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


     

DOI: 10.1148/radiol.2412051330
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
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 HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bauer, J. S.
Right arrow Articles by Link, T. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bauer, J. S.
Right arrow Articles by Link, T. M.
(Radiology 2006;241:399-406.)
© RSNA, 2006

Experimental Studies

Volumetric Cartilage Measurements of Porcine Knee at 1.5-T and 3.0-T MR Imaging: Evaluation of Precision and Accuracy1

Jan S. Bauer, MD, Stefanie J. Krause, MD, Christian J. Ross, MD, Roland Krug, PhD, Julio Carballido-Gamio, PhD, Eugene Ozhinsky, BA, Sharmila Majumdar, PhD and Thomas M. Link, MD

1 From Musculoskeletal and Quantitative Imaging Research, Department of Radiology, University of California, San Francisco, Calif. Received August 9, 2005; revision requested October 17; revision received November 4; accepted December 1; final version accepted February 6, 2006. Supported by National Institutes of Health grants AR46905 and AG17762. Address correspondence to J.S.B., Institut für Röntgendiagnostik, Technische Universität München, Ismaninger Strasse 22, 81675 München, Germany (e-mail: jsb{at}roe.med.tum.de).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
Purpose: To compare the precision and accuracy of 3.0-T and 1.5-T magnetic resonance (MR) imaging in the quantification of cartilage volume by using direct volumetric measurements as a reference standard.

Materials and Methods: The local animal experimentation committee did not require its approval for this study. Porcine knees were obtained from an abattoir. These specimens were used to optimize imaging parameters regarding effective signal-to-noise ratio (SNRE) and contrast-to-noise ratio (CNRE) for a fat-saturated spoiled gradient-recalled acquisition in the steady state (SPGR) sequence, a water excitation SPGR sequence, and a fast spin-echo sequence at 3.0 T and a fat-saturated SPGR sequence at 1.5 T. By using the optimized sequences, 18 specimens were imaged in less than 6 minutes per sequence. A fivefold repetition of measurements of four specimens was performed for precision analysis. Cartilage was segmented by using semiautomatic software to calculate the volume. After imaging, the cartilage was scraped off and the volume was measured directly by using a saline-displacement method to calculate accuracy. Precision and accuracy errors were calculated as the root-mean-squares of the single errors per specimen.

Results: SNRE and CNRE values, respectively, were highest for the water excitation sequence at 3.0 T (1.81 sec–1/2 and 1.27 sec–1/2), followed by the fat-saturated SPGR sequence (1.52 sec–1/2 and 1.07 sec–1/2). The fast spin-echo sequence and the fat-saturated SPGR sequence at 1.5 T had lower SNRE (1.27 sec–1/2 and 0.59 sec–1/2, respectively). Accuracy error for MR-based volume calculation at the femur was 5.0%, 3.0%, 21%, and 16% for the water excitation, fat-saturated SPGR, and fast spin-echo sequences at 3.0 T and the fat-saturated SPGR sequence at 1.5 T, respectively.

Conclusion: MR imaging at 3.0 T was shown in our study to better quantify cartilage volume. SNRE and CNRE were substantially improved, resulting in significantly higher accuracy in determining cartilage volume.

© RSNA, 2006


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
Osteoarthritis is a slowly progressing disease that is characterized by morphologic and molecular changes of cartilage and adjacent bone. The prevalence increases with advancing age to more than 20% for men older than 65 years and to more than 30% for women older than 65 years (1). Similar percentages are found in younger populations with previous knee injuries (1). Pain, stiffness, and disability associated with osteoarthritis markedly reduce life quality. This impact on our society led to the declaration of the "Bone and Joint Decade" and motivated the medical and pharmacologic community to develop new drugs and therapies (2). The assessment of the outcome of these new therapies, however, is still challenging, and there is a need for biomarkers that can be used to evaluate early osteoarthritis. Current studies are thus focusing on the development and validation of noninvasive imaging methods (3).

Imaging methods generally can be classified as those that are used to assess biochemical parameters of cartilage and those that are used to assess morphologic parameters. Tests for biochemical markers focus on determination of glycosaminoglycan or water content and matrix composition. In terms of magnetic resonance (MR) imaging, T2 and T1{rho} quantification, as well as the delayed gadolinium-enhanced MR imaging of cartilage method, have shown great potential in the determination of cartilage properties (3,4). Further studies must determine if these properties are predictive of final clinical outcome and thus have potential as biomarkers for osteoarthritis. The same applies to such morphologic measures as cartilage volume and thickness, although with these measures, at least, some initial data exist and the pathophysiologic mechanisms are well understood (5,6).

Cartilage loss is one of the early alterations in osteoarthritis and can be noninvasively monitored with volumetric measurements obtained at high-spatial-resolution MR imaging (7–10). Indeed, it has been proposed that MR imaging is the "optimal modality for assessing articular cartilage because of superior tissue contrast, direct visualization of articular cartilage and multiplanar capability" (11). Results of different studies have shown the feasibility of this method, but, because high spatial resolution and signal-to-noise ratio are needed, imaging time was a major drawback in these studies. However, recent developments in MR imaging, including new pulse sequences and coils and higher field strengths, can improve these imaging variables (12–14). In particular, MR imaging at 3.0 T has been identified as having great potential in optimizing cartilage imaging (12,14–16); nevertheless, there is little experience with high-spatial-resolution cartilage imaging at 3.0 T, and, to our knowledge, no study as yet has assessed the accuracy of volumetric cartilage measurements at 3.0 T. The purpose of our study, therefore, was to compare 3.0-T and 1.5-T MR imaging in terms of precision and accuracy in quantifying cartilage volume by using direct volumetric measurements as the reference standard.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
Specimens
The anatomy of porcine knee joints at MR imaging is similar to that of humans. Porcine knee joints are therefore well suited as an animal model for cartilage quantification. Eighteen fresh porcine knee specimens were obtained at a local abattoir. Thus, the local animal experimentation committee did not require its approval for our study. Although skin and muscles had been removed, the knee joints remained intact. Between measurements, the knees were stored at –80°C. For imaging, the knees were thawed to room temperature in a water bath that contained physiologic saline.

Imaging Parameters
MR images of the porcine knee specimens were obtained at 1.5 and 3.0 T (Signa; GE Healthcare, Milwaukee, Wis) by using quadrature knee coils (GE Healthcare; and Clinical MR Solutions, Brookfield, Wis). For volumetric cartilage measurements, high-spatial-resolution sequences that yield homogeneous high signal intensity (SI) throughout the cartilage and high contrast with the surrounding tissues are required. Spoiled gradient-recalled acquisition in the steady state (SPGR) sequences have been previously advocated for this purpose (7,10,17). Images were acquired by using a three-dimensional fat-saturated SPGR sequence, a three-dimensional water excitation SPGR sequence, and a two-dimensional fast spin-echo sequence at 3.0 T and a fat-saturated SPGR sequence at 1.5 T. All images were obtained in the sagittal plane with a field of view of 10 cm, an imaging matrix of 512 x 256 pixels, and a section thickness of 1.5 mm.

To optimize the sequences used, the properties of each sequence were first simulated by numeric integration of the Bloch equation, as previously described (18). SIs and tissue contrasts were calculated for different flip angles and echo and repetition times. For the fast spin-echo sequence, different echo train lengths were also simulated. The simulation was based on T1 and T2 values previously published for cartilage, bone marrow, synovial fluid, and muscle (12). The best parameters were chosen, and, with these parameters, optimization was repeated with both imaging units. Images for all sequences were acquired with varying bandwidths, flip angles, and echo times. The repetition time was chosen to be as short as possible within a range of 22–30 msec because this resulted in the highest contrast-to-noise ratio in the theoretic optimization procedure. Other parameters were kept constant.

So that we could compare the different settings, effective signal-to-noise ratio (SNRE) and contrast-to-noise ratio (CNRE) were calculated. SNRE was defined as the SI of cartilage divided by the standard deviation of the background and the square root of the imaging time. CNRE was defined as the SI of the cartilage minus the SI of the adjacent tissue (including synovial fluid, fat, menisci, muscles, and bone marrow) divided by the standard deviation of the background and the square root of the imaging time. CNRE was calculated independently for all adjacent tissues, and an average CNRE was calculated by weighting all single CNRE values with regard to the contact area with the cartilage, as previously described by Kornaat et al (19). After this optimization, the best parameters with an imaging time of less than 6 minutes were chosen for each sequence.

Cartilage Imaging and Volume Measurement
Eighteen knees were imaged by using the optimized sequences. To determine the precision error for each sequence, imaging of four randomly selected knees was repeated five times. That is, the knee was removed from the knee coil, the joint was flexed and extended, and the knee was put back into the coil and repositioned.

The segmentation of the cartilage in the MR images and the consecutive calculation of cartilage volumes were performed by using a semiautomatic spline-based method previously described by Ozhinsky et al (20). In this procedure, the cartilage surface and cartilage-bone interface were represented by cubic Bézier splines. The splines were selected manually in representative key sections, while the program interpolated the splines for the remaining sections. Only minor adjustments were required in these sections. Because all imaging sequences were not useful for a differentiation between articular and epiphyseal cartilage, an anatomic landmark was defined at the posterior corner of the thick triangular cartilage region that extends from the femoral epiphyseal growth plate. This landmark was used during the scraping-off process, as well as for all imaging sequences. Segmentation of one patella required approximately 10 minutes, while segmentation of the femoral cartilage required 25 minutes. This segmentation was performed by three authors (J.S.B., S.J.K., and C.J.R., with 3 years, 1 year, and 1 year of experience in musculoskeletal MR imaging, respectively) under the direct supervision of another author (T.M.L., with 15 years of experience in musculoskeletal MR imaging). After the segmentation was complete, the volume of the cartilage was calculated with a pixel-based approach.

Reference-Standard Direct Volumetric Measurements
The volume of the cartilage was measured by using a water displacement technique previously described by Peterfy et al (10) and Burgkart et al (7). After all imaging studies had ended, the femoral and patellar cartilages were harvested separately. Ligaments and menisci were resected so that there was full access to the joint surface. A scalpel was used to remove the cartilage; great care was taken not to damage or resect the subchondral bone. The tissue specimens were placed in plastic bags filled with physiologic saline to avoid shrinkage of cartilage before the specimens were placed in a graduated cylinder filled with physiologic saline. The cylinder had a scale that measured volume in 0.1-mL increments. To optimize measurement precision, the cartilage specimens were evacuated of air before the measuring procedure by using a standard vacuum device (model 5KH33DN16GX, GE, Fairfield, Conn). The increase in the water level was measured to obtain cartilage volume. All procedures for the direct volume measurements were performed by two authors (J.S.B. and C.J.R.).

Statistical Analysis
To determine measurement precision with the different sequences, imaging and analysis of four specimens were repeated five times. The coefficient of variation (standard deviation divided by mean volume) was calculated as the reproducibility error for each of the four specimens. The precision error of the method was determined as the root-mean-square of the four measured individual reproducibility errors (21). The correlation between the direct and the MR-based volumetric measurements was determined by calculating Pearson correlation coefficients. Differences between these correlations were analyzed by using the Fisher Z transformation, with P < .05 indicating a significant difference. The accuracy error was considered to be the root-mean-square of the individual relative differences between directly measured and MR-based cartilage volumes. Differences between the precision and accuracy errors of each technique were assessed with the Wilcoxon signed rank test and a significance threshold of P < .05. All statistical computations were performed by using software (JMP, version 5.1, SAS Institute, Cary, NC; and SPSS, version 11.5, SPSS, Chicago, Ill).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
Imaging Parameters
Optimal SI was found for a flip angle of 13° with an echo time of 10.6 msec and a repetition time of 27.4 msec (Fig 1). The best calculated tissue contrast, however, was found with higher flip angles. For example, the best contrast between cartilage and fluid was at a flip angle of 20°. Although calculating SI and tissue contrast and measuring SNRE and CNRE gave similar results for optimizing repetition time and flip angle, the hardware limited the usage of the optimal echo time and bandwidth. In general, lower echo times and bandwidth both result in higher SI, but adequate acquisition of the Fourier space limits shortening of the echo time (Table 1).


Figure 1
View larger version (64K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1: Graph shows calculated SI for the 3.0-T fat-saturated SPGR sequence with different repetition times (TR) and flip angles ({alpha}) at an echo time of 10.6 msec; higher SIs are plotted as lighter grays.

 

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

  		Table 1. Sequence Parameters after Optimization

 
If only a part of the Fourier space was acquired, CNRE in particular decreased (Fig 2). For the fat-saturated SPGR sequence at 3.0 T, the measured CNRE between cartilage and surrounding tissue decreased by a factor of four when partial rather than full Fourier space was acquired. A bandwidth of ±15.6 kHz resulted in the highest SNRE and CNRE without a relevant increase in artifacts. The highest measured CNRE was found for a flip angle of 20° with the fat-saturated SPGR sequences at 1.5 and 3.0 T and for a flip angle of 12° with the water excitation sequence (Fig 2). By using these optimized parameters, both fat-saturated SPGR and water excitation sequences at 3.0 T could be performed within an imaging time of 6 minutes. To achieve the same imaging time at 1.5 T, the bandwidth had to be increased to ±31.25 kHz.


Figure 2
View larger version (22K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2a: Graphs show calculated SNRE (left) and CNRE (right) values (plotted along y-axis in sec–1/2) for (a) different bandwidths, (b) different flip angles, and (c) different echo times (TE). Required echo times for partial and full Fourier space acquisition, respectively, were 3.5 and 11.5 msec (for the 1.5-T fat-saturated [fs] SPGR sequence), 3.2 and 10.6 msec (for the 3.0-T fat-saturated SPGR sequence), and 5.8 and 11.0 msec (for the 3.0-T water excitation [WE] sequence). FSE = fast spin echo.

 

Figure 2
View larger version (19K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2b: Graphs show calculated SNRE (left) and CNRE (right) values (plotted along y-axis in sec–1/2) for (a) different bandwidths, (b) different flip angles, and (c) different echo times (TE). Required echo times for partial and full Fourier space acquisition, respectively, were 3.5 and 11.5 msec (for the 1.5-T fat-saturated [fs] SPGR sequence), 3.2 and 10.6 msec (for the 3.0-T fat-saturated SPGR sequence), and 5.8 and 11.0 msec (for the 3.0-T water excitation [WE] sequence). FSE = fast spin echo.

 

Figure 2
View larger version (17K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2c: Graphs show calculated SNRE (left) and CNRE (right) values (plotted along y-axis in sec–1/2) for (a) different bandwidths, (b) different flip angles, and (c) different echo times (TE). Required echo times for partial and full Fourier space acquisition, respectively, were 3.5 and 11.5 msec (for the 1.5-T fat-saturated [fs] SPGR sequence), 3.2 and 10.6 msec (for the 3.0-T fat-saturated SPGR sequence), and 5.8 and 11.0 msec (for the 3.0-T water excitation [WE] sequence). FSE = fast spin echo.

 
For the fast spin-echo sequence, repetition time and the echo train length were important parameters during optimization. An echo train length of seven in combination with a repetition time of 2000 msec resulted in the highest SNRE. The resulting SNRE and CNRE, respectively, were highest for the water excitation sequence at 3.0 T (1.81 sec–1/2 ± 0.13 [standard deviation] and 1.27 sec–1/2 ± 0.13), followed by the fat-saturated SPGR sequence at 3.0 T (1.52 sec–1/2 ± 0.13 and 1.07 sec–1/2 ± 0.13 [Table 1, Fig 3]). SNRE and CNRE, respectively, were lower for the fast spin-echo sequence (1.27 sec–1/2 ± 0.20 and 0.74 sec–1/2 ± 0.24) and lowest for the SPGR sequence at 1.5 T (0.59 sec–1/2 ± 0.16 and 0.34 sec–1/2 ± 0.26). Both SNRE and CNRE were significantly different among all sequences (P < .05).


Figure 3
View larger version (167K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3: Representative sagittal images obtained with optimized MR imaging sequences. A, 1.5-T fat-saturated SPGR sequence, B, 3.0-T fat-saturated SPGR sequence, C, 3.0-T water excitation sequence, and, D, 3.0-T fast spin-echo sequence. Contrast between bone marrow and cartilage (solid arrowheads) was compromised by noise in the 1.5-T SPGR sequence and by blurring in the fast spin-echo sequence. Delineation of adjacent cartilage layers (open arrowheads) was better with the fat-saturated SPGR sequence than with the water excitation sequence at 3.0 T.

 
Cartilage Volumetry
The femoral and the patellar cartilage, respectively, had an average volume of 2.2 mL ± 0.34 and 7.1 mL ± 1.14, as measured with the water displacement method. The cartilage volume was underestimated with all MR imaging sequences, but the error was relatively small for the fat-saturated SPGR sequence and the water excitation sequence at 3.0 T (0.08 mL [0.9%] and 0.25 mL [2.3%], respectively [Table 2]). The error was substantially higher for the fat-saturated SPGR sequence at 1.5 T and the fast spin-echo sequence at 3.0 T (1.23 mL [13.2%] and 1.49 mL [16.0%]).


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

  		Table 2. Absolute Cartilage Volume and Relative Differences between Direct and MR-based Volume Measurements

 
Measurement of the cartilage volume in four specimens after repositioning them five times revealed that the precision error was 4.8% for the fat-saturated SPGR sequence at 1.5 T, 2.7% for the fat-saturated SPGR sequence at 3.0 T, and 4.0% for the water excitation sequence at 3.0 T. A slight increase in all precision errors was found when the patella and femur were analyzed separately (0.5%–1.0%). Differences were not statistically significant (P > .05, Wilcoxon signed rank test [Table 3]).


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

  		Table 3. Precision Errors Calculated as Root-Mean-Squares of Individual Reproducibility Errors

 
Correlation with the volume measured with the water displacement method was significantly higher for the fat-saturated SPGR sequence at 3.0 T (r = 0.99) and the water excitation sequence at 3.0 T (r = 0.97) than for the fast spin-echo sequence at 3.0 T (r = 0.76) and the fat-saturated SPGR sequence at 1.5 T (r = 0.88) (P < .05 for all [Fig 4, Table 4]). The accuracy error for the total cartilage volume was significantly smaller for the fat-saturated SPGR sequence at 3.0 T (3.0%) than for the same sequence at 1.5 T (14%, P < .01) (Table 4). At 3.0 T, the fat-saturated SPGR and water excitation sequences had similar accuracy errors (3.0% and 4.0%, respectively), while the accuracy error of the fast spin-echo sequence was significantly (P < .01) higher (19%). The differences between the accuracy errors were more pronounced at the femur (P < .01) than at the patella, where only a trend was shown (P = .06).


Figure 4
View larger version (8K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4: Graphs show correlation of volumetric cartilage measurements at fat-saturated (fs) SPGR MR imaging at 1.5 T (left) and 3.0 T (right) with cartilage volumes determined by using the water displacement method. The specific correlation coefficients for different volumes are shown in Tables 2 and 4.

 

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

  		Table 4. Accuracy Errors Calculated as Root-Mean-Squares of Relative Differences and Correlations between Direct and MR-based Volume Measurements

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
ADVANCES IN KNOWLEDGE
 References
 
In this study, the potential of MR imaging at 3.0 T in the optimization of volumetric cartilage measurements was demonstrated: SNRE and CNRE improved more than twofold compared with standard imaging at 1.5 T, and significantly higher accuracy in cartilage volume quantification was observed.

Although cartilage morphometry has been extensively studied and validated at 1.5 T (6,7,10,17,22,23), experience with 3.0-T systems is limited (14). Theoretically, the doubled field strength results in twice the signal-to-noise ratio, but other parameters such as radiofrequency coil design and changing T1 and T2 times have to be considered (12,14,24). We therefore optimized our sequence protocol both at 1.5 and 3.0 T before the quantitative imaging studies. Because our final goal was the accurate determination of cartilage volume with a protocol suitable for clinical routine, all sequences had to be shorter than 6 minutes, in combination with a high in-plane spatial resolution (0.19 x 0.38 mm2/pixel) and minimal section thickness (1.5 mm). Results of previous studies (25) have shown that especially high spatial in-plane resolution is required for volume measurements of knee articular cartilage. By using such protocols, good accuracy and precision have been achieved with 1.5-T systems, but imaging time has been a limiting factor in those studies, ranging from about 11 to 31 minutes (6,7,10,17,22,26). At 3.0 T, we achieved about the same precision and slightly better accuracy in only about half (ie, 6 minutes) of the imaging time achieved in previous studies (7,10).

The precision of volumetric cartilage assessment was compared at 1.5 and 3.0 T in a previous study involving five healthy volunteers (14). In that study, no significant difference in precision was found. Of note, the fat-saturated SPGR sequence in that study had a longer acquisition time at 1.5 than at 3.0 T (11 minutes 47 seconds vs 9 minutes 41 seconds), and both sequences had high signal-to-noise ratios because relatively large voxels (0.66 x 0.66 x 1.5 mm3) were acquired with three signal averages. Because the 1.8-times-higher SNRE of cartilage at 3.0 T did not improve precision in that study, this benefit in SNRE may be used to decrease the imaging time.

In our study, comparison with direct volume measurements revealed that volumetric measurements at 3.0 T had significantly better accuracy than those at 1.5 T. This underlines the fact that—as recent studies (15,16,27) have demonstrated for bone and cartilage imaging—higher field strengths in musculoskeletal MR imaging improve diagnostic and quantitative accuracy. The comparison between 1.5 and 3.0 T was limited to the fat-saturated SPGR sequence in our study because the fast spin-echo sequence at 1.5 T with the required short imaging time and spatial resolution was insufficient to allow cartilage segmentation. A water excitation sequence was not available for 1.5-T imaging at the time our study was performed. At 3.0 T, however, only small differences were found between the water excitation and the fat-saturated SPGR sequences. The optimized water excitation sequence had a slightly higher signal-to-noise ratio and contrast-to-noise ratio than the fat-saturated SPGR sequence, but the qualitative delineation of adjacent cartilage surfaces was better with the fat-saturated SPGR sequence.

Although gradient-echo sequences yielded higher SNRE and CNRE values than did the fast spin-echo sequence in our study, other authors (15,19,28–30) found contrary results when they used lower spatial resolution or compared sequences of different spatial resolutions. Of note, none of those studies involved analysis of a fast spin-echo sequence with a section thickness smaller than 2 mm. One study (28) involved comparison of fat-saturated SPGR and fast spin-echo sequences that were performed with the same section thickness (2 mm). In that study, the author found that the fast spin-echo sequence had better CNRE for cartilage versus fluid, while the fat-saturated SPGR sequence had better CNRE for cartilage versus bone marrow. In summarizing these previous results performed at 1.5 T, the finding that many studies of quantification of articular cartilage (6,7,10,17,22,23) included gradient-echo but not fast spin-echo sequences is not unexpected.

Promising results for cartilage imaging with newer sequences, such as driven-equilibrium Fourier transform and steady-state free precession, have been published (12–14,19,29,31,32). These sequences have been shown to yield an increase of more than 30% in cartilage SNRE and 100% in CNRE compared with standard fat-saturated SPGR or fast spin-echo sequences (13,14). Nevertheless, an improvement in diagnostic or quantitative accuracy has not yet been shown. Wagner et al (33) reported a trend of lower reproducibility of cartilage morphometric measurements when they compared a balanced steady-state free precession with a water excitation sequence at 3.0 T. The authors explained this trend by citing differences in the contrast of synovial fluid versus cartilage.

In our study, a suitable balanced steady-state free precession sequence could not be performed because, for the spatial resolution required, the minimal echo time and repetition time were not short enough to avoid artifacts. However, in a clinical setting, with a larger field of view and smaller matrix (eg, a field of view of 13 cm and a matrix of 320 x 320) promising images without artifacts may be acquired. Whether the higher SNRE and CNRE of balanced steady-state free precession sequences can help improve accuracy in cartilage imaging has to be investigated in further studies.

Owing to the experimental setup involving skeletal immature porcine knees, potential limitations of this study have to be considered. The size and shape of the articular cartilage did not match exactly that of mature human knee joints. The shapes of the patellar and femoral cartilage were quite similar, but volumes were only half of those in human knees. Because at the tibia the articular cartilage covering the bone was very thin and had adjacent thicker segments of epiphyseal cartilage that were difficult to separate from the articular cartilage, the tibia was excluded from the analysis.

Practical application: Long acquisition times and low signal-to-noise ratios have limited the practical application of cartilage volumetric measurements at MR imaging thus far. With an imaging time of less than 6 minutes at 3.0 T, cartilage volumetric measurements are clinically feasible. With an accuracy error of 3%, changes of cartilage volume down to approximately 8% may be measured with statistical significance. Clinical studies are currently underway to investigate whether this accuracy is sufficient to monitor disease progression or cartilage therapy effects.

In conclusion, our study results have shown the potential of 3.0-T imaging in improving cartilage quantification. Compared with 1.5-T imaging, SNRE and CNRE were improved more than twofold, and the accuracy in determining cartilage volume was substantially increased.


   ADVANCES IN KNOWLEDGE
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
References
 


   ACKNOWLEDGMENTS
 
We thank Belinda S. Y. Li, PhD, of GE Healthcare for providing us with the water excitation SPGR sequence for 3.0 T.


   FOOTNOTES
 

Abbreviations: CNRE = effective contrast-to-noise ratio • SI = signal intensity • SNRE = effective signal-to-noise ratio • SPGR = spoiled gradient-recalled acquisition in the steady state

Authors stated no financial relationship to disclose.

Author contributions: Guarantors of integrity of entire study, C.J.R., S.M., T.M.L.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manuscript, all authors; literature research, J.S.B., S.J.K., C.J.R., T.M.L.; clinical studies, C.J.R., R.K.; experimental studies, J.S.B., S.J.K., C.J.R., R.K., J.C., S.M., T.M.L.; statistical analysis, J.S.B., C.J.R., E.O.; and manuscript editing, J.S.B., S.J.K., C.J.R., R.K., J.C., S.M., T.M.L.


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 

  1. Roos EM. Joint injury causes knee osteoarthritis in young adults. Curr Opin Rheumatol 2005;17(2):195–200.[CrossRef][Medline]
  2. Harris ED Jr. The bone and joint decade: a catalyst for progress. Arthritis Rheum 2001;44(9):1969–1970.[CrossRef][Medline]
  3. Gray ML, Eckstein F, Peterfy C, Dahlberg L, Kim YJ, Sorensen AG. Toward imaging biomarkers for osteoarthritis. Clin Orthop Relat Res 2004;427(suppl):S175–S181.
  4. Dunn TC, Lu Y, Jin H, Ries MD, Majumdar S. T2 relaxation time of cartilage at MR imaging: comparison with severity of knee osteoarthritis. Radiology 2004;232(2):592–598.[Abstract/Free Full Text]
  5. Peterfy CG. Imaging of the disease process. Curr Opin Rheumatol 2002;14(5):590–596.[CrossRef][Medline]
  6. Cicuttini FM, Jones G, Forbes A, Wluka AE. Rate of cartilage loss at 2 years predicts subsequent total knee arthroplasty: a prospective study. Ann Rheum Dis 2004;63(9):1124–1127.[Abstract/Free Full Text]
  7. Burgkart R, Glaser C, Hyhlik-Durr A, Englmeier KH, Reiser M, Eckstein F. Magnetic resonance imaging-based assessment of cartilage loss in severe osteoarthritis: accuracy, precision, and diagnostic value. Arthritis Rheum 2001;44(9):2072–2077.[CrossRef][Medline]
  8. Cicuttini F, Morris KF, Glisson M, Wluka AE. Slice thickness in the assessment of medial and lateral tibial cartilage volume and accuracy for the measurement of change in a longitudinal study. J Rheumatol 2004;31(12):2444–2448.[Medline]
  9. Eckstein F, Schnier M, Haubner M, et al. Accuracy of cartilage volume and thickness measurements with magnetic resonance imaging. Clin Orthop Relat Res 1998;352:137–148.
  10. Peterfy CG, van Dijke CF, Janzen DL, et al. Quantification of articular cartilage in the knee with pulsed saturation transfer subtraction and fat-suppressed MR imaging: optimization and validation. Radiology 1994;192(2):485–491.[Abstract/Free Full Text]
  11. Recht M, Bobic V, Burstein D, et al. Magnetic resonance imaging of articular cartilage. Clin Orthop Relat Res 2001;391(suppl):S379–S396.
  12. Gold GE, Han E, Stainsby J, Wright G, Brittain J, Beaulieu C. Musculoskeletal MRI at 3.0 T: relaxation times and image contrast. AJR Am J Roentgenol 2004;183(2):343–351.[Abstract/Free Full Text]
  13. Hargreaves BA, Gold GE, Beaulieu CF, Vasanawala SS, Nishimura DG, Pauly JM. Comparison of new sequences for high-resolution cartilage imaging. Magn Reson Med 2003;49(4):700–709.[CrossRef][Medline]
  14. Kornaat PR, Reeder SB, Koo S, et al. MR imaging of articular cartilage at 1.5T and 3.0T: comparison of SPGR and SSFP sequences. Osteoarthritis Cartilage 2005;13(4):338–344.[CrossRef][Medline]
  15. Fischbach F, Bruhn H, Unterhauser F, et al. Magnetic resonance imaging of hyaline cartilage defects at 1.5T and 3.0T: comparison of medium T2-weighted fast spin echo, T1-weighted two-dimensional and three-dimensional gradient echo pulse sequences. Acta Radiol 2005;46(1):67–73.[CrossRef][Medline]
  16. Masi JN, Sell CA, Phan C, et al. Cartilage MR imaging at 3.0 versus that at 1.5 T: preliminary results in a porcine model. Radiology 2005;236(1):140–150.[Abstract/Free Full Text]
  17. Eckstein F, Heudorfer L, Faber SC, Burgkart R, Englmeier KH, Reiser M. Long-term and resegmentation precision of quantitative cartilage MR imaging (qMRI). Osteoarthritis Cartilage 2002;10(12):922–928.[CrossRef][Medline]
  18. Hargreaves BA, Vasanawala SS, Pauly JM, Nishimura DG. Characterization and reduction of the transient response in steady-state MR imaging. Magn Reson Med 2001;46(1):149–158.[CrossRef][Medline]
  19. Kornaat PR, Doornbos J, van der Molen AJ, et al. Magnetic resonance imaging of knee cartilage using a water selective balanced steady-state free precession sequence. J Magn Reson Imaging 2004;20(5):850–856.[CrossRef][Medline]
  20. Ozhinsky E, Majumdar S, Carballido-Gamio J. 3D visualization of cartilage thickness in knee joint using bezier spline segmentation [abstr]. In: Proceedings of the Eleventh Meeting of the International Society for Magnetic Resonance in Medicine. Berkeley, Calif: International Society for Magnetic Resonance in Medicine, 2003; 776.
  21. Gluer CC, Blake G, Lu Y, Blunt BA, Jergas M, Genant HK. Accurate assessment of precision errors: how to measure the reproducibility of bone densitometry techniques. Osteoporos Int 1995;5(4):262–270.[CrossRef][Medline]
  22. Glaser C, Burgkart R, Kutschera A, Englmeier KH, Reiser M, Eckstein F. Femoro-tibial cartilage metrics from coronal MR image data: technique, test-retest reproducibility, and findings in osteoarthritis. Magn Reson Med 2003;50(6):1229–1236.[CrossRef][Medline]
  23. Raynauld JP, Martel-Pelletier J, Berthiaume MJ, et al. Quantitative magnetic resonance imaging evaluation of knee osteoarthritis progression over 2 years and correlation with clinical symptoms and radiologic changes. Arthritis Rheum 2004;50(2):476–487.[CrossRef][Medline]
  24. Collins CM, Smith MB. Signal-to-noise ratio and absorbed power as functions of main magnetic field strength, and definition of "90 degrees" RF pulse for the head in the birdcage coil. Magn Reson Med 2001;45(4):684–691.[CrossRef][Medline]
  25. Hardya PA, Newmark R, Liu YM, et al. The influence of the resolution and contrast on measuring the articular cartilage volume in magnetic resonance images. Magn Reson Imaging 2000;18(8):965–972.[CrossRef][Medline]
  26. Raynauld JP, Kauffmann C, Beaudoin G, et al. Reliability of a quantification imaging system using magnetic resonance images to measure cartilage thickness and volume in human normal and osteoarthritic knees. Osteoarthritis Cartilage 2003;11(5):351–360.[CrossRef][Medline]
  27. Phan CM, Matsuura M, Bauer JS, et al. Trabecular bone structure of the calcaneus: comparison of MR imaging at 3.0 and 1.5 T with micro-CT as the standard of reference. Radiology 2006;239(2):488–496.[Abstract/Free Full Text]
  28. Mohr A. The value of water-excitation 3D FLASH and fat-saturated PDw TSE MR imaging for detecting and grading articular cartilage lesions of the knee. Skeletal Radiol 2003;32(7):396–402.[CrossRef][Medline]
  29. 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(5):857–864.[CrossRef][Medline]
  30. Yoshioka H, Stevens K, Genovese M, Dillingham MF, Lang P. Articular cartilage of knee: normal patterns at MR imaging that mimic disease in healthy subjects and patients with osteoarthritis. Radiology 2004;231(1):31–38.[Abstract/Free Full Text]
  31. Gold GE, Fuller SE, Hargreaves BA, Stevens KJ, Beaulieu CF. Driven equilibrium magnetic resonance imaging of articular cartilage: initial clinical experience. J Magn Reson Imaging 2005;21(4):476–481.[CrossRef][Medline]
  32. Hargreaves BA, Gold GE, Lang PK, et al. MR imaging of articular cartilage using driven equilibrium. Magn Reson Med 1999;42(4):695–703.[CrossRef][Medline]
  33. Wagner S, Mendlik T, Wagner S, Horger W, Reiser M, Glaser C. Reproducibility of patellar cartilage volume and thickness at 3.0T: 3D-FLASH vs. 3D-TrueFISP [abstr]. In: Proceedings of the Eleventh Meeting of the International Society for Magnetic Resonance in Medicine. Berkeley, Calif: International Society for Magnetic Resonance in Medicine, 2003; 395.



This article has been cited by other articles:


Home page
 Am. J. Roentgenol.Home page
D. B. Siepmann, J. McGovern, J. H. Brittain, and S. B. Reeder
High-Resolution 3D Cartilage Imaging with IDEAL SPGR at 3 T
Am. J. Roentgenol., December 1, 2007; 189(6): 1510 - 1515.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
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 HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bauer, J. S.
Right arrow Articles by Link, T. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bauer, J. S.
Right arrow Articles by Link, T. M.

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