = .06), or cortex (P = .44) perfusion values were evident between the two examinations. Conversely, perfusion in the medulla was consistently different from perfusion in the cortex at both breath-hold and non–breath-hold imaging (P < .001) (Table 1). The difference in perfusion values between the left and right kidneys at breath-hold and non–breath-hold FAIR perfusion imaging was less than 5% in the patients without RAS (Table 2).


View this table:
[in this window]
[in a new window]

  		Table 1. Medulla, Cortex, and Total Kidney Perfusion Values at FAIR True FISP MR Imaging in Six Patients without RAS

 

View this table:
[in this window]
[in a new window]

  		Table 2. Quantitative MR Perfusion Values

 
Patients with RAS
Detailed information about the perfusion values for the left and right kidneys at breath-hold and non–breath-hold imaging in the patients with RAS is given in Table 2. The percentages of RAS derived from the angiographic examinations, as well as the relative perfusion values derived from the SPECT examinations, also are given.

In contrast to the perfusion values measured in the patients with unaffected renal arteries, significant side-based differences in perfusion were obvious in the patients with RAS (Table 3). Furthermore, the Kruskal-Wallis test revealed significant differences (P < .005) in perfusion among the patients with no or mild, moderate, or severe RAS. At post hoc testing, no significant differences between the patients with no or mild RAS and those with moderate RAS were found. However, absolute perfusion values were significantly different between the kidneys with high-grade (>70%) stenotic arteries and the kidneys supplied by renal arteries with no or low-grade (≤70%) stenosis (P < .005) (Fig 1).


View this table:
[in this window]
[in a new window]

  		Table 3. Absolute and Relative Perfusion Values at FAIR True FISP MR and SPECT Imaging in 12 Patients with RAS

 

Figure 1
View larger version (24K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1: Box-Whiskers plot of renal perfusion grouped according to RAS grade. For each perfusion group, perfusion values are shown ({square}, {blacksquare}, bullet). Central box represents values from lower to upper quartile (25%–75% percentile). Middle line represents median. Vertical line extends from minimal to maximal value. Significant differences were found between the grade A and grade C stenosis groups, as well as between the grade B and grade C stenosis groups (P < .005 for both comparisons). Grade A represents no or mild (0%–30%) RAS; grade B, moderate (31%–70%) RAS; and grade C, severe (71%–100%) RAS.

 
A comparison of the side-based differences in perfusion between the FAIR and SPECT examinations revealed a high correlation (r = 0.83; P < .001; 95% confidence interval: 0.63, 0.93). Furthermore, a moderate correlation between FAIR perfusion values and angiographic stenosis grades was found (r = –0.76; P < .001; 95% confidence interval: –0.51, –0.89).

Patients without significant RAS had comparable perfusion values between the two kidneys (Fig 2). In the patients with significant (>70%) RAS, decreased perfusion values were measured in the kidney supplied by the severely stenotic renal arteries (Fig 3).


Figure 2
View larger version (135K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2a: MR images acquired in 71-year-old man with non–hemodynamically significant (50%) left RAS. Total perfusion values for left and right kidneys are 210 and 213 mL/(100 g · min–1), respectively. (a) Tissue equilibrium magnetization image (4.6/2.3) obtained in tilted coronal orientation by using true FISP without an inversion pulse. (b) T1 map and (c) FAIR true FISP image averaged from 36 data acquisitions after section-selective inversion. (d) FAIR true FISP image of perfusion in both kidneys shows differences in perfusion between cortex and medulla. (e) FAIR perfusion and (f) tissue classification maps of both kidneys, generated by assigning each pixel to tissue categories according to the T1 value of the pixel. Yellow pixels = cortex, orange pixels = medulla, brown pixels = macroscopic blood vessels, blue pixels = no convergence of T1 curve fitting. In d and e, color bar represents perfusion, in milliliters per 100 g per minute.

 

Figure 2
View larger version (167K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2b: MR images acquired in 71-year-old man with non–hemodynamically significant (50%) left RAS. Total perfusion values for left and right kidneys are 210 and 213 mL/(100 g · min–1), respectively. (a) Tissue equilibrium magnetization image (4.6/2.3) obtained in tilted coronal orientation by using true FISP without an inversion pulse. (b) T1 map and (c) FAIR true FISP image averaged from 36 data acquisitions after section-selective inversion. (d) FAIR true FISP image of perfusion in both kidneys shows differences in perfusion between cortex and medulla. (e) FAIR perfusion and (f) tissue classification maps of both kidneys, generated by assigning each pixel to tissue categories according to the T1 value of the pixel. Yellow pixels = cortex, orange pixels = medulla, brown pixels = macroscopic blood vessels, blue pixels = no convergence of T1 curve fitting. In d and e, color bar represents perfusion, in milliliters per 100 g per minute.

 

Figure 2
View larger version (150K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2c: MR images acquired in 71-year-old man with non–hemodynamically significant (50%) left RAS. Total perfusion values for left and right kidneys are 210 and 213 mL/(100 g · min–1), respectively. (a) Tissue equilibrium magnetization image (4.6/2.3) obtained in tilted coronal orientation by using true FISP without an inversion pulse. (b) T1 map and (c) FAIR true FISP image averaged from 36 data acquisitions after section-selective inversion. (d) FAIR true FISP image of perfusion in both kidneys shows differences in perfusion between cortex and medulla. (e) FAIR perfusion and (f) tissue classification maps of both kidneys, generated by assigning each pixel to tissue categories according to the T1 value of the pixel. Yellow pixels = cortex, orange pixels = medulla, brown pixels = macroscopic blood vessels, blue pixels = no convergence of T1 curve fitting. In d and e, color bar represents perfusion, in milliliters per 100 g per minute.

 

Figure 2
View larger version (111K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2d: MR images acquired in 71-year-old man with non–hemodynamically significant (50%) left RAS. Total perfusion values for left and right kidneys are 210 and 213 mL/(100 g · min–1), respectively. (a) Tissue equilibrium magnetization image (4.6/2.3) obtained in tilted coronal orientation by using true FISP without an inversion pulse. (b) T1 map and (c) FAIR true FISP image averaged from 36 data acquisitions after section-selective inversion. (d) FAIR true FISP image of perfusion in both kidneys shows differences in perfusion between cortex and medulla. (e) FAIR perfusion and (f) tissue classification maps of both kidneys, generated by assigning each pixel to tissue categories according to the T1 value of the pixel. Yellow pixels = cortex, orange pixels = medulla, brown pixels = macroscopic blood vessels, blue pixels = no convergence of T1 curve fitting. In d and e, color bar represents perfusion, in milliliters per 100 g per minute.

 

Figure 2
View larger version (77K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2e: MR images acquired in 71-year-old man with non–hemodynamically significant (50%) left RAS. Total perfusion values for left and right kidneys are 210 and 213 mL/(100 g · min–1), respectively. (a) Tissue equilibrium magnetization image (4.6/2.3) obtained in tilted coronal orientation by using true FISP without an inversion pulse. (b) T1 map and (c) FAIR true FISP image averaged from 36 data acquisitions after section-selective inversion. (d) FAIR true FISP image of perfusion in both kidneys shows differences in perfusion between cortex and medulla. (e) FAIR perfusion and (f) tissue classification maps of both kidneys, generated by assigning each pixel to tissue categories according to the T1 value of the pixel. Yellow pixels = cortex, orange pixels = medulla, brown pixels = macroscopic blood vessels, blue pixels = no convergence of T1 curve fitting. In d and e, color bar represents perfusion, in milliliters per 100 g per minute.

 

Figure 2
View larger version (53K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2f: MR images acquired in 71-year-old man with non–hemodynamically significant (50%) left RAS. Total perfusion values for left and right kidneys are 210 and 213 mL/(100 g · min–1), respectively. (a) Tissue equilibrium magnetization image (4.6/2.3) obtained in tilted coronal orientation by using true FISP without an inversion pulse. (b) T1 map and (c) FAIR true FISP image averaged from 36 data acquisitions after section-selective inversion. (d) FAIR true FISP image of perfusion in both kidneys shows differences in perfusion between cortex and medulla. (e) FAIR perfusion and (f) tissue classification maps of both kidneys, generated by assigning each pixel to tissue categories according to the T1 value of the pixel. Yellow pixels = cortex, orange pixels = medulla, brown pixels = macroscopic blood vessels, blue pixels = no convergence of T1 curve fitting. In d and e, color bar represents perfusion, in milliliters per 100 g per minute.

 

Figure 3
View larger version (149K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3a: Mild (30%) left RAS and severe (90%) right RAS in 70-year-old man. (a) Tissue equilibrium magnetization image (4.6/2.3) obtained in tilted coronal angulation by using true FISP without inversion pulse. (b) T1 map and (c) FAIR true FISP image averaged from 36 data acquisitions after section-selective inversion. (d) Perfusion-weighted map shows differences in perfusion between right (191 mL/[100 g · min–1]) and left (270 mL/[100 g · min–1]) kidneys. Color bar represents perfusion, in milliliters per 100 g per minute.

 

Figure 3
View larger version (115K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3b: Mild (30%) left RAS and severe (90%) right RAS in 70-year-old man. (a) Tissue equilibrium magnetization image (4.6/2.3) obtained in tilted coronal angulation by using true FISP without inversion pulse. (b) T1 map and (c) FAIR true FISP image averaged from 36 data acquisitions after section-selective inversion. (d) Perfusion-weighted map shows differences in perfusion between right (191 mL/[100 g · min–1]) and left (270 mL/[100 g · min–1]) kidneys. Color bar represents perfusion, in milliliters per 100 g per minute.

 

Figure 3
View larger version (111K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3c: Mild (30%) left RAS and severe (90%) right RAS in 70-year-old man. (a) Tissue equilibrium magnetization image (4.6/2.3) obtained in tilted coronal angulation by using true FISP without inversion pulse. (b) T1 map and (c) FAIR true FISP image averaged from 36 data acquisitions after section-selective inversion. (d) Perfusion-weighted map shows differences in perfusion between right (191 mL/[100 g · min–1]) and left (270 mL/[100 g · min–1]) kidneys. Color bar represents perfusion, in milliliters per 100 g per minute.

 

Figure 3
View larger version (107K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3d: Mild (30%) left RAS and severe (90%) right RAS in 70-year-old man. (a) Tissue equilibrium magnetization image (4.6/2.3) obtained in tilted coronal angulation by using true FISP without inversion pulse. (b) T1 map and (c) FAIR true FISP image averaged from 36 data acquisitions after section-selective inversion. (d) Perfusion-weighted map shows differences in perfusion between right (191 mL/[100 g · min–1]) and left (270 mL/[100 g · min–1]) kidneys. Color bar represents perfusion, in milliliters per 100 g per minute.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
References
 
Our study results show that an ASL MR technique, FAIR true FISP, enables reliable quantitative and spatially resolved examination of renal perfusion without administration of contrast material. The quantitative renal perfusion data derived at FAIR true FISP had excellent correlation with scintigraphic data and reasonable correlation with angiographic RAS grade.

True FISP
ASL perfusion techniques used in combination with echo-planar data recording are sensitive to magnetic susceptibility artifacts. Consequently, examinations of body regions with inherent magnetic field inhomogeneities are critical when echo-planar imaging is used. The half-Fourier rapid acquisition with relaxation enhancement sequence enables perfusion imaging without susceptibility artifacts (26,27). However, long acquisition times may result in blurring artifacts. Fast low-angle shot data acquisition with repeated small-angle radiofrequency excitation (8,15,16) has been shown to produce undesired saturation effects on the magnetization at perfusion imaging. Moreover, the signal-to-noise ratio per measurement time with fast low-angle shot imaging is lower than that with other imaging strategies.

The true FISP sequence is not associated with the described disadvantages but rather yields a high signal-to-noise ratio per measurement time (18). Martirosian et al (18) reported higher signal-to-noise ratios and relative signal intensity changes on true FISP renal perfusion images compared with corresponding parameters on echo-planar images, and an even higher spatial resolution was achieved with true FISP in the phase-encoding direction. Furthermore, unlike the findings seen on FAIR echo-planar images, no signal losses in the medulla or geometric image distortions were present on the FAIR true FISP images.

Quantification of Renal Perfusion
The cortex perfusion data acquired with FAIR true FISP in the current study were well within the values reported in the literature (9,28,29), but some discrepancy between the medulla perfusion data acquired in our study and the physiologic data reported in two studies (9,29) can be observed. It should be considered, however, that quantification of renal perfusion might be susceptible to errors due to partial volume effects (voxel size, 2.8 x 2.8 x 8.0 mm), which may increase the estimated perfusion rate in voxels of the medulla, secondary to signal contributions from the cortex or adjacent macroscopic blood vessels. In particular, the contamination of adjacent voxels from macroscopic blood vessels can easily lead to higher perfusion values because macroscopic blood vessels have 10–50-fold higher perfusion values. Although pixels with non–physiologically high perfusion in the medulla and the cortex were excluded from our evaluation, the increased perfusion values in the medulla may have been secondary to partial volume errors. Roberts et al (8) reported mean perfusion rates of 278 mL/(100 g · min–1) ± 55 (standard deviation) in the cortex and 55 mL/(100 g · min–1) ± 25 in the medulla in a series of seven volunteers. However, in their study, signals from the intravascular compartment were successfully attenuated by applying a bipolar gradient pulse after excitation to reduce the effects of macroscopic blood vessels on the perfusion measurements.

T1 Measurements
Correct interactive segmentation of the medulla and the cortex is not trivial using only perfusion-weighted images. Owing to differences in T1 between the medulla and the cortex, these tissues can be differentiated by using T1 maps. Therefore, a T1 map was calculated for each patient so that the perfusion value of each pixel could be assigned to the medulla or the cortex and macroscopic vessels and perirenal tissue could be excluded from the perfusion calculations. This approach is an objective means of differentiating between the cortex and the medulla, and, in our opinion, it is superior to interactive region-of-interest placement and leads to more standardized image postprocessing with reduced user interaction. The T1 map values calculated in our study correlated reasonably well with the T1 data reported in the literature. However, we chose to use the T1 values described in the literature (24) to calculate perfusion values. Noise on T1 maps, as well as blurring caused by minimal residual misregistration and partial volume effects, might introduce additional error or increase the noise from perfusion data.

Mean T1 values for the total kidney, medulla, and cortex were similar between the normally perfused and hypoperfused kidneys. Therefore, the segmentation and calculation of perfusion data were not influenced by the RAS grade.

Clinical Implications
It is commonly acknowledged in the literature that less than 70% RAS can be sufficiently compensated and therefore does not lead to reduced kidney perfusion (30–32). Results of the present study confirm this statement because kidney perfusion was not significantly reduced in the patients with less than 70% RAS. However, the perfusion measurements in the patients who had one main renal artery with luminal narrowing that exceeded 70% were significantly different between the two kidneys.

Our study results suggest that the noninvasive quantification of renal perfusion with FAIR true FISP in patients with RAS is feasible. Use of this sequence can help to differentiate patients with no, mild, or moderate RAS from those with severe RAS. This is important because the extent of renovascular disease needs to be accurately determined, especially in patients with moderate stenosis, in whom intervention may be the most beneficial (33). Serial investigations have been performed to assess the severity of perfusion impairment in patients with moderate stenosis by using a contrast material bolus (34,35). Despite the benign pharmacologic properties of the gadolinium chelates used for contrast material–enhanced renal perfusion measurements, patients with impaired renal function in particular may benefit from examinations performed by using the described FAIR true FISP approach.

Although ASL sequences still need to be tested more rigorously, previous work has already shown the value of ASL with use of a semiquantitative extrasection spin-tagging technique in the monitoring of organ perfusion in transplanted rat kidneys (5) and in the detection of kidneys supplied by stenotic renal arteries (27). In the latter study, Berr et al (27) reported a coefficient of 0.76 for the correlation between the stenosis degree measured with conventional angiography and that measured with ASL perfusion MR imaging. Berr et al concluded that semiquantitative extrasection spin-tagging techniques may be useful for routine screening of patients for renal vascular disease and for assessment of angiographically equivocal RAS (27).

Limitations and Future Aspects
One limitation of the study was the relatively small number of patients who were included. Despite this limitation, however, several significant results were obtained. Studies with more patients are needed to confirm these data. Compared with techniques based on the administration of contrast material, ASL methods yield lower contrast effects per measurement time. Like all ASL techniques, the FAIR approach is used to measure only the perfusion component perpendicular to the imaging section. Depending on the orientation of the imaging section in relation to the main direction of perfusion, different results may be obtained. Owing to the course of the renal arteries, coronal or sagittal section orientations are preferred for FAIR perfusion imaging.

Another limitation of the true FISP method is that unlike with echo-planar ASL approaches, with true FISP ASL techniques, there are signal contributions from macroscopic blood vessels. These contributions cannot be dephased by using conventional bipolar crusher gradients because such gradients lead to an undesired prolongation of the repetition time (18). Although it is possible to eliminate the perfusion values from macroscopic blood vessels with image postprocessing, it is preferable to reduce the contributions from macroscopic blood vessels at the time of image acquisition, with resultant images that depict capillary perfusion only. This could be achieved with suitable preparation of fast flow, as recently suggested by Pell et al (36).

In conclusion, noninvasive quantification of renal perfusion without contrast material administration seems to be a promising technique for the screening and follow-up of patients known or suspected to have renal artery disease or impaired renal function. Because we observed good correlation between ASL perfusion values and reference examination values and because ASL techniques do not require the use of contrast agents, the described true FISP ASL technique could be valuable for extended applications in routine investigations of RAS.


   FOOTNOTES
 

Abbreviations: ASL = arterial spin labeling • FAIR = flow-sensitive alternating inversion recovery • FISP = fast imaging with steady-state precession • RAS = renal artery stenosis • TI = inversion time

Authors stated no financial relationship to disclose.

Author contributions: Guarantors of integrity of entire study, M.F., P.M., J.L., C.D.C., F.S.; 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, M.F., P.M., J.G., S.M., N.I.S., U.K.; clinical studies, all authors; statistical analysis, M.F., P.M., J.G., S.M., N.I.S., U.K.; and manuscript editing, all authors


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 

  1. Mitchell DG, Tobin M, Le Veen R, et al. Induced renal artery stenosis in rabbits: magnetic resonance imaging, angiography, and radionuclide determination of blood volume and blood flow. Magn Reson Imaging 1988;6:113–124.[CrossRef][Medline]
  2. Gruenewald SM, Collins LT, Antico VF, Farlow DC, Fawdry RM. Can quantitative renography predict the outcome of treatment of atherosclerotic renal artery stenosis? J Nucl Med 1989;30:1946–1954.[Abstract/Free Full Text]
  3. Powers TA, Lorenz CH, Holburn GE, Price RR. Renal artery stenosis: in vivo perfusion MR imaging. Radiology 1991;178:543–548.[Abstract/Free Full Text]
  4. Semelka RC, Shoenut JP, Kroeker MA, MacMahon RG, Greenberg HM. Renal lesions: controlled comparison between CT and 1.5-T MR imaging with nonenhanced and gadolinium-enhanced fat suppressed spin-echo and breath-hold FLASH techniques. Radiology 1992;182:425–430.[Abstract/Free Full Text]
  5. Wang JJ, Hendrich KS, Jackson EK, Ildstad ST, Williams DS, Ho C. Perfusion quantitation in transplanted rat kidney by MRI with arterial spin labeling. Kidney International 1998;53:1783–1791.[CrossRef][Medline]
  6. Hanna S, Helenon O, Legendre C, et al. MR imaging of renal transplant rejection. Acta Radiol 1991;32:42–46.[Medline]
  7. Kubota K, Atkins HL, Anaise D, Oster ZH, Pollack W. Quantitative evaluation of renal excretion on the dynamic DTPA renal scan. Clin Nucl Med 1989;14:8–12.[CrossRef][Medline]
  8. Roberts DA, Detre JA, Bolinger L, et al. Renal perfusion in humans: MR imaging with spin tagging of arterial water. Radiology 1995;196:281–286.[Abstract/Free Full Text]
  9. Karger N, Biederer J, Lüsse S, et al. Quantitation of renal perfusion using arterial spin labeling with FAIR-UFLARE. Magn Reson Imaging 2000;18:641–647.[CrossRef][Medline]
  10. Detre JA, Leigh JS, Williams DS, Koretsky AP. Perfusion imaging. Magn Reson Med 1992;23:37–45.[Medline]
  11. Calamante F, Williams SR, van Bruggen N, Kwong KK, Turner R. A model for quantification of perfusion in pulsed labelling techniques. NMR Biomed 1996;9(2):79–83. [Published correction appears in NMR Biomed 1996;9(6):277.][CrossRef][Medline]
  12. Edelman RR, Siewert B, Darby DG, et al. Qualitative mapping of cerebral blood flow and functional localization with echo-planar MR imaging and signal targeting with alternating radio frequency. Radiology 1994;192:513–520.[Abstract/Free Full Text]
  13. Kim SG. Quantification of relative cerebral blood flow change by flow-sensitive alternating inversion recovery (FAIR) technique: application to functional mapping. Magn Reson Med 1995;34:293–301.[Medline]
  14. Kwong KK, Chesler DA, Weisskoff DM, et al. MR perfusion studies with T1-weighted echo planar imaging. Magn Reson Med 1995;34:878–887.[Medline]
  15. Schwarzbauer C, Morrissey SP, Haase A. Quantitative magnetic resonance imaging of perfusion using magnetic labeling of water proton spins within the detection slice. Magn Reson Med 1996;35:540–546.[Medline]
  16. Tsekos NV, Zhang F, Merkle H, Nagayama M, Iadecola C, Kim SG. Quantitative measurements of cerebral blood flow in rats using the FAIR technique: correlation with previous iodoantipyrine autoradiographic studies. Magn Reson Med 1998;39:564–573.[Medline]
  17. Martirosian P, Schick F, Klose U. True-FISP sequences applied for data recording in FAIR perfusion imaging (abstr). In: Proceedings of the Ninth Meeting of the International Society for Magnetic Resonance in Medicine. Berkeley, Calif: International Society for Magnetic Resonance in Medicine, 2001; 1560.
  18. Martirosian P, Klose U, Mader I, Schick F. FAIR true-FISP perfusion imaging of the kidneys. Magn Reson Med 2004;51:353–361.[CrossRef][Medline]
  19. Briguori C, Tavano D, Colombo A. Contrast agent-associated nephrotoxicity. Prog Cardiovasc Dis 2003;45:493–503.[CrossRef][Medline]
  20. Sam AD, Morasch MD, Collins J, Song G, Chen R, Pereles FS. Safety of gadolinium contrast angiography in patients with chronic renal insufficiency. J Vasc Surg 2003;38:313–318.[CrossRef][Medline]
  21. Oppelt A, Graumann R, Barfuss H, Fischer H, Hartl W, Schajor W. FISP: a new fast MRI sequence. Electromedica 1986;54:15–18.
  22. Holsinger AE, Riederer SJ. The importance of phase-encoding order in ultra-short TR snapshot MR imaging. Magn Reson Med 1990;16:481–488.[Medline]
  23. Bjornerud A, Johansson LO, Briley-Saebo K, Ahlstrom HK. Assessment of T1 and T2* effects in vivo and ex vivo using iron oxide nanoparticles in steady state: dependence on blood volume and water exchange. Magn Reson Med 2002;47:461–471.[CrossRef][Medline]
  24. de Bazelaire CM, Duhamel GD, Rofsky NM, Alsop DC. MR imaging relaxation times of abdominal and pelvic tissues measured in vivo at 3.0 T: preliminary results. Radiology 2004;230:652–659.[Abstract/Free Full Text]
  25. Bubeck B, Brandau W, Weber E, Kälble T, Parekh N, Georgi P. Pharmacokinetics of technetium-99m-MAG3 in humans. J Nucl Med 1990;31:1285–1293.[Abstract/Free Full Text]
  26. Chen Q, Siewert B, Warach S, Edelman RR. STAR-HASTE: perfusion imaging without magnetic susceptibility artifact. Magn Reson Med 1997;38:404–408.[Medline]
  27. Berr SS, Hagspiel KD, Mai VM, et al. Perfusion of the kidney using extraslice spin tagging (EST) magnetic resonance imaging. J Magn Reson Imaging 1999;10:886–891.[CrossRef][Medline]
  28. Lorenz CH, Powers TA, Partain CL. Quantitative imaging of renal blood flow and function. Invest Radiol 1992;27(suppl 2):S109–S114.[Medline]
  29. Dworkin LD, Brenner BM. The renal circulation. In: Benner BM, Rector FC, eds. The kidney. Philadelphia, Pa: Saunders, 1981.
  30. 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]
  31. Davidson RA, Wilcox CS. Newer tests for the diagnosis of renovascular disease. JAMA 1992;268:3353–3358.[Abstract/Free Full Text]
  32. Haimovici H, Zinicola N. Experimental renal-artery stenosis diagnostic significance of arterial hemodynamics. J Cardiovasc Surg (Torino) 1962;3:259–262.[Medline]
  33. Martin LG, Rundback JH, Sacks D, et al. Quality improvement guidelines for angiography, angioplasty, and stent placement in the diagnosis and treatment of renal artery stenosis in adults. J Vasc Interv Radiol 2002;13:1069–1083.[Medline]
  34. Prasad PV, Cannillo J, Chavez DR, et al. First-pass renal perfusion imaging using MS-325, an albumin-targeted MRI contrast agent. Invest Radiol 1999;34:566–571.[CrossRef][Medline]
  35. Miller S, Schick F, Duda SH, et al. Gd-enhanced 3D phase-contrast MR angiography and dynamic perfusion imaging in the diagnosis of renal artery stenosis. Magn Reson Imaging 1998;16:1005–1012.[CrossRef][Medline]
  36. Pell GS, Lewis DP, Branch CA. Pulsed arterial spin labeling using TurboFLASH with suppression of intravascular signal. Magn Reson Med 2003;49:341–350.[CrossRef][Medline]




This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2382041623v1
238/3/1013    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Fenchel, M.
Right arrow Articles by Schick, F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Fenchel, M.
Right arrow Articles by Schick, F.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE