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


     

Published online before print April 10, 2008, 10.1148/radiol.2473071013
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2473071013v1
247/3/706    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 Andreisek, G.
Right arrow Articles by Nanz, D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Andreisek, G.
Right arrow Articles by Nanz, D.
(Radiology 2008;247:706-716.)
© RSNA, 2008

Experimental Studies

Direct MR Arthrography at 1.5 and 3.0 T: Signal Dependence on Gadolinium and Iodine Concentrations—Phantom Study1

Gustav Andreisek, MD, Johannes M. Froehlich, PhD, Juerg Hodler, MD, MBA, Dominik Weishaupt, MD, Verena Beutler, RT, Christian W. A. Pfirrmann, MD, Chris Boesch, MD, PhD, and Daniel Nanz, PhD

1 From the Institute for Diagnostic Radiology (G.A., D.W.) and Department of Medical Radiology (D.N.), University Hospital Zurich, Raemistrasse 100, 8091 Zurich, Switzerland; Guerbet, Zurich, Switzerland (J.M.F.); Department of Radiology, Orthopedic University Hospital Zurich, Zurich, Switzerland (J.H., C.W.A.P.); and Department of Clinical Research, University of Bern, Bern, Switzerland (V.B., C.B.). Received June 11, 2007; revision requested August 20; revision received October 3; final version accepted December 16. Address correspondence to G.A. (e-mail: gustav{at}andreisek.de).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 
Purpose: To prospectively quantify in vitro the influence of gadopentetate dimeglumine and ioversol on the magnetic resonance (MR) imaging signal observed with a variety of musculoskeletal pulse sequences to predict optimum gadolinium concentrations for direct MR arthrography at 1.5 and 3.0 T.

Materials and Methods: In an in vitro study, T1 and T2 relaxation times of three dilution series of gadopentetate dimeglumine (concentration, 0–20.0 mmol gadolinium per liter) at ioversol concentrations with iodine concentration of 0, 236.4, and 1182 mmol iodine per liter (corresponding to 0, 30, and 150 mg of iodine per milliliter) were measured at 1.5 and 3.0 T. The relaxation rate dependence on concentrations of gadolinium and iodine was analytically modeled, and continuous profiles of signal versus gadolinium concentration were calculated for 10 pulse sequences used in current musculoskeletal imaging. After fitting to experimental discrete profiles, maximum signal-to-noise ratio (SNR), gadolinium concentration with maximum SNR, and range of gadolinium concentration with 90% of maximum SNR were derived. The overall influence of field strength and iodine concentration on these parameters was assessed by using t tests. The deviation of simulated from experimental signal-response profiles was assessed with the autocorrelation of the residuals.

Results: The model reproduced relaxation rates of 0.37–38.24 sec–1, with a mean error of 4.5%. Calculated SNR profiles matched the discrete experimental profiles, with autocorrelation of the residuals divided by the mean of less than 5.0. Admixture of ioversol consistently reduced T1 and T2, narrowed optimum gadolinium concentration ranges (P = .004–.006), and reduced maximum SNR (P < .001 to not significant). Optimum gadolinium concentration was 0.7–3.4 mmol/L at both field strengths. At 3.0 T, maximum SNR was up to 75% higher than at 1.5 T.

Conclusion: Admixture of ioversol to gadopentetate dimeglumine solutions results in a consistent additional relaxation enhancement, which can be analytically modeled to allow a near-quantitative a priori optimized match of contrast media concentrations and imaging protocol for a broad variety of pulse sequences.

© RSNA, 2008


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 
Iodinated contrast agents alter gadolinium-related signal enhancement in standard spin-echo (SE), fast SE, and gradient-recalled echo (GRE) magnetic resonance (MR) imaging sequences. T1-, T2-, and intermediate-weighted SE or fast SE sequences with and without fat-signal suppression and GRE sequences are commonly used to diagnose joint disorders, with good accuracy (1–10). Current reports on the effect of iodinated contrast agents do not allow conclusions for imaging protocols that are only marginally changed (8,9,11,12). In addition, newer sequences, such as fat-signal–suppressing spoiled GRE (spoiled gradient-recalled acquisition in the steady state, or SPGR, fast low-angle shot [FLASH], and fast field-echo, or FFE), steady-state free-precession (true fast imaging with steady-state precession [FISP], fast imaging employing steady-state acquisition, or FIESTA, and balanced fast field-echo, or bFFE), multiple-echo GRE (dual-echo imaging in the steady state, multiecho data image combination [MEDIC], and multiple-echo recalled gradient-echo acquisition, or MERGE), and short–inversion time (short inversion time inversion-recovery, or STIR, and turbo inversion-recovery magnitude [TIRM]) sequences, have been increasingly used for joint imaging (13–17), with widely differing parameters (1,2,6,18–20). For these sequences, optimum gadolinium (eg, gadopentetate dimeglumine [Magnevist; Schering, Berlin, Germany]) concentrations and effects of iodine-based contrast agents (eg, ioversol [Optiray 350; Guerbet, Roissy, France]) injected for MR arthrography remain to be determined.

Thus, the purpose of our study was to prospectively quantify in vitro the influence of gadopentetate dimeglumine and ioversol on the MR imaging signal observed with a variety of musculoskeletal pulse sequences to predict optimum gadolinium concentrations for direct MR arthrography at 1.5 and 3.0 T.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 
Experimental Phantom
A phantom with three series of 11 sterile gadopentetate dimeglumine solutions in normal saline (0.9% sodium chloride) was prepared at gadolinium concentrations of 0, 0.05, 0.10, 0.25, 0.50, 0.75, 1.00, 2.00, 5.00, 10.0, and 20.0 mmol/L (10 000- to 25-fold dilution) (Fig 1) by one of the authors (J.M.F.). The three dilution series differed by their content of the iodinated contrast agent ioversol at iodine concentrations of 0 mmol/L (corresponding to 0 mg I/mL) (series 1), 236.4 mmol/L (corresponding to 30 mg I/mL) (series 2), and 1182 mmol/L (corresponding to 150 mg I/mL) (series 3).


Figure 1
View larger version (49K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1: Sketch of experimental phantom with three series of 11 sterile gadopentetate dimeglumine solutions in normal saline at ioversol concentrations of 0 mmol iodine per liter (0 mg I/mL) for series 1, 236.4 mmol iodine per liter (30 mg I/mL) for series 2, and 1182 mmol iodine per liter (150 mg I/mL) for series 3. Circles represent test tubes (plastic vials), and top and bottom numbers in each circle represent corresponding gadopentetate dimeglumine (millimoles per liter) and iodine (milligrams per milliliter) concentrations, respectively.

 
The solutions were kept in tightly sealed 7-mL plastic vials. Three additional vials contained normal saline, a fat specimen of wool grease (Haenseler, Herisau, Switzerland), and a muscle-tissue–mimicking preparation containing, per 100 mL of aqueous gel solution, 0.01 mmol of gadoterate meglumine (Dotarem; Guerbet), 0.385 mg of iron in the form of ferumoxsil (eg, corresponding to 2.2 mL ferumoxsil [GastroMark, Mallinckrodt, St Louis, Mo; or Lumirem, Guerbet]), and 3 g of hydroxyethylcellulose, and the resulting T1 and T2 relaxation times were 831.2 msec and 46.97 msec at 1.5 T, respectively. The 36 vials were immobilized in a plastic container.

Measurement of Relaxation Times
Relaxation times of the phantom solutions were measured with the whole-body transmit-receive coils of 1.5- and 3.0-T systems (Achieva; Philips Medical Systems, Best, the Netherlands) by one author (D.N.). Data were acquired at both field strengths on the same morning with the phantom placed in a water bath within a plastic Dewar container. The measured bath temperature was between 19.9°C and 20°C before and after each acquisition. The T1 determination was based on imaging with single-section inversion-recovery turbo SE MR imaging (repetition time msec/echo time msec/inversion time msec, 8000/15.6/50–5000; number of inversion-recovery times, 14; fast SE factor, seven; water-fat shift, 770 Hz/pixel; and nominally acquired voxel size, 1.0 x 1.0 x 10.0 mm). The T2 determination was based on single-section turbo SE MR imaging (repetition time, 6000 msec; four acquisitions with echo time values at multiples of 10, 15, 25, and 40 msec; number of echoes, eight; fast SE factor, eight; refocusing control on [180° pulses]; water-fat shift, 1066 Hz/pixel; and nominally acquired voxel size, 1.2 x 1.2 x 10.0 mm). T1 and T2 were estimated from three-parameter monoexponential fits of signal-to-noise ratio (SNR) versus inversion time and SNR versus echo time, respectively, by using standard interactive data language routines (IDL, version 6.3, 2006; ITT Visual Information Solutions, Boulder, Colo).

Modeling of Combined Effects of Gadopentetate Dimeglumine and Ioversol on Relaxation Rates
Test functions were fitted to experimental T1 and T2 values, with gadolinium and iodine concentrations as independent variables in search of analytic expressions that best reproduced the experimental results. Expressions were finally selected of the following form:

Formula
where Tx denotes T1 or T2; Txnat denotes T1 or T2 relaxation time of the saline solutions in the absence of any gadolinium- or iodine-based contrast agent; rx denotes r1 or r2 interpreted as the molar relaxivity of gadopentetate dimeglumine; cGd and cI are the concentrations of gadolinium and iodine, respectively; Ax represents scalar constant A1 or A2; and Bx represents scalar constant B1 or B2.

The coefficients of the Equation were optimized four times to model longitudinal and transverse relaxation rates at 1.5 and 3.0 T, respectively. The 21 T1 times at the gadolinium concentration range of 0.05–2.00 mmol/L and the 18 T2 times at the gadolinium concentration range of 0.10–2.00 mmol/L, which were assumed to be determined with highest accuracy, were included in the analysis. For these calculations, mathematic software (Mathematica, release 4.1.0; Wolfram Research, Champaign, Ill) was used.

In combination with standard pulse sequence–specific signal equations (18,21–24), the Equation allowed the calculation of continuous signal profiles as a function of gadolinium concentration at the three iodine concentrations for 10 clinical imaging sequences.

MR Imaging at 1.5 and 3.0 T
The phantom was imaged at a range of 19°–21°C with a 1.5-T MR unit (Avanto; Siemens Medical Solutions, Erlangen, Germany) (maximum slew rate, 200 mT/m/sec) by four authors (G.A., J.H., C.W.A.P., D.N.) and with a 3.0-T MR unit (Trio; Siemens Medical Solutions) and a maximum slew rate of 200 mT/m/sec (Fig 2) by four authors (G.A., V.B., C.B., D.N.) with 10 sequences (Table 1). Identical sequence parameters were targeted at both field strengths as far as allowed by the software of the MR unit. The phantom was placed in a plastic bowl flooded with a 0.5 mmol/L solution of gadopentetate dimeglumine in normal saline, which minimized susceptibility artifacts and allowed assessment and minimization of spatial signal inhomogeneities. Further attempts at minimizing such distortions included placement of the phantom in the isocenter of the MR unit, use of whole-body radiofrequency coils for radiofrequency irradiation and signal reception, and fourfold performance of all sequences with phantom orientations that differed by a rotation of 0°, 90°, 180°, and 270° around the central vertical short axis of the phantom.


Figure 2
View larger version (118K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2: TIRM MR image (4930/67/195) of phantom at 3.0 T. In each series, local SNR maximum (max) and SNR minimum (min) were observed. Minimum SNR was observed for test tubes with 0.75, 0.50, and 0.25 mmol/L gadopentetate dimeglumine for series 1, 2, and 3, respectively. Local SNR maxima were observed for test tubes with 2 mmol/L gadopentetate dimeglumine (series 1 and 2) and 0.75 mmol/L gadopentetate dimeglumine (series 3). Iodine concentration is in milligrams per milliliter.

 

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

  		Table 1. Pulse Sequences and Parameters in Evaluation of MR Signal Dependence on Gadopentetate Dimeglumine Concentration at 1.5 and 3.0 T

 
Experimental SNR Profiles for Clinical Sequences at 1.5 and 3.0 T
Standard software on the workstations (Syngo MR 2004V; Siemens Medical Solutions) of the MR unit was used to measure mean signal intensity (SI) values in regions of interest with an area of 46 mm2 positioned in the center of the test-vial images for each of the four phantom orientations, which were averaged for the evaluation. Four additional regions of interest of identical size were placed in the air outside the phantom to determine the average standard deviation (SD) of the background noise. All region of interest measurements were performed by the same author (G.A.). SNRs were calculated as SNR = SItsol/SDbn, where SItsol is the SI of the test solution and SDbn is the SD of background noise, to obtain 11 SNR values at different concentrations of gadolinium for each dilution series.

Theoretic SNR Profiles for Clinical Sequences at 1.5 and 3.0 T
For each dilution series and imaging sequence, the continuous analytic profiles were scaled to best fit the discrete experimental data by using least-squares optimization. Maximum SNR values and the gadolinium concentration yielding maximum SNR, as well as the concentration range yielding an SNR of 90% or higher of the maximum SNR, were then derived (Table 2). The global SNR maximum was taken as reference, except for the TIRM sequence. For the TIRM sequence, SNR profiles showed a decrease of near-zero magnitude (Fig 3), and the local maximum above this gadolinium concentration with minimum SNR served as reference. Gray-scale value representations of the relative gadopentetate dimeglumine–dependent signal variations were produced for each sequence, iodine concentration, and field strength (Figs 3, 4). Finally, the projected optimum gadolinium concentration and maximum SNR at varying iodine concentrations for T1-weighted SE MR imaging were evaluated (Fig 5). All of these evaluations were performed by one author (D.N.) with the mathematic software package mentioned previously.


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

  		Table 2. Gadolinium Concentrations and SNR Values for Pulse Sequences at Three Iodine Concentrations

 

Figure 3
View larger version (33K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3: Plot of experimental (bullet) and fitted (solid line) theoretical profiles of SNR (y-axis) at 1.5 and 3.0 T versus gadopentetate dimeglumine concentration (cGd, x-axis) obtained with TIRM sequences. Note logarithmic scaling of x-axis. Gray-scale value representation on top of each profile clarifies relationship between these two representations. Although white areas in gray-scale value representations encode SNR of 90% or more of the maximum SNR, gray-scale values in all other areas directly reflect ratio of SNR divided by maximum SNR of corresponding profile. Iodine concentration is in milligrams per milliliter. cI = iodine concentration, mM = millimoles per liter.

 

Figure 4A
View larger version (44K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4a: Gray-scale value representations of SNR versus gadopentetate dimeglumine concentration profiles obtained with different imaging sequences at three ioversol concentrations and at field strengths of (a) 1.5 T and (b) 3.0 T. Iodine concentration is in milligrams per milliliter. cGd = gadolinium concentration, cI = iodine concentration, DESS = dual-echo imaging in the steady state, fs = fat saturated, FSE = fast SE, mM = millimoles per liter.

 

Figure 4B
View larger version (44K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4b: Gray-scale value representations of SNR versus gadopentetate dimeglumine concentration profiles obtained with different imaging sequences at three ioversol concentrations and at field strengths of (a) 1.5 T and (b) 3.0 T. Iodine concentration is in milligrams per milliliter. cGd = gadolinium concentration, cI = iodine concentration, DESS = dual-echo imaging in the steady state, fs = fat saturated, FSE = fast SE, mM = millimoles per liter.

 

Figure 5A
View larger version (7K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5a: Projected effect of ioversol on T1-weighted SE imaging. (a) Gadopentetate dimeglumine concentration resulting in maximum SI, (b) as well as maximum SI reachable at a given iodine concentration (cI) decrease with increasing iodine concentration at 1.5 T (solid line) and 3.0 T (dashed line). Maximum SI on b is given in relation to global maximum SI (100%). cGd = gadolinium concentration.

 

Figure 5B
View larger version (6K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5b: Projected effect of ioversol on T1-weighted SE imaging. (a) Gadopentetate dimeglumine concentration resulting in maximum SI, (b) as well as maximum SI reachable at a given iodine concentration (cI) decrease with increasing iodine concentration at 1.5 T (solid line) and 3.0 T (dashed line). Maximum SI on b is given in relation to global maximum SI (100%). cGd = gadolinium concentration.

 
Statistical Analysis
The accuracy with which the Equation reproduced the experimental relaxation rates was assessed with the calculation of the mean and SD of the absolute relative residuals, given as follows: 100 · |Rxfit – Rxexp|/Rxexp, where Rxfit is the fitted R1 or R2 and Rxexp is the experimental R1 or R2, for the longitudinal and transverse relaxation rates, respectively, at both field strengths, by using the mathematic software mentioned previously. The deviation of the theoretic from the experimental SNR profiles was assessed with the autocorrelation of the residuals (25), including all data points with an experimental SNR of 5 or more. To compare the goodness of the fit between the sequences, the autocorrelation of the residuals for each profile was divided by the mean SNR of the experimental data points of the profile (Table 3). To test for overall effects of field strength, data in Table 2 were analyzed with two-tailed paired Student t tests (a pair consisting of data obtained at 1.5 T and at 3.0 T, respectively, with a given sequence). Two-tailed P < .05 values were considered to indicate a significant difference. Overall effects of increasing iodine concentration at a given field strength were tested with t tests with column data from Table 2 across all sequences but specific for field strength and iodine concentration. Also, means and SDs for all field-specific column data were evaluated for all column entries in Table 2. These evaluations were performed with spreadsheet software (Excel, release 12.0.6024.500 [EC] 0; Microsoft, Redmond, Wash).


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

  		Table 3. Goodness of Fit of Theoretic to Experimental Profiles with Autocorrelation of the Residuals Divided by Mean SNR at Three Iodine Concentrations

 

   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 ADVANCES IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 
With the coefficients in Table 4, the Equation reproduced the experimental relaxation rates (range, 0.37–38.24 seconds–1) with a mean absolute residual of 4.5% (SD, 5.4%). The analytic signal profiles matched the discrete experimental SNR profiles well (autocorrelation of the residuals divided by a mean of < 5.0), with the exception of the true FISP sequence, in which systematic deviations were observed (Fig 3, Table 3).


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

  		Table 4. Least-Squares Optimized Coefficients for Calculation of Relaxation Rates at Arbitrary Gadopentetate Dimeglumine and Ioversol Concentrations in Normal Saline at 1.5 and 3.0 T by Using Our Equation

 
Dependence of Relaxation Times on Gadopentetate Dimeglumine and Ioversol Concentrations
The presence of ioversol consistently increased the longitudinal and transverse relaxation rates (ie, shortened relaxation times T1 and T2). The slope of the relaxation rate increase for growing gadolinium concentration increased with iodine concentration (A1 or A2 > 0). An additional relaxation acceleration, independent of gadolinium concentration, needed to be considered for a good fit (B1 or B2). This latter effect was stronger for transverse relaxation than it was for longitudinal relaxation and was better reproduced with a second-order term in iodine concentration than with a first-order term.

Overall MR Signal Dependence on Gadopentetate Dimeglumine and Ioversol Concentrations
A stepwise increase in iodine concentration from 0 to 236.4 and to 1182 mmol iodine per liter, corresponding to 0, 30, and 150 mg I/mL, respectively, decreased the mean maximum SNR for all sequences from 177 to 174 (difference was not significant) and 143 (P = .02) at 1.5 T and from 321 to 305 (difference was not significant) and 230 (P < .001) at 3.0 T, respectively. The mean gadolinium concentration, for which maximum SNR was observed, decreased from 1.20 to 0.91 (P = .002) and 0.40 (P = .002) mmol/L at 1.5 T and from 1.12 to 0.90 (P = .005) and 0.54 (P = .01) mmol/L at 3.0 T, respectively. The width of the gadolinium concentration range, within which at least 90% of the maximum SNR was observed, decreased from 3.23 to 2.50 (P = .004) and 1.19 (P = .006) mmol/L at 1.5 T and from 3.25 to 2.57 (P = .006) and 1.42 (P = .006) mmol/L at 3.0 T, respectively.

Profiles of SNR versus Gadopentetate Dimeglumine Concentration for Specific Sequences
Some findings are highlighted here as follows (Table 2; Figs 4, 5). The SNR profiles for T1-weighted SE imaging demonstrated the expected maximum at an intermediate gadolinium concentration, which shifted to lower concentrations and narrowed with increasing iodine concentration. The projected maximum SNR decreased more rapidly at 3.0 T than at 1.5 T for increasing iodine concentration (Fig 5), whereas the optimum gadolinium concentration steadily decreased in parallel at the two field strengths. Optimum gadolinium concentration was observed at 1.4 and 1.3 mmol/L at 1.5 and 3.0 T, respectively. T2-weighted turbo SE sequences with and without fat saturation resulted in near-identical signal profiles. Results for the non–fat-saturating acquisitions were omitted for clarity. The dominant feature of the SNR profile was a sharp decrease at a low gadolinium concentration, which may allow imaging of the intraarticular cavity with negative contrast. The profiles for intermediate-weighted turbo SE sequences with fat saturation resembled those of the T2-weighted sequences, with an onset of the SNR decrease at a higher gadolinium concentration. The profiles for the TIRM sequence were more complex than others (Fig 3). The local maxima to the right of the SNR minima occurred at a comparatively high gadolinium concentration and had the third lowest width after T2- and intermediate-weighted MR imaging.

The SNR for 2D FLASH was prominently higher at 3.0 T than it was at 1.5 T. The profiles resembled those of T1-weighted fast SE MR imaging. However, the optimum gadolinium concentration range was consistently broader for FLASH. True FISP with water-selective excitation showed consistently high SNR for a gadolinium concentration between 0.05 and approximately 3 mmol/L at all iodine concentrations and both field strengths. At the highest iodine concentration, there was a tendency toward forming an SNR maximum at an intermediate gadolinium concentration, which was more pronounced at 3.0 T than at 1.5 T. Signal equations tended to underestimate the signal at a gadolinium concentration of more than approximately 2.00 mmol/L. The SNR profiles for dual-echo imaging with water-selective excitation displayed similarities to those for true FISP, such as a broad gadolinium concentration range with a high SNR (approximately 0.3 and 2.6 mmol/L, independent of iodine concentration and field strength) and a tendency for an initial SNR increase and formation of a peak maximum that was more pronounced at a higher iodine concentration and higher field strength.

The SNR for MEDIC was comparatively high at the two lower iodine concentrations with both field strengths for a gadolinium concentration between 0.10 and 2.00 mmol/L. At the highest ioversol concentration, the profiles were compressed to considerably lower gadolinium concentration, and maximum SNR was reduced. Maximum SNR was only moderately higher at 3.0 T than it was at 1.5 T. In comparison with the other GRE sequences, the width of the gadopentetate dimeglumine concentration range with high SI was narrow. The SNR profiles from 3D VIBE resembled those from 2D FLASH with a pronounced SNR maximum. The 90% SNR range was significantly broader than it was for the T1-weighted SE sequences.

Field Strength Dependence
The maximum SNR of 291 across all sequences and iodine concentrations was higher at 3.0 T than the maximum SNR of 167 at 1.5 T (P < .001). The mean maximum SNR for the iodine concentrations of 0, 236.4, and 1182 mmol/L, corresponding to 0, 30, and 150 mg I/mL, was larger by 81% (P = .009), 75% (P = .008), and 61% (P = .03), respectively, at 3.0 T. No significant differences were found between the two field strengths for gadolinium concentration, resulting in a maximum SNR (P = .92) or mean width of the 90% SNR range (P = .79) across all sequences and iodine concentrations.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
ADVANCES IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 
In an in vitro study, we aimed at characterizing the SNR dependence of a variety of MR imaging sequences commonly employed in musculoskeletal imaging on gadopentetate dimeglumine and ioversol concentrations to facilitate an optimized use of these respective MR imaging and radiographic contrast agents in direct MR arthrography.

As consistent with previous results (8,26), ioversol did not reduce the T1-shortening effect of the gadolinium-based contrast agent (9) but rather boosted gadolinium-induced relaxation enhancement. The main concern in attempts at visualizing the intraarticular cavity with positive contrast in the presence of an iodinated contrast agent is, therefore, the faster decay of transverse magnetization. Thus, minimization of the echo time is of increasing importance at higher iodine concentrations.

The contrast agent ioversol is commercially available with 240–350 mg I/mL. Thus our concentrations of 30 and 150 mg I/mL correspond to a dilution by a factor of 8.0–11.7 and 1.6–2.3, respectively.

The presence of the second-order term in iodine concentration (B1 or B2) in the Equation suggests a capability of ioversol to enhance the relaxation of water magnetization in aqueous solutions even in the absence of gadopentetate dimeglumine, although this enhancement is clearly weaker than that for gadopentetate dimeglumine.

The admixture of ioversol changed three main parameters of the SNR versus gadolinium concentration profiles: (a) The optimum gadolinium concentration range shifted to lower concentrations (8,9). (b) The width of this range narrowed. (c) The maximum SNR most often decreased. A larger SNR loss at 3.0 T compared with that at 1.5 T for most sequences was in agreement with findings in an earlier publication (8). Only moderate SNR reduction or even increased maximum SNR can be expected for T1-weighted short–echo time sequences and low iodine concentrations (8,9).

An optimum gadolinium concentration in the range of 1.25–2.0 mmol/L has been suggested for T1-weighted SE imaging at 1.5 and 3.0 T (8,9) in the absence of iodine, whereas we found, in good agreement, an optimum range of 0.7–3.4 mmol/L, peaking at 1.4 and 1.3 mmol/L at 1.5 and 3.0 T, respectively.

Related in vivo investigations are rare. At first glance, our results and the results of others (8,9) appear to contradict findings in a previous study by Kopka et al (27), who reported "no evident increase in intra-articular signal intensity" in T1-weighted SE images of shoulder joints at 1.5 T obtained 30–45 minutes after direct administration of a 2.5 mmol/L gadopentetate dimeglumine solution. The injected solution appears to have contained 120–150 mg I/mL (27). When using the data and repetition and echo times from Kopka et al (27) for a calculation on the basis of our Equation, our model predicts a narrow SNR maximum, which peaked at a gadolinium concentration of 0.50 mmol/L and near-identical SI values at gadolinium concentrations of 0 and 2.5 mmol/L. This prediction well matches the reported results (27), even though a different iodinated contrast material was used. Thus, although the 2.5 mmol/L gadopentetate dimeglumine concentration was the lowest among three investigated in the study of Kopka et al (27), it might still have been detrimentally high because of high iodine concentration.

With the exception of MEDIC, the GRE sequences (FLASH, true FISP, dual-echo imaging in the steady state, and VIBE) showed a broader gadolinium concentration range with close to maximum SI than did T1-weighted SE imaging at all ioversol concentrations and both field strengths. They may still image the intraarticular cavity with a positive contrast when insufficient enhancement is already observed on T1-weighted SE images (28). In case of insufficient signal enhancement, examinations in joints may therefore be "rescued" with one of these sequences.

MEDIC depends most heavily on T2*, as opposed to T2. Because we determined only T2 values, significant deviations of our results from results with in vivo situations are more likely than for other sequences. However, the theoretic profiles still matched the experimental results reasonably well, with no obvious systematic deviations in the corresponding plots.

The SNR observed in TIRM acquisitions most critically depended on iodine concentration. Gadolinium concentrations producing maximum SNR at high iodine concentrations may well result in near-minimum SNR in the absence of ioversol, and a careful optimization of the examination protocol is therefore required.

Although the maximum SNR at 3.0 T was significantly higher than that at 1.5 T, width and peak location of the optimum gadopentetate dimeglumine range were only slightly changed. This finding is in good agreement with previously published results (8,29,30) and supports the use of identical contrast agent concentrations at the two field strengths.

Our study had limitations. Because data were collected for only three ioversol concentrations, projections to other concentrations are of speculative character. In combination with the unidentified mechanism of the ioversol-induced relaxation enhancement, findings in more detailed studies may result in changes to the form of our Equation and/or to the magnitude of the corresponding coefficients in our study.

Imaging in vivo occurs at a higher temperature than the temperature we used in this study. This difference may partly explain the higher longitudinal molar relaxivity of gadopentetate dimeglumine measured in this work than the typical values reported for albumin or human blood plasma solutions at 37°C (31). Although the differences are not dramatic, appropriate caution should be taken in direct adaptations to in vivo situations. The influence of high-molecular-weight proteins on relaxation rates and SNR profiles has been repeatedly studied in vitro and was found to be comparatively small (8,9,18). Lidocaine or epinephrine usually is not administered intraarticularly during MR arthrography in our institutions, and the influence of either medication has not been investigated in our study (7).

Our results may allow near-quantitative estimations of optimum gadolinium concentrations on the basis of well-established pulse sequence–specific signal equations when arthrograms are acquired with new sequences or new sets of sequence parameters. The documented broad gadolinium concentration range with high SI with some of the sequences evaluated for the first time in our work may allow rescuing examinations in case of inadvertent application of high contrast agent concentrations injected into small joints or unexpected imaging delays. Finally, our results corroborate previous indications that solutions with identical iodine concentrations may be used at 1.5 and 3.0 T.

In conclusion, admixture of ioversol to gadopentetate dimeglumine solutions results in a consistent additional relaxation enhancement, which can be analytically modeled to allow a near-quantitative a priori optimized match of contrast media concentrations and imaging protocol for a broad variety of pulse sequences.

Practical application: A prominent concern for clinical application is the ill-characterized change of contrast agent concentration. A change in such concentration can be caused by an immediate mixing of the injected solution with fluids already present in the intraarticular space, by a small joint space (eg, distal radioulnar joint) with a high iodine concentration relative to the gadolinium concentration, or by a delay between contrast agent application and imaging during which the contrast agent concentration may decrease (32). An injection with higher than optimum contrast agent concentrations may result in improved SNR for T1-weighted SE imaging that is delayed by 90–180 minutes (27). Immediate dilution may be minimized by aspiration of joint effusion prior to contrast agent injection (29). This is of increasing concern at higher iodine concentrations, because of the generally reduced width of the optimum gadolinium concentration range. However, considering the large gadolinium concentration range, within which some GRE sequences promise high SNR, aspiration might often not be necessary.


   ADVANCES IN KNOWLEDGE
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
IMPLICATION FOR PATIENT CARE
 References
 


   IMPLICATION FOR PATIENT CARE
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
References
 


   ACKNOWLEDGMENTS
 
We thank Roger Luechinger, PhD, for his help at the MR imaging units for the determination of the relaxation times.


   FOOTNOTES
 

Abbreviations: FISP = fast imaging with steady-state precession • FLASH = fast low-angle shot • GRE = gradient-recalled echo • MEDIC = multiecho data image combination • SD = standard deviation • SE = spin echo • SI = signal intensity • SNR = signal-to-noise ratio • 3D = three-dimensional • TIRM = turbo inversion-recovery magnitude • 2D = two-dimensional • VIBE = volumetric interpolated breath-hold examination

Author contributions: Guarantors of integrity of entire study, G.A., D.N.; 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, all authors; literature research, G.A., J.H., D.W., D.N.; experimental studies, all authors; statistical analysis, G.A., D.N.; and manuscript editing, G.A., J.M.F., J.H., D.W., C.W.A.P., C.B., D.N.


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 

  1. Grainger AJ, Elliott JM, Campbell RS, Tirman PF, Steinbach LS, Genant HK. Direct MR arthrography: a review of current use. Clin Radiol 2000;55:163–176.[CrossRef][Medline]
  2. Steinbach LS, Palmer WE, Schweitzer ME. Special focus session. MR arthrography. RadioGraphics 2002;22:1223–1246.
  3. Petersilge CA. MR arthrography for evaluation of the acetabular labrum. Skeletal Radiol 2001;30:423–430.[CrossRef][Medline]
  4. Waldt S, Bruegel M, Mueller D, et al. Rotator cuff tears: assessment with MR arthrography in 275 patients with arthroscopic correlation. Eur Radiol 2007;17:491–498.[CrossRef][Medline]
  5. Waldt S, Burkart A, Lange P, Imhoff AB, Rummeny EJ, Woertler K. Diagnostic performance of MR arthrography in the assessment of superior labral anteroposterior lesions of the shoulder. AJR Am J Roentgenol 2004;182:1271–1278.[Abstract/Free Full Text]
  6. Schulte-Altedorneburg G, Gebhard M, Wohlgemuth WA, et al. MR arthrography: pharmacology, efficacy and safety in clinical trials. Skeletal Radiol 2003;32:1–12.[Medline]
  7. Brown RR, Clarke DW, Daffner RH. Is a mixture of gadolinium and iodinated contrast material safe during MR arthrography? AJR Am J Roentgenol 2000;175:1087–1090.[Abstract/Free Full Text]
  8. Masi JN, Newitt D, Sell CA, et al. Optimization of gadodiamide concentration for MR arthrography at 3 T. AJR Am J Roentgenol 2005;184:1754–1761.[Abstract/Free Full Text]
  9. Montgomery DD, Morrison WB, Schweitzer ME, Weishaupt D, Dougherty L. Effects of iodinated contrast and field strength on gadolinium enhancement: implications for direct MR arthrography. J Magn Reson Imaging 2002;15:334–343.[CrossRef][Medline]
  10. Rafii M, Minkoff J. Advanced arthrography of the shoulder with CT and MR imaging. Radiol Clin North Am 1998;36:609–633.[CrossRef][Medline]
  11. Shahbazi-Gahrouei D, Williams M, Allen BJ. In vitro study of relationship between signal intensity and gadolinium-DTPA concentration at high magnetic field strength. Australas Radiol 2001;45:298–304.[CrossRef][Medline]
  12. May DA, Pennington DJ. Effect of gadolinium concentration on renal signal intensity: an in vitro study with a saline bag model. Radiology 2000;216:232–236.[Abstract/Free Full Text]
  13. Nishii T, Tanaka H, Nakanishi K, Sugano N, Miki H, Yoshikawa H. Fat-suppressed 3D spoiled gradient-echo MRI and MDCT arthrography of articular cartilage in patients with hip dysplasia. AJR Am J Roentgenol 2005;185:379–385.[Abstract/Free Full Text]
  14. Schmid MR, Pfirrmann CW, Koch P, Zanetti M, Kuehn B, Hodler J. Imaging of patellar cartilage with a 2D multiple-echo data image combination sequence. AJR Am J Roentgenol 2005;184:1744–1748.[Abstract/Free Full Text]
  15. Yoshioka H, Stevens K, Hargreaves BA, et al. Magnetic resonance imaging of articular cartilage of the knee: comparison between fat-suppressed three-dimensional SPGR imaging, fat-suppressed FSE imaging, and fat-suppressed three-dimensional DEFT imaging, and correlation with arthroscopy. J Magn Reson Imaging 2004;20:857–864.[CrossRef][Medline]
  16. Wutke R, Fellner FA, Fellner C, Stangl R, Dobritz M, Bautz WA. Direct MR arthrography of the shoulder: 2D vs. 3D gradient-echo imaging. Magn Reson Imaging 2001;19:1183–1191.
  17. Recht MP, Goodwin DW, Winalski CS, White LM. MRI of articular cartilage: revisiting current status and future directions. AJR Am J Roentgenol 2005;185:899–914.[Abstract/Free Full Text]
  18. Binkert CA, Zanetti M, Gerber C, Hodler J. MR arthrography of the glenohumeral joint: two concentrations of gadoteridol versus Ringer solution as the intraarticular contrast material. Radiology 2001;220:219–224.[Abstract/Free Full Text]
  19. White LM, Schweitzer ME, Weishaupt D, Kramer J, Davis A, Marks PH. Diagnosis of recurrent meniscal tears: prospective evaluation of conventional MR imaging, indirect MR arthrography, and direct MR arthrography. Radiology 2002;222:421–429.[Abstract/Free Full Text]
  20. Applegate GR, Flannigan BD, Tolin BS, Fox JM, Del Pizzo W. MR diagnosis of recurrent tears in the knee: value of intraarticular contrast material. AJR Am J Roentgenol 1993;161:821–825.[Abstract/Free Full Text]
  21. Bernstein MA, King KF, Zhou XJ. Handbook of MRI pulse sequences. Burlington, Mass: Elsevier, 2004.
  22. Haacke EM, Brown RW, Thompson MR, Venkatesan R. Magnetic resonance imaging: physical principles and sequence design. New York, NY: Wiley, 1990.
  23. Hardy PA, Recht MP, Piraino D, Thomasson D. Optimization of a dual echo in the steady state (DESS) free-precession sequence for imaging cartilage. J Magn Reson Imaging 1996;6:329–335.[Medline]
  24. Nitz WR. Magnetic resonance imaging: sequence acronyms and other abbreviations in MR imaging [in German]. Radiologe 2003;43:745–763.[CrossRef][Medline]
  25. Femenias JL. Goodness of fit: analysis of residuals. J Mol Spectrosc 2003;217:32–42.[CrossRef]
  26. Jinkins JR, Robinson JW, Sisk L, Fullerton GD, Williams RF. Proton relaxation enhancement associated with iodinated contrast agents in MR imaging of the CNS. AJNR Am J Neuroradiol 1992;13:19–27.[Abstract]
  27. Kopka L, Funke M, Fischer U, Keating D, Oestmann J, Grabbe E. MR arthrography of the shoulder with gadopentetate dimeglumine: influence of concentration, iodinated contrast material, and time on signal intensity. AJR Am J Roentgenol 1994;163:621–623.[Abstract/Free Full Text]
  28. Aydingoz U, Kerimoglu U, Canyigit M. "Black" contrast effect during magnetic resonance arthrography attributable to inadvertent administration of excessive gadolinium chelates. J Comput Assist Tomogr 2005;29:333–335.[CrossRef][Medline]
  29. Stecco A, Brambilla M, Puppi AM, Lovisolo M, Boldorini R, Carriero A. Shoulder MR arthrography: in vitro determination of optimal gadolinium dilution as a function of field strength. J Magn Reson Imaging 2007;25:200–207.[CrossRef][Medline]
  30. Trattnig S, Pinker K, Ba-Ssalamah A, Nobauer-Huhmann IM. The optimal use of contrast agents at high field MRI. Eur Radiol 2006;16:1280–1287.[CrossRef][Medline]
  31. Pintaske J, Martirosian P, Graf H, et al. Relaxivity of gadopentetate dimeglumine (Magnevist), gadobutrol (Gadovist), and gadobenate dimeglumine (MultiHance) in human blood plasma at 0.2, 1.5, and 3 Tesla. Invest Radiol 2006;41:213–221.[CrossRef][Medline]
  32. Andreisek G, Duc SR, Froehlich JM, Hodler J, Weishaupt D. MR arthrography of the shoulder, hip, and wrist: evaluation of contrast dynamics and image quality with increasing injection-to-imaging time. AJR Am J Roentgenol 2007;188:1081–1088.[Abstract/Free Full Text]




This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2473071013v1
247/3/706    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 Andreisek, G.
Right arrow Articles by Nanz, D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Andreisek, G.
Right arrow Articles by Nanz, D.

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