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


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
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 Michaely, H. J.
Right arrow Articles by Schoenberg, S. O.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Michaely, H. J.
Right arrow Articles by Schoenberg, S. O.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?
DOI:10.2214/AJR.06.0567
AJR 2007; 188:919-926
© American Roentgen Ray Society

Original Research

Semiquantitative Assessment of First-Pass Renal Perfusion at 1.5 T: Comparison of 2D Saturation Recovery Sequences With and Without Parallel Imaging

Henrik J. Michaely1, Harald Kramer1, Niels Oesingmann2, Klaus-Peter Lodemann3, Maximilian F. Reiser1 and Stefan O. Schoenberg1

1 Department of Clinical Radiology, University of Munich, Grosshadern-Campus, Marchionistrasse 15, Munich, Germany, 81377.
2 Siemens Medical Solutions, Malvern, PA.
3 Bracco-Altana Pharma, Konstanz, Germany.

Received April 26, 2006; accepted after revision August 31, 2006.

 
Address correspondence to H. J. Michaely (henrik.michaely{at}med.uni-muenchen.de).

Sponsored through a grant from Bracco-Altana Pharma, Konstanz, Germany.

The employment of N. Oesingmann at Siemens Medical Solutions and K. P. Lodemann at Bracco-Altana Pharma did not influence the data in this study.


Abstract
Top
 Abstract
Introduction
 Subjects and Methods
 Results
 Discussion
 References
 
OBJECTIVE. The purpose of this study was to assess the feasibility and reliability of measurements performed with true fast imaging with steady-state free precession (FISP) and turbo fast low-angle shot (FLASH) sequences with parallel imaging compared with those obtained with turbo FLASH sequences without parallel imaging in first-pass renal perfusion MRI.

SUBJECTS AND METHODS. The subjects in this prospective study were 15 healthy men who volunteered to undergo MRI for acquisition of renal perfusion measurements. Imaging was performed at 1.5 T with the following three techniques after administration of gadobenate dimeglumine at 4 mL/s: saturation recovery (SR) turbo FLASH sequences without parallel imaging, SR turbo FLASH sequences with parallel imaging, and SR true FISP sequences. The spatial resolution was 2.3 x 2.6 x 8 mm with a temporal resolution of four slices per second (turbo FLASH without parallel imaging and true FISP) or six slices per second (turbo FLASH with parallel imaging). The semiquantitative perfusion parameters mean transit time and maximal upslope were determined. Signal-to-noise ratio (SNR), delta ratio, and time to maximal signal intensity also were determined. Image quality was rated in consensus.

RESULTS. Image quality was best for turbo FLASH sequences without parallel imaging compared with true FISP and turbo FLASH sequences with parallel imaging. True FISP sequences yielded the highest baseline SNR (26.7) but the lowest delta ratio (3.2). Turbo FLASH sequences without and with parallel imaging had significantly lower SNRs (9.6 and 9.3) and significantly higher delta ratios (5.1 and 5.0). The first-pass perfusion parameters mean transit time and time to maximal signal intensity were independent of the technique used.

CONCLUSION. It seems that at 1.5 T, turbo FLASH sequences without parallel imaging are the best approach to renal first-pass perfusion imaging.

Keywords: genitourinary tract imaging • kidney • MRI technique • perfusion-weighted MRI • renal perfusion


Introduction
Top
 Abstract
 Introduction
Subjects and Methods
 Results
 Discussion
 References
 
Despite technical advances, diagnosis of renovascular hypertension with MR angiography remains a challenge [1]. Therefore, additional dynamic, first-pass imaging with gadolinium enhancement is being used increasingly to assess organ perfusion in patients with suspected renal artery stenosis [2]. The clinical importance of perfusion measurements is emphasized by the fact that reduced renal perfusion is associated with increased morbidity and mortality [3, 4]. In previous studies, various imaging techniques, such as multiphasic T1-weighted sequences [2, 5, 6] and volume-inter-polated breath-hold examination sequences [7], have been used. Both approaches are based on spoiled gradient-echo sequences. Fast low-angle shot (FLASH) sequences, particularly turbo FLASH sequences, which are characterized by an additional magnetization preparation, are known to show a high degree of linearity between signal intensity and gadolinium concentration [8, 9]. These sequences are robust, fast, and not prone to specific absorption rate (SAR) restrictions; they do, however, have a moderate signal-to-noise-ratio (SNR) [10].

Compared with renal perfusion imaging, cardiac perfusion imaging has become established as a standard clinical tool [11]. In cardiac imaging, fast imaging with steady-state free precession (FISP) sequences, such as true FISP, have become standard because they have a higher inherent SNR than turbo FLASH sequences [12, 13]. Parallel imaging, introduced to perfusion imaging to improve temporal resolution and to increase the number of slices simultaneously acquired [14], is considered standard technique for cardiac perfusion imaging.

The aim of this intraindividual comparison study was twofold. The first goal was to compare the feasibility and reliability of true FISP renal first-pass perfusion imaging with those of a turbo FLASH sequence, the standard of reference. The second aim was to assess the feasibility and value of parallel imaging in measurement of renal perfusion.


Subjects and Methods
Top
 Abstract
 Introduction
 Subjects and Methods
Results
 Discussion
 References
 
Study Concept
The aim of this prospective, intraindividual comparison study was to investigate the best sequence for renal perfusion measurements. Fifteen healthy male volunteers underwent renal perfusion measurements on three different days with a different imaging sequence on each day. The sequences were saturation recovery (SR) turbo FLASH without parallel imaging, SR true FISP, and SR turbo FLASH with parallel imaging. Because, to our knowledge, all published data on renal perfusion have been obtained with fast spoiled gradient-echo sequences, the turbo FLASH sequence without parallel imaging was considered the standard of reference against which a true FISP sequence without parallel imaging and a turbo FLASH sequence with a parallel imaging acceleration factor of 2 were tested.

Subjects
After internal review board approval was granted, informed consent was obtained from 15 healthy men who volunteered for the study. The subjects were 25-38 years old (mean age, 29.1 ± 3.1 years) and had no history of renal or vascular disease. Because of internal review board restrictions, only men were eligible for this study. All volunteers were well hydrated at the time of the examinations. Vital signs were checked before imaging and 10 minutes after the volunteers left the MRI machine. A 20-gauge needle was placed in an antecubital vein of all volunteers for administration of the contrast agent. For each examination, a blood sample was drawn from each volunteer and analyzed in the hospital laboratory to measure serum creatinine level to rule out previously unknown renal disease.

MRI
Measurements were performed on a 32-channel 1.5-T whole-body MRI unit (Magnetom Avanto, 76 x 32, Siemens Medical Solutions) with a gradient strength of 45 mT/m and slew rate of 200 mT/m/ms. For signal reception, one body matrix with six independent receiver elements and six elements of the spine matrix were used—that is, a total of 12 independent receiver elements. Before perfusion imaging, true FISP sequences in the coronal and transverse orientations were acquired for proper positioning. The following three sequences for measurement of renal perfusion were used on the three examination days, which were separated by at least 1 day and an average of 3 days: day 1, SR turbo FLASH sequence without parallel imaging and with MR angiography; day 2, SR true FISP; day 3, SR turbo FLASH with parallel imaging.

The parameters for all sequences are shown in Table 1. The slices of the perfusion measurement sequences were positioned in an oblique coronal orientation over the kidneys with a single slice in an axial orientation through the kidneys.


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

  		TABLE 1: Sequence Parameters for Perfusion MRI

 

All sequences were performed with the SR technique, which grants complete and homogeneous saturation over the entire selected imaging volume and thereby minimizes inflow-related artifacts. For magnetization preparation, a pulse train of three short {pi}/2 pulses with constant amplitude and phase cycling in a phase angle of {pi}/2 was applied. The pulse train was followed by a 10-mT/m spoiler over 1 millisecond. More than three pulses did not significantly improve the saturation effect but lengthened the preparation phase. The time between the preparation pulses and acquisition of the central image line (inversion time) is given in Table 1.

For the turbo FLASH sequences without parallel imaging, a linear relation between contrast medium concentration and signal intensity for a wide range of concentrations up to 1.1 mmol/L is well known [8]. A standardized contrast bolus of 7 mL of gadobenate dimeglumine (MultiHance, Bracco Pharma) was injected into an antecubital vein at a flow rate of 4 mL/s for the perfusion measurements and was followed by 30 mL of saline solution. Owing to weak protein binding, gadobenate dimeglumine has higher relaxivity than conventional gadolinium chelates [15] and therefore yields stronger enhancement.

For two reasons, contrast agent injection was started simultaneously with the start of the imaging sequences. First, magnetization is driven into the steady state before arrival of contrast agent in the kidney; second, unenhanced images are needed for successful postprocessing. The volunteers were instructed to hold their breath as long as possible to allow motion-free monitoring of the first pass of the contrast agent.

A phantom containing eight vials (2 x 11 cm) of albumin-gadobenate dimeglumine solution was placed under the volunteers. The vials had the following increasing concentrations of gadobenate dimeglumine: 0.03125, 0.125, 0.25, 0.5, 1, 4, 8, and 16 mmol/L. The phantom was tempered to 37°C. The axial slice of the perfusion sequence was positioned to include the phantom (Fig. 1A, 1B, 1C).


Figure 1
View larger version (98K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 1A —28-year-old man in good health. Axial slice of turbo fast low-angle shot (FLASH) perfusion sequence shows kidneys and phantom.

 

Figure 2
View larger version (11K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 1B —28-year-old man in good health. Graph shows measured mean peak signal intensities in aorta and both kidneys. True fast imaging with steady-state free precession (FISP) sequence yields highest absolute signal from enhanced kidneys and enhanced aorta. Turbo FLASH sequences with and without parallel imaging (PI) perform in similar way. AU = arbitrary units.

 

Figure 3
View larger version (13K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 1C —28-year-old man in good health. Graph shows calibration curve with measured signal intensities from first five vials (0.032-1.0 mmol/L of gadobenate dimeglumine). Peak signal intensity measured in aorta and kidneys (B) are well within increasing part of calibration curve. T2*-related loss of signal intensity can therefore safely be ruled out. Solid line indicates regression for true FISP sequence; dashed line, regression for turbo FLASH sequence; dotted line, regression for turbo FLASH sequence with parallel imaging.

 
During the first examination, contrast-enhanced MR angiography was performed to rule out renal artery stenosis. To allow the contrast agent to leave the vessel bed, the examination was stopped for at least 5 minutes after initial contrast administration. A test bolus technique was used to determine circulation time. A full 3D MR angiographic data set (spatial resolution, 1.0 x 0.8 x 1.0 mm; generalized autocalibrating partially parallel acquisition [16]; parallel imaging acceleration factor, 3; acquisition time, 19 seconds) was acquired. For MR angiography, a standardized 20-mL bolus of gadobenate dimeglumine was injected into an antecubital vein at a flow rate of 2 mL/s and was followed by a 30-mL saline chaser administered at 2 mL/s.

Perfusion Data Analysis
MR perfusion measurements were analyzed on an offline workstation with MERZ software (version 0.99, Siemens Medical Solutions) [2]. According to this software, region of interest (ROI)-based evaluation of contrast enhancement on a pixel-by-pixel basis yields a signal intensity-time curve. The sizes of the ROIs were chosen to include at least 400 pixels. The ROIs were positioned over the renal cortex, sparing the surrounding fat and renal medulla. Figure 2 shows positioning in the ROI. The ROIs were automatically corrected for motion. This motion correction algorithm is based on a next-neighbor approach.


Figure 4
View larger version (86K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 2 —27-year-old man in good health. Coronal MR image shows sample of placement of region of interest over both renal cortices.

 
MERZ operates with a Marquardt-Levenberg least squares [17] algorithm for automatic acquisition of a gamma variate fit [18] of the initial signal intensity-time curve. If desired, it is possible to interact with the fitting process by confining the time intervals and parameters for finding reasonable start values. Goodness of fit is presented after successful fitting. Examples of signal intensity-time curves and the derived fitted curves for all three techniques are shown in Figure 3A, 3B. The gamma variate fit founded on the indicator dilution theory describes the first-pass component of tracer kinetics and has been used for description of renal blood flow in imaging techniques such as CT and electron beam tomography [19, 20].


Figure 5
View larger version (10K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 3A —25-year-old man in good health. Signal intensity-time curves for three MRI sequences. Dotted lines indicate true fast imaging with steady-state free precession (FISP), solid gray lines indicate turbo fast low-angle shot (FLASH) without parallel imaging, and solid black lines indicate turbo FLASH with parallel imaging. Graph shows marked differences in configuration of first pass peak. AU = arbitrary units.

 

Figure 6
View larger version (8K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 3B —25-year-old man in good health. Signal intensity-time curves for three MRI sequences. Dotted lines indicate true fast imaging with steady-state free precession (FISP), solid gray lines indicate turbo fast low-angle shot (FLASH) without parallel imaging, and solid black lines indicate turbo FLASH with parallel imaging. Gamma variate fit of curves in A. True FISP sequence yielded lowest maximal signal intensity and slowest upslope of curve. Mean transit time and time to peak were equal for all techniques. Scale of x-axis is slightly different from that in A.

 

The four following characteristic curve parameters can be derived from the gamma variate fit: time to maximal signal intensity, maximal upslope (ds/dt), maximal signal intensity (baseline corrected), and first moment of function that corresponds to the mean transit time of the tracer. To compensate for intraobserver variability, the mean values of two ROIs were used for further analysis.

SNR Measurements
The signal intensity of the eight vials of the phantom was measured for each volunteer and plotted against concentration. Because dynamic imaging always represents a problem for objective SNR measurements, a previously described [21] difference method was used for all sequences. This method eliminates the problem of inhomogeneous distribution of noise. Two baseline images obtained before arrival of contrast agent in the kidney were subtracted. The mean SD of an ROI drawn in the region of the kidneys was used as image noise. To avoid saturation-related artifacts, the first four images were not used for image noise calculation. Signal intensity for the baseline SNR calculation was measured as the mean value of an ROI over the unenhanced kidneys and the aorta. For peak SNR calculation, ROIs were placed over the aorta and kidneys at the point of maximal enhancement, and mean signal intensity again was measured. Baseline and peak SNRs were calculated as follows:

Formula

To compare the relative enhancement between baseline SNR and peak SNR, the delta ratio was computed for the kidney and aorta as follows:

Formula

Visual Assessment
The image quality of the perfusion sequences was rated by two radiologists in consensus. Overall image quality was based on the presence and degree of artifacts and on demarcation of the kidney from the surrounding tissue. An ordinal scale with three grades was used: 1, no artifacts, kidneys easily differentiated from the surrounding tissues; 2, visible artifacts, kidneys easily differentiated from the surrounding tissues or no artifacts but kidneys difficult to differentiate from surrounding tissues; 3, visible artifacts and kidneys difficult to differentiate from surrounding tissues.

Statistical Analysis
All values are mean and SD. Results of the Kolmogorov-Smirnov test proved that the data followed gaussian distribution. For that reason, Student's t tests were used for further comparison. Without correction there would be a 46.0% chance of finding one or more significant differences in the 12 tests performed. Therefore, Bonferroni correction [22] for multiple comparisons was performed and resulted in a significance level of p < 0.0042. For comparison of perfusion parameters, the SR turbo FLASH sequence was considered the standard of reference. The median was used to describe the visual assessment data. All statistical analyses were performed with SPSS version 12.0.1 software (SPSS, Inc.).


Results
Top
 Abstract
 Introduction
 Subjects and Methods
 Results
Discussion
 References
 
The perfusion measurements were obtained without problems in all volunteers. No adverse effects occurred. The additional MR angiography did not reveal renal artery disease. Serum creatinine levels were within normal limits for all volunteers on all examination days (mean serum creatinine concentration, 1.1 ± 0.1 mg/dL).

Measurement of peak signal intensity revealed a peak in the kidneys of 56.7 ± 26.4 arbitrary units (AU) for the turbo FLASH sequence with parallel imaging, 60.1 ± 25.0 AU for turbo FLASH without parallel imaging, and 109.2 ± 20.5 AU for true FISP. In the aorta, peak signal intensity was 104.0 ± 17.8 AU for turbo FLASH with parallel imaging, 94.0 ± 19.9 AU for turbo FLASH without parallel imaging, and 189.9 ± 49.5 AU for true FISP. As Figure 1A, 1B, 1C shows, these values were within the rising part of the calibration curve. T2*-related alteration of measured signal intensity therefore was ruled out. A similar result was found for SNR, which was significantly (p < 0.0001) higher for true FISP in unenhanced kidney, enhanced kidney, and the aorta (Fig. 4A, 4B). In contrast, the delta ratio, which describes relative enhancement normalized to baseline SNR, was significantly lower for the true FISP (3.2 ± 0.8) sequence compared with the turbo FLASH sequences with (5.1 ± 1.1) and without (5.0 ± 2.2) parallel imaging (Fig. 3A, 3B).


Figure 7
View larger version (12K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 4A —Comparison of sequences. Graph shows comparison of measured signal-to-noise ratio (SNR) at baseline and peak enhancement in renal cortex and abdominal aorta. True fast imaging with steady-state free precession (FISP) sequence yields highest SNR. Turbo fast low-angle shot (FLASH) sequences with and without parallel imaging (PI) behave similarly to each other.

 

Figure 8
View larger version (8K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 4B —Comparison of sequences. Graph shows delta ratio comparisons for three sequences. True FISP yielded highest SNR at baseline and during peak enhancement. Relative enhancement—that is, delta ratio—however, was only 3.2 for true FISP. Turbo FLASH sequences had 59% higher delta ratios of 5.1 with and 5.0 without parallel imaging.

 
The exact values of the semiquantitative perfusion parameters with mean and SD are shown in Table 2. Almost equal values for mean transit time of approximately 14.8 seconds and time to maximal signal intensity of approximately 11.1 seconds were found for all three sequences. For maximal signal intensity, no significant difference was found between the turbo FLASH sequences with and without parallel imaging. The maximal upslope of the signal intensity-time curve also showed no significant differences between these two sequences. In contrast, the true FISP sequence had statistically significantly (p < 0.00002) lower maximal signal intensity and maximal upslope than the turbo FLASH sequences with and without parallel imaging.


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

  		TABLE 2: Overview of Results of Semiquantitative Perfusion Parameters Measured with Three Sequences

 

Examples of image quality are shown in Figure 5A, 5B, 5C, 5D, 5E, 5F, 5G, 5H, 5I. Visual assessment of image quality yielded the highest rating for the turbo FLASH sequence without parallel imaging, a median of 1. Both true FISP and turbo FLASH with parallel imaging were given a median rating of 2. True FISP was rated lower because it yielded lower relative signal enhancement after contrast administration, weak contrast between the kidneys and perirenal fat tissue, susceptibility artifacts, and banding artifacts. The turbo FLASH sequence with parallel imaging suffered from increased, perceivable image noise and reconstruction artifacts, which became especially visible after first-pass perfusion (Fig. 6A, 6B, 6C, 6D). The kidneys became very difficult to differentiate with turbo FLASH sequences without parallel imaging in the late phase of perfusion, in which most of the contrast agent had left the vascular system.


Figure 9
View larger version (89K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 5A —29-year-old man in good health. Unenhanced (A), early arterial phase (B), and medullary phase (C) MR images obtained with turbo fast low-angle shot (FLASH) sequence.

 

Figure 10
View larger version (93K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 5B —29-year-old man in good health. Unenhanced (A), early arterial phase (B), and medullary phase (C) MR images obtained with turbo fast low-angle shot (FLASH) sequence.

 

Figure 11
View larger version (88K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 5C —29-year-old man in good health. Unenhanced (A), early arterial phase (B), and medullary phase (C) MR images obtained with turbo fast low-angle shot (FLASH) sequence.

 

Figure 12
View larger version (92K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 5D —29-year-old man in good health. Unenhanced (D), early arterial phase (E), and medullary phase (F) MR images obtained with turbo FLASH sequence with parallel imaging show higher noise level than A-C.

 

Figure 13
View larger version (94K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 5E —29-year-old man in good health. Unenhanced (D), early arterial phase (E), and medullary phase (F) MR images obtained with turbo FLASH sequence with parallel imaging show higher noise level than A-C.

 

Figure 14
View larger version (89K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 5F —29-year-old man in good health. Unenhanced (D), early arterial phase (E), and medullary phase (F) MR images obtained with turbo FLASH sequence with parallel imaging show higher noise level than A-C.

 

Figure 15
View larger version (85K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 5G —29-year-old man in good health. Unenhanced (G), early arterial phase (H), and medullary phase (I) MR images obtained with true fast imaging with steady-state free precession sequence yield best signal-to-noise ratio, but because of higher background signal intensity, kidneys are difficult to differentiate.

 

Figure 16
View larger version (87K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 5H —29-year-old man in good health. Unenhanced (G), early arterial phase (H), and medullary phase (I) MR images obtained with true fast imaging with steady-state free precession sequence yield best signal-to-noise ratio, but because of higher background signal intensity, kidneys are difficult to differentiate.

 

Figure 17
View larger version (86K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 5I —29-year-old man in good health. Unenhanced (G), early arterial phase (H), and medullary phase (I) MR images obtained with true fast imaging with steady-state free precession sequence yield best signal-to-noise ratio, but because of higher background signal intensity, kidneys are difficult to differentiate.

 

Figure 18
View larger version (92K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 6A —29-year-old man in good health. MR image obtained with single true fast imaging with steady-state free precession sequence shows susceptibility artifacts that occur when kidney is close to air-filled large bowel. Margins of renal cortex (arrows) are not clearly defined on either side.

 

Figure 19
View larger version (82K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 6B —29-year-old man in good health. Medullary phase (D) MR images obtained with turbo fast low-angle shot sequence with parallel imaging show typical bandlike reconstruction artifact (arrows). Artifacts were seen especially in late phases of perfusion measurement after intravascular concentration of contrast agent had markedly decreased.

 

Figure 20
View larger version (82K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 6C —29-year-old man in good health. Medullary phase (D) MR images obtained with turbo fast low-angle shot sequence with parallel imaging show typical bandlike reconstruction artifact (arrows). Artifacts were seen especially in late phases of perfusion measurement after intravascular concentration of contrast agent had markedly decreased.

 

Figure 21
View larger version (88K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 6D —29-year-old man in good health. Medullary phase (D) MR images obtained with turbo fast low-angle shot sequence with parallel imaging show typical bandlike reconstruction artifact (arrows). Artifacts were seen especially in late phases of perfusion measurement after intravascular concentration of contrast agent had markedly decreased.

 


Discussion
Top
 Abstract
 Introduction
 Subjects and Methods
 Results
 Discussion
References
 
The main focus of this study was to evaluate the value of true FISP and turbo FLASH sequences in first-pass renal perfusion at 1.5 T. The results study shows that for assessment of first-pass renal perfusion, turbo FLASH without parallel imaging appears to be a better sequence than true FISP and turbo FLASH with parallel imaging. The parameters' mean transit time and time to maximal signal intensity were found equal for all techniques used. In another study [23], these parameters were also found equal for measurements with a time-resolved angiographic technique with high temporal resolution at 3.0 T. In contrast, the maximum upslope and maximal signal intensity were significantly different between the turbo FLASH sequences and the true FISP sequence. With equal time to maximum signal intensity, higher maximum signal intensity automatically leads to increased maximum upslope.

The results may seem confusing in light of the SNR values, the highest measured signal intensity and SNR values being found for the true FISP sequence. The solution to this paradox lies in the postprocessing program MERZ, which performs a baseline correction during the fitting and yields signal intensity values normalized to the unenhanced images. A closer look at the delta ratio confirms these conclusions. The delta ratio of the kidneys, the relative enhancement rate, was 38% lower with true FISP than with turbo FLASH sequences. This result may be a consequence of the mixed contrast behavior of true FISP, which is determined by T1 and T2 weighting [24], whereas turbo FLASH sequences are heavily T1 weighted [25]. As a consequence, unenhanced true FISP images reveal higher baseline signal intensity in the kidney because of high renal water content. In contrast, true FISP has shown the greatest enhancement in myocardial perfusion studies [13]. This contradiction can be explained in the same way as for renal studies. The myocardium shows only little signal intensity without contrast material but is highly vascularized and shows strong enhancement.

In our study, the weaker peak arterial contrast enhancement was one reason the true FISP images were scored worse than the turbo FLASH images. In addition to weak relative enhancement, the true FISP images often had banding and susceptibility artifacts. Although they did not hinder assessment of renal perfusion in this study, these artifacts can be troublesome if air-filled large bowel is adjacent to the kidneys and if higher field strengths are used [26]. In addition, because of the high intrinsic SNR of true FISP sequences [27, 28] and the high fat content, differentiating enhanced kidney from perinephric fat was more difficult with true FISP than with turbo FLASH sequences.

The high signal intensity of the kidneys with true FISP can be explained by the mixed contrast behavior of steady-state free precession, which depends on T2 and T1 weighting. Therefore, the low T1 relaxation time of water does not necessarily lead to a lower SNR. Images obtained with the turbo FLASH sequence with parallel imaging were scored equally as low as true FISP images, but for another reason. Even though the SNR of turbo FLASH with parallel imaging was lower at baseline than the SNR of turbo FLASH without parallel imaging, the peak arterial enhancement for turbo FLASH with parallel imaging was the highest of all three sequences investigated. In the late phases, which are used to determine renal excretory function, it became very difficult to differentiate kidney from surrounding tissues with use of turbo FLASH with parallel imaging.

Use of parallel imaging produced artifacts, which were mainly seen after the main fraction of contrast material had left the vessel bed. These artifacts were particularly pronounced in the image noise and were present to a lesser extent in the spleen and kidneys. The use of newer parallel imaging reconstruction algorithms, such as temporal sensitivity encoding [29], may reduce these artifacts in the future. Temporal sensitivity encoding also may be more suitable than the generalized autocalibrating partially parallel acquisition algorithm for dynamic measurements because temporal sensitivity encoding obviates acquisition of reference lines and hence allows faster imaging with a higher temporal resolution [29]. Initial results in cardiac functional cine studies have proved acceleration factors of 4 to be feasible at 1.5 T [21].

Overall, our results imply that maximal upslope and maximal signal intensity seem to vary with the technique used and that mean transit time and time to maximal signal intensity seem to be independent of field strength. Maximal upslope has often been reported as a sensitive marker for perfusion deficits [6, 9]. However, in view of our results, care should be taken when maximal upslope values are compared.

This study was designed as a volunteer comparison study. Because the results are promising, patient studies should be undertaken to evaluate perfusion abnormalities in renal parenchymal and vascular diseases. In this context, determination of segmental contrast-to-noise changes, as in segmental renal transplant dysfunction or accessory renal artery stenosis, should be valuable in defining extent of disease.

In view of our results, future studies with turbo FLASH sequences are likely to profit from transition to 3.0 T in terms of increased image quality. A turbo FLASH sequence that spoils all remaining transverse magnetization after each readout is prone to neither increased susceptibility nor specific absorption rate limitations at 3.0 T. Higher field strength may allow use of parallel imaging with less SNR penalty and fewer artifacts than at 1.5 T. It also may be possible to use increased SNR at higher field strength to increase spatial resolution with less SNR penalty and less deterioration of image quality. In contrast, because of enhanced susceptibility and the problem of the high specific absorption rate from radiofrequency deposition with steady-state free precession sequences, such as true FISP, use of a true FISP sequence for renal perfusion measurements theoretically is less advantageous at 3.0 T.

A potential limitation of this study was that all volunteers were given a 7-mL bolus of contrast agent. A separate study of optimal contrast agent dose would be desirable. Assessment of contrast dose was not one of the aims of our investigation.

In conclusion, this study showed that first-pass renal perfusion measurements with a true FISP sequence and a turbo FLASH sequence with parallel imaging are feasible and have equal results for mean transit time and time to maximal signal intensity. Maximal upslope and maximal signal intensity were significantly different between the turbo FLASH sequences with and without parallel imaging and the true FISP sequence. In contrast to myocardial perfusion measurements, renal perfusion measurements with true FISP sequences do not seem favorable because of the 38% lower delta ratio compared with that of turbo FLASH sequences. Use of parallel imaging allows coverage of more slices simultaneously but at the cost of image quality. Therefore, because of advantageous contrast behavior and the absence of artifacts, SR turbo FLASH without parallel imaging appears to be the most appropriate sequence for assessment of first-pass renal perfusion at 1.5 T.


References
Top
 Abstract
 Introduction
 Subjects and Methods
 Results
 Discussion
 References
 

  1. Vasbinder GB, Nelemans PJ, Kessels AG, et al. Accuracy of computed tomographic angiography and magnetic resonance angiography for diagnosing renal artery stenosis. Ann Intern Med2004; 141:674 -682[Abstract/Free Full Text]
  2. Michaely HJ, Schoenberg SO, Oesingmann N, et al. Renal artery stenosis: functional assessment with dynamic MR perfusion measurements—feasibility study. Radiology2006; 238:586 -596[Abstract/Free Full Text]
  3. 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]
  4. 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]
  5. Vallee JP, Lazeyras F, Khan HG, Terrier F. Absolute renal blood flow quantification by dynamic MRI and Gd-DTPA. Eur Radiol 2000; 10:1245 -1252[CrossRef][Medline]
  6. 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]
  7. 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]
  8. 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]
  9. Wintersperger BJ, Penzkofer HV, Knez A, Weber J, Reiser MF. Multislice MR perfusion imaging and regional myocardial function analysis: complimentary findings in chronic myocardial ischemia. Int J Card Imaging 1999; 15:425 -434[CrossRef][Medline]
  10. Perrin RL, Ivancevic MK, Kozerke S, Vallee JP. Comparative study of FAST gradient echo MRI sequences: phantom study. J Magn Reson Imaging 2004; 20:1030 -1038[CrossRef][Medline]
  11. Wagner A, Mahrholdt H, Sechtem U, Kim RJ, Judd RM. MR imaging of myocardial perfusion and viability. Magn Reson Imaging Clin N Am 2003; 11:49 -66[CrossRef][Medline]
  12. Barkhausen J, Ruehm SG, Goyen M, Buck T, Laub G, Debatin JF. MR evaluation of ventricular function: true fast imaging with steady-state precession versus fast low-angle shot cine MR imaging: feasibility study. Radiology 2001;219 : 264-269[Abstract/Free Full Text]
  13. Fenchel M, Helber U, Simonetti OP, et al. Multislice first-pass myocardial perfusion imaging: comparison of saturation recovery (SR)-trueFISP-two-dimensional (2D) and SR-turboFLASH-2D pulse sequences. J Magn Reson Imaging 2004;19 : 555-563[CrossRef][Medline]
  14. Kostler H, Sandstede JJ, Lipke C, Landschutz W, Beer M, Hahn D. Auto-SENSE perfusion imaging of the whole human heart. J Magn Reson Imaging 2003; 18:702 -708[CrossRef][Medline]
  15. Rohrer M, Bauer H, Mintorovitch J, Requardt M, Weinmann HJ. Comparison of magnetic properties of MRI contrast media solutions at different magnetic field strengths. Invest Radiol2005; 40:715 -724[CrossRef][Medline]
  16. Griswold MA, Jakob PM, Heidemann RM, et al. Generalized autocalibrating partially parallel acquisitions (GRAPPA). Magn Reson Med 2002; 47:1202 -1210[CrossRef][Medline]
  17. Press WH, Teukolsky SA, Vetterling WT, Flannery BP. Modeling of data. In: Numerical recipes in C: the art of scientific computing. Cambridge, UK: Cambridge University Press,1992 : 521-529
  18. Thompson HK, Starmer CF, Whalen RE, McIntosh HD. Indicator transit time considered as a gamma variate. Circ Res1964; 14:502 -515[Free Full Text]
  19. Romero JC, Lerman LO. Novel noninvasive techniques for studying renal function in man. Semin Nephrol2000; 20:456 -462[Medline]
  20. 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]
  21. Reeder SB, Wintersperger BJ, Dietrich O, et al. Practical approaches to the evaluation of signal-to-noise ratio performance with parallel imaging: application with cardiac imaging and a 32-channel cardiac coil. Magn Reson Med 2005;54 : 748-754[CrossRef][Medline]
  22. Bland JM, Altman DG. Multiple significance tests: the Bonferroni method. BMJ 1995;310 : 1073[Free Full Text]
  23. Michaely HJ, Nael K, Schoenberg SO, et al. Renal perfusion: comparison of saturation-recovery turboFLASH measurements at 1.5T with saturation-recovery turboFLASH and time-resolved echo-shared angiographic technique (TREAT) at 3.0T. J Magn Reson Imaging2006; 24:1413 -1419[CrossRef][Medline]
  24. Huang TY, Huang IJ, Chen CY, Scheffler K, Chung HW, Cheng HC. Are trueFISP images T2/T1-weighted? Magn Reson Med2002; 48:684 -688[CrossRef][Medline]
  25. Chien D, Edelman RR. Ultrafast imaging using gradient echoes. Magn Reson Q 1991;7 : 31-56[Medline]
  26. Uematsu H, Takahashi M, Dougherty L, Hatabu H. High field body MR imaging: preliminary experiences. Clin Imaging2004; 28:159 -162[CrossRef][Medline]
  27. Reeder SB, Herzka DA, McVeigh ER. Signal-to-noise ratio behavior of steady-state free precession. Magn Reson Med2004; 52:123 -130[CrossRef][Medline]
  28. Hunold P, Maderwald S, Eggebrecht H, Vogt FM, Barkhausen J. Steady-state free precession sequences in myocardial first-pass perfusion MR imaging: comparison with turboFLASH imaging. Eur Radiol 2004; 14:409 -416[CrossRef][Medline]
  29. Kellman P, Derbyshire JA, Agyeman KO, McVeigh ER, Arai AE. Extended coverage first-pass perfusion imaging using slice-interleaved TSENSE. Magn Reson Med 2004;51 : 200-204[CrossRef][Medline]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?



This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
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 Michaely, H. J.
Right arrow Articles by Schoenberg, S. O.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Michaely, H. J.
Right arrow Articles by Schoenberg, S. O.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS