| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Developments |
1 From the Department of Radiology, Stanford University School of Medicine, 300 Pasteur Dr, SO-68B, Stanford, CA 94305-5105 (G.E.G., S.B.R., H.Y., N.J.P., C.F.B.); Department of Radiology, University of Leiden, Leiden, the Netherlands (P.K.); and GE Healthcare Applied Sciences Laboratory West, Menlo Park, Calif (A.S.S., J.W.J., J.H.B.). Received February 18, 2005; revision requested April 19; revision received August 26; accepted September 22; final version accepted October 19. Supported by NIH grant EB002524-01 and the Whitaker Foundation. Address correspondence to G.E.G. (e-mail: gold{at}stanford.edu).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCE IN KNOWLEDGEReferences |
|---|
© RSNA, 2006
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCE IN KNOWLEDGEReferences |
|---|
Two-dimensional techniques for imaging articular cartilage in patients with osteoarthritis include intermediate-weighted fast spin echo (4), T2-weighted fast spin echo (5), and short echo time projection reconstruction (6). Although these techniques provide excellent image quality, the presence of section gaps leads to lower anatomic coverage and less quantitative information than are available with three-dimensional (3D) techniques.
Considerable work in osteoarthritis has been devoted to screening high-risk patients with 3D imaging techniques. High accuracy for depiction of cartilaginous lesions has been shown with fat-suppressed spoiled gradient-echo (GRE) imaging (7). There are two main disadvantages to this approach: lack of reliable contrast between cartilage and fluid that outlines surface defects and long imaging times (approximately 8–12 minutes) to achieve the image resolution and the signal-to-noise ratio (SNR) needed for imaging cartilage. Unfortunately, overall signal intensity is reduced with fat-suppressed spoiled GRE, compared with balanced steady-state free precession (bSSFP) techniques (8).
Several other 3D imaging methods have been proposed for the assessment of cartilage, including driven-equilibrium Fourier transform (9–11). Another 3D GRE technique that provides bright synovial fluid is 3D dual-echo imaging (12). Fluctuating equilibrium MR imaging is a variant of bSSFP and has been used to image cartilage at 1.5 T (13). Phase-sensitive bSSFP also has been used to provide fat suppression for imaging cartilage (14).
At 3.0 T, techniques for fat saturation with bSSFP such as fluctuating equilibrium MR imaging and phase-sensitive bSSFP are impractical because of the short repetition time required to place fat and water in different pass bands (13,14). Other bSSFP approaches that may provide more reliable fat suppression at 3.0-T MR imaging are linear combination bSSFP (15), Dixon MR imaging (16,17), or intermittent fat-suppressed bSSFP (18).
The purpose of our study was to compare iterative decomposition of water and fat with echo asymmetry and least-squares estimation (IDEAL) bSSFP, fat-suppressed bSSFP, and fat-suppressed spoiled GRE sequences for 3D MR imaging of articular cartilage at 3.0 T.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCE IN KNOWLEDGEReferences |
|---|
Volunteers and Imaging
Our Health Insurance Portability and Accountability Act–compliant study received approval of the institutional review board; patient informed consent was obtained. We imaged the right and left knees of five healthy volunteers (four men, one woman; age range, 26–42 years). To test use of our sequences in volunteers without osteoarthritis, we recruited volunteers who were asymptomatic and younger than 50 years old and had no history of knee injury or prior surgery. All images were acquired with a 3.0-T MR unit (Signa; GE Healthcare, Milwaukee, Wis) with high-performance gradients (maximum gradient strength of 40 mT/m and slew rate of 150 mT/m/sec) by using a transmit-receive quadrature knee coil (GE Healthcare Technologies, Waukesha, Wis). Acquisition resolution was kept constant among all imaging sequences, with a matrix of 256 x 256, a section thickness of 1.5 mm, a field of view of 17 cm, a receive bandwidth of ±62.5 kHz, and 52 sagittal sections over the knee joint.
IDEAL bSSFP images were acquired with a repetition time of 5.1 msec and a flip angle of 30°. Three echoes were acquired at 1.3, 1.7, and 2.0 msec, and imaging time was 3 minutes 40 seconds (Fig 1). A homodyne reconstruction algorithm was used to prevent image blurring caused by the partial echo acquisition (19). Three-dimensional fat-suppressed spoiled GRE images were acquired with repetition time msec/echo time msec of 13.5/1.5, three acquisitions, a flip angle of 10°, and fat suppression; total imaging time was 9 minutes 40 seconds. Fat-suppressed bSSFP images were acquired with 5.6/1.5, a flip angle of 30°, and an imaging time of 2 minutes 13 seconds. For fat-suppressed bSSFP, we used an intermittent chemically selective fat suppression pulse with transitions in and out of the steady state to minimize artifacts. Flip angles were chosen to maximize cartilaginous signal for the fat-suppressed spoiled GRE and bSSFP techniques (20,21). Images were acquired in the sagittal plane to evaluate the articular cartilage of the knee.
|
 |
Image quality was assigned a grade for all 10 sets of images on a scale of 0–3 as follows: grade 0, high image noise with artifacts; grade 1, high image noise with minor artifacts; grade 2, high image noise and no artifacts; and grade 3, low image noise and no artifacts. Uniformity of fat suppression or fat-water separation was assigned a grade on a scale of 0–3 as follows: grade 0, water suppression; grade 1, fat near cartilage not suppressed; grade 2, patchy failure of fat suppression not near cartilage; and grade 3, near-perfect fat suppression. Differences in the assignment of a grade were resolved in consensus. The same radiologists as previously mentioned assigned grades for both image quality and uniformity of fat saturation, and the radiologists were blinded as to the sequence used to obtain the images for which they were assigning a grade.
Statistical Analysis
Statistical analysis was performed by using software (Excel 11.1.1, Microsoft, Redmond, Wash; SPSS 11.0, SPSS, Chicago, Ill). For each of the variables, analysis of variance or analysis with a nonparametric analogue (Friedman test) was performed. For variables that were significant (P < .05) with analysis of variance or the Friedman test, the post hoc pairwise comparison was performed. The three imaging sequences were pairwise compared with respect to SNR, CNR, SNR efficiency, and CNR efficiency of cartilage and fluid by using a paired-sample t test. The mean scores of the two radiologists for the three imaging sequences also were pairwise compared with respect to image quality and uniformity of fat suppression or fat-water separation by using a Wilcoxon signed rank test. A difference with a P value of less than .05 was considered significant for the paired t tests and the Wilcoxon signed rank test.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCE IN KNOWLEDGEReferences |
|---|
Mean SNR efficiency of cartilage was statistically equal for IDEAL bSSFP (1.5 ± 0.3) and fat-suppressed bSSFP sequences (1.3 ± 0.5) (Fig 2), but it was significantly higher than that for the fat-suppressed spoiled GRE sequence (1.0 ± 0.2, P < .01). Mean SNR efficiency of fluid was significantly higher for the IDEAL bSSFP sequence (5.3 ± 1.7) than it was for the fat-suppressed bSSFP sequence (4.4 ± 1.4, P < .01). Mean SNR efficiency of fluid was higher for both of the bSSFP techniques than it was for the fat-suppressed spoiled GRE sequence (0.7 ± 0.2, P < .01), as was expected because the fat-suppressed spoiled GRE sequence is not sensitive to fluid. The CNR efficiency between fluid and cartilage was higher for the IDEAL bSSFP sequence (3.8 ± 1.4) (Fig 3) than it was for the fat-suppressed bSSFP (3.0 ± 0.9) and the fat-suppressed spoiled GRE sequence (0.2 ± .04, P < .01), primarily because of lower fluid signal intensity on the fat-suppressed spoiled GRE images. Fluid was easily distinguished from cartilage with IDEAL bSSFP and fat-suppressed bSSFP sequences but not with the fat-suppressed spoiled GRE sequence.
|
|
IDEAL bSSFP images showed excellent cartilaginous signal intensity and uniformity (Fig 4) and had the capability to correct for chemical shift (23). Images obtained with all three sequences showed excellent cartilaginous signal intensity (Fig 5), whereas those obtained with the bSSFP-based techniques also showed bright synovial fluid.
|
|
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCE IN KNOWLEDGEReferences |
|---|
IDEAL bSSFP and fat-suppressed bSSFP images provide bright fluid and an arthrogram-like effect. This effect is similar to the contrast seen on images obtained with driven-equilibrium Fourier transform and intermediate-weighted fast spin echo. This effect may highlight surface lesions and outline cartilaginous fissures (25) but could reduce the dynamic range in the cartilage overall. The SNR of fluid on fat-suppressed spoiled GRE images is low, as this technique is insensitive to fluid. Comparison of IDEAL bSSFP with fat-suppressed spoiled GRE MR imaging for detection of cartilaginous lesions will be the subject of further study.
The SNR efficiency of the IDEAL bSSFP and fat-suppressed bSSFP sequences for MR imaging of cartilage was much higher than it was for the fat-suppressed spoiled GRE sequence. The IDEAL bSSFP sequence had a slightly higher SNR efficiency of cartilage, compared with that of the fat-suppressed bSSFP sequence, although the difference was not statistically significant. The SNR efficiency of fluid and the fluid-cartilage CNR efficiency were significantly higher for the IDEAL bSSFP sequence than they were for the fat-suppressed bSSFP sequence. This difference is likely caused by the loss of imaging efficiency due to coming out of the steady state to apply the fat saturation pulse and saturation of a portion of the water signal with the fat saturation pulse. Overall, IDEAL bSSFP images received significantly higher image quality grades and fat suppression grades compared with those assigned to fat-suppressed bSSFP images.
The IDEAL bSSFP sequence increases repetition time slightly to accommodate shifts in echo time, with a reduction in SNR efficiency. However, performance of multiple acquisitions at different echo times improves SNR through effective averaging (17). Unlike with fat-suppressed bSSFP, the steady state is maintained throughout, and this feature allows elimination of the need for discarded acquisitions and prevention of artifacts caused by transient magnetization. The IDEAL bSSFP sequence also provides lipid images and images corrected for chemical shift (23) that may be useful in areas of thin cartilage.
Intermediate-weighted fast spin-echo imaging without fat saturation is a popular technique for imaging articular cartilage (4). Compared with imaging at 1.5 T, however, chemical shift is doubled at 3.0 T. IDEAL bSSFP and IDEAL fast spin-echo (26) techniques allow correction of chemical shift (23). Although the IDEAL technique requires multiple acquisitions, imaging time can be reduced by using parallel imaging (17).
The limitations of this study included the small number of normal volunteers. In patients with osteoarthritis, signal characteristics of the cartilage and fluid may be different, and this difference reduces the SNR or CNR advantages of the bSSFP-based techniques. Segmentation of cartilage was not performed in this study, so volume and thickness results were not compared. Finally, patients with cartilaginous defects were not compared.
Volume and thickness measurements from 3D techniques may be useful in following up subjects with osteoarthritis. The fat-suppressed spoiled GRE sequence commonly is used for such measurements, but it is time consuming and does not provide bright fluid to outline surface defects (27). The IDEAL bSSFP sequence has the potential to replace fat-suppressed spoiled GRE for quantification of thickness and volume at 3.0 T, with a much more reasonable imaging time. The capability of bSSFP-based techniques to aid the accurate measurement of thickness and volume of cartilage may depend on the sensitivity of these techniques to cartilage in the tidemark zone and will be the subject of further study.
In conclusion, the IDEAL bSSFP sequence is considered a promising method for the evaluation of articular cartilage. The SSFP sequence provides excellent contrast between joint fluid and articular cartilage that is helpful for the detection of surface defects (25). Studies with more subjects to further demonstrate the capability of the IDEAL bSSFP technique to help depict cartilaginous lesions and to be used for measurement of thickness and volume of cartilage in subjects with osteoarthritis are important for the improvement of this technique.
|
ADVANCE IN KNOWLEDGE |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCE IN KNOWLEDGEReferences |
|---|
|
FOOTNOTES |
|---|
Abbreviations: bSSFP = balanced steady-state free precession • CNR = contrast-to-noise ratio • GRE = gradient echo • IDEAL = iterative decomposition of water and fat with echo asymmetry and least-squares estimation • SNR = signal-to-noise ratio • 3D = three-dimensional
See Materials and Methods for pertinent disclosures.
Author contributions: Guarantor of integrity of entire study, G.E.G.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, G.E.G., S.B.R., H.Y., C.F.B.; clinical studies, all authors; and manuscript editing, G.E.G., S.B.R., J.H.B.
|
References |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCE IN KNOWLEDGEReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. A. Chen, W. Lu, C. T. John, B. A. Hargreaves, S. B. Reeder, S. L. Delp, R. A. Siston, and G. E. Gold Multiecho IDEAL Gradient-Echo Water-Fat Separation for Rapid Assessment of Cartilage Volume at 1.5 T: Initial Experience Radiology, June 15, 2009; (2009) 2522081424. [Abstract] [Full Text]
|
|
|
|
|
|
R. Kijowski, D. G. Blankenbaker, J. L. Klaers, K. Shinki, A. A. De Smet, and W. F. Block Vastly Undersampled Isotropic Projection Steady-State Free Precession Imaging of the Knee: Diagnostic Performance Compared with Conventional MR Radiology, April 1, 2009; 251(1): 185 - 194. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. R. Duc, C. W. A. Pfirrmann, P. P. Koch, M. Zanetti, and J. Hodler Internal Knee Derangement Assessed with 3-minute Three-dimensional Isovoxel True FISP MR Sequence: Preliminary Study Radiology, February 1, 2008; 246(2): 526 - 535. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. E. Anderson, B. J. Ellis, C. L. Peters, and J. A. Weiss Cartilage Thickness: Factors Influencing Multidetector CT Measurements in a Phantom Study Radiology, January 1, 2008; 246(1): 133 - 141. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. M. Gerdes, R. Kijowski, and S. B. Reeder IDEAL Imaging of the Musculoskeletal System: Robust Water Fat Separation for Uniform Fat Suppression, Marrow Evaluation, and Cartilage Imaging Am. J. Roentgenol., November 1, 2007; 189(5): W284 - W291. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Biswal, D. L. Resnick, J. M. Hoffman, and S. S. Gambhir Molecular Imaging: Integration of Molecular Imaging into the Musculoskeletal Imaging Practice Radiology, September 1, 2007; 244(3): 651 - 671. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
