Published online before print March 30, 2007, 10.1148/radiol.2432060274
(Radiology 2007;243:475-482.)
© RSNA, 2007
Diagnosis of Articular Cartilage Abnormalities of the Knee: Prospective Clinical Evaluation of a 3D Water-Excitation True FISP Sequence1
Sylvain R. Duc, MD,
Peter Koch, MD,
Marius R. Schmid, MD,
Wilhelm Horger, PhD,
Juerg Hodler, MD, and
Christian W. A. Pfirrmann, MD
1 From the Departments of Radiology (S.R.D., M.R.S., J.H., C.W.A.P.) and Orthopedic Surgery (P.K.), University Hospital, Balgrist, Forchstrasse 340, CH-8008 Zurich, Switzerland; and Siemens Medical Solutions, Erlangen, Germany (W.H.). Received February 13, 2006; revision requested April 13; revision received May 19; accepted June 8; final version accepted September 12.
Address correspondence to S.R.D. (e-mail: Sylvain.Duc{at}hcuge.ch).
 |
ABSTRACT
|
|---|
Purpose: To prospectively evaluate the accuracy of three-dimensional (3D) water-excitation true fast imaging with steady-state precession (FISP) in the assessment of cartilage abnormalities of the knee, by using surgery as the reference standard.
Materials and Methods: The study was approved by the hospital institutional review board. Written informed consent was obtained from all patients. Twenty-nine patients (30 knees) with a mean age of 56 years (range, 18–86 years) were prospectively evaluated with a sagittal 3D true FISP magnetic resonance (MR) sequence. The mean interval between MR imaging and surgery was 1 day (range, 0–9 days). During surgery, the articular surfaces of the knee were evaluated by using a modified Noyes score. The MR images were evaluated by two blinded readers on two separate occasions. Diagnostic performance was evaluated by setting the cutoff for abnormality between grade 1 (intact cartilage surface) and grade 2 (cartilage defects). Statistical methods used included calculation of sensitivity, specificity, and accuracy, with 95% confidence intervals (Wilson score method) and calculation of 
values with standard errors.
Results: Overall sensitivity, specificity, and accuracy for the two readers and the two evaluations ranged from 56% to 66%, 78% to 93%, and 71% to 75%, respectively. Interobserver agreement was substantial for both the first ( = 0.73) and the second ( = 0.65) evaluation. Intraobserver agreement was almost perfect ( = 0.84) for reader 1 and moderate ( = 0.60) for reader 2.
Conclusion: The 3D water-excitation true FISP MR sequence allows assessment of the articular cartilage of the knee with moderate-to-high specificity and low-to-moderate sensitivity.
© RSNA, 2007
|
INTRODUCTION
|
|---|
After cardiovascular diseases, osteoarthritis is the second most common cause of disability for subjects older than 50 years (1). The diagnosis of early osteoarthritic changes may alter the therapeutic decision. Medication (2) and surgical interventions (3) potentially delay disease progression.
True fast imaging with steady-state precession (FISP) is an efficient sequence performed to obtain three-dimensional (3D) magnetic resonance (MR) images. The method belongs to the family of steady-state free precession (SSFP) sequences using balanced gradients (4). Other terms are balanced fast field echo imaging or fast imaging employing steady-state acquisition. True FISP is increasingly used for cardiac imaging, abdominal imaging, MR angiography, and real-time imaging (5–7).
Several investigators have described the use of balanced SSFP–type sequences in combination with different kinds of fat-suppression techniques for cartilage imaging, such as fluctuating-equilibrium MR or iterative decomposition of water and fat with echo asymmetry and least-squares estimation SSFP (8–11). Balanced SSFP images are characterized by a bright fluid signal intensity with preservation of the signal intensity from cartilage (11). With advances in MR gradient coils, shorter repetition times have become possible, and these allow one to use balanced SSFP without off-resonance artifacts caused by field inhomogeneity, such as banding artifacts. Thus, the purpose of our study was to prospectively evaluate the accuracy of 3D water-excitation true FISP MR imaging for the assessment of cartilage abnormalities of the knee, by using surgery as the reference standard.
|
MATERIALS AND METHODS
|
|---|
The study was approved by the hospital's institutional review board. Written informed consent was obtained from all patients. One author (W.H.) is an employee of Siemens Medical Solutions (Erlangen, Germany) and has a direct interest in the sequence under investigation. Authors who were not employees of Siemens Medical Solutions had control of the data and information presented in this article.
Patients
From February to July 2005, 29 consecutive patients (30 knees) were prospectively included in our study. There were 12 men and 17 women (mean age, 56 years; range, 18–86 years). Inclusion criteria were as follows: (a) patient age of 18 years and older and (b) referral for open surgery (nine patients) or arthroscopy (20 patients) of the knee. Exclusion criteria were as follows: surgery that did not allow one to review articular cartilage (such as tibial osteotomy), prior knee surgery, unavailability of the MR imager, and contraindications for MR imaging (Fig 1).
MR Imaging
All MR examinations were performed before surgery. The mean interval between MR imaging and surgery was 1 day (range, 0–9 days). The images were acquired with a 1.5-T MR imager (Avanto; Siemens Medical Solutions) that was equipped with 18 independent radiofrequency channels. An eight-channel receive-only extremity coil was used for this study.
The sagittal 3D true FISP sequence with water excitation was performed with the following imaging parameters: repetition time msec/echo time msec, 9.24/3.17; one signal acquired; field of view, 180 x 180 mm; matrix, 512 x 256; pixel size, 0.35 x 0.70 mm; no interpolation; partition thickness, 1.7 mm; partitions per slab, 48; flip angle, 28°; water excitation; and bandwidth, 199 Hz/pixel. Imaging time was 3 minutes 8 seconds.
The true FISP sequence is a gradient-echo sequence with balanced-gradient waveforms in all directions (read, phase, and section directions). This leads to a complete refocusing of the phase of stationary spins after repetition time so that a steady state of the magnetization will be established after several repetition time periods.
The value of longitudinal and transverse magnetization is preserved. Therefore, image contrast contains both T1 and T2 components. The image contrast with true FISP imaging is determined by means of the T2*/T1 ratio and depends on flip angle. With the parameters stated above, joint fluid is hyperintense, which is similar to the signal intensity seen with T2-weighted sequences. To suppress the fat signal, water excitation is used by applying a short spatial and frequency-selective 1-1 binomial radiofrequency-excitation pulse with monopolar section-select gradients to minimize chemical shift effects and phase evolution between the radiofrequency pulses of 180°.
Surgery
Four orthopedic surgeons (P.K., with 12 years of experience in orthopedic surgery, and three surgeons with 10, 7, and 7 years of experience) from the same knee surgery service were involved in the study. Before surgery, all patients underwent the clinically accepted work-up required for making a therapeutic decision, including the appropriate imaging. The informed consent form signed by all patients stated that the true FISP sequence performed for the study does not yet have a well-defined role in cartilage imaging and that the evaluation of the true FISP MR images would not be performed until after the surgical intervention. Accordingly, the surgeons were blinded to the MR imaging findings and proceeded with therapeutic decisions, per the routine, on the basis of the findings at clinically accepted work-up.
For cartilage evaluation during surgery, a form with a schematic drawing of the articular surfaces of the knee was used. The orthopedic surgeon marked the location and the degree of cartilage damage on this drawing. The medial and lateral femoral condyles, the medial and lateral surfaces of the tibial plateau, the femoral groove, and the patella were separately evaluated. A modified Noyes grading system was used (12), as follows: grade 0, normal cartilage; grade 1, cartilage softening without surface abnormalities; grade 2, erosions or fissures with a depth of not more than 50% of the cartilage thickness; grade 3, erosions or fissures with a depth of greater than 50% but less than 100%; and grade 4, cartilage lesion to the subchondral bone. If different grades of cartilage lesions were present within the same articular surface, the worst lesion was used for grading. Surgical grading served as the standard of reference. Two patellar surfaces were not properly visible during surgery and were not evaluated.
Evaluation of MR Images
After completion of the data acquisition in all patients, the MR images were evaluated at once by two radiologists (J.H., M.R.S.) with 18 and 5 years of experience in musculoskeletal MR imaging. The readers were blinded with regard to clinical and intraoperative data. The evaluation was performed independently and was repeated after 3 weeks. The grading of the cartilage lesions was performed in correspondence with surgical grading, with minor adaptations required according to the MR appearance of abnormalities: grade 0 (Fig 2), normal cartilage (homogeneous signal intensity, intact cartilage surface and thickness); grade 1 (Fig 3), abnormal signal intensity (focal alteration of the cartilage signal intensity) but normal surface (13,14); grade 2 (Fig 4), superficial fraying, erosion, or ulceration, with a depth of not more than 50% of cartilage thickness; grade 3 (Fig 5), defect of more than 50% but less than 100% of cartilage thickness; and grade 4, full-thickness cartilage damage (Fig 6).
View larger version (129K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2: Normal (grade 0) cartilage. Sagittal true FISP MR image (9.24/3.17, 28° flip angle) shows intact femoral and tibial cartilage (upper and lower arrowheads, respectively). F = femoral condyle, T = tibial plateau.
|
|
View larger version (130K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3: Grade 1 cartilage lesion. Sagittal true FISP MR image (9.24/3.17, 28° flip angle) shows localized cartilage hyperintensity of the trochlear (black arrowheads) and patellar (white arrowheads) articular surfaces. F = femoral condyle, P = patella.
|
|
View larger version (145K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4: Grade 2 cartilage lesion. Sagittal true FISP MR image (9.24/3.17, 28° flip angle) shows localized thinning (arrowheads) of the condylar cartilage that involves less than 50% of the cartilage thickness. F = femoral condyle, T = tibial plateau.
|
|
View larger version (105K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5a: Grade 3 cartilage lesion. Sagittal true FISP MR images (9.24/3.17, 28° flip angle). F = femoral condyle, T = tibial plateau. (a) Image shows widespread thinning of the cartilage of the medial tibial plateau (arrowheads) that involves more than 50% of the cartilage thickness. (b) For comparison, image of normal cartilage at the lateral tibial plateau (arrowheads) in the same patient demonstrates the thickness of nondamaged cartilage.
|
|
View larger version (91K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5b: Grade 3 cartilage lesion. Sagittal true FISP MR images (9.24/3.17, 28° flip angle). F = femoral condyle, T = tibial plateau. (a) Image shows widespread thinning of the cartilage of the medial tibial plateau (arrowheads) that involves more than 50% of the cartilage thickness. (b) For comparison, image of normal cartilage at the lateral tibial plateau (arrowheads) in the same patient demonstrates the thickness of nondamaged cartilage.
|
|
View larger version (115K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6: Grade 4 cartilage lesion. Sagittal true FISP MR image (9.24/3.17, 28° flip angle) shows extensive cartilage lesion of the femoral and tibial cartilage (arrowheads) that involves the entire cartilage thickness to the subchondral bone. Arrow = hyperintense subchondral bone marrow, F = femoral condyle, T = tibial plateau.
|
|
Error analysis was performed for the first blinded MR evaluation. Two radiologists (C.W.A.P., S.R.D., with 7 and 2 years of experience in musculoskeletal MR imaging, respectively) who were not involved in the blinded readings analyzed all studies with false-positive or false-negative results in consensus. The most probable reason for the diagnostic error was recorded.
Statistical Analysis
Diagnostic performance for MR imaging was evaluated by setting the cutoff for cartilage abnormality between grades 0 and 1 (intact cartilage surface) and grades 2, 3, and 4 (cartilage defects). Sensitivity, specificity, and accuracy of 3D water-excitation true FISP MR imaging were calculated for both readers. The 95% confidence intervals for sensitivity and specificity were determined by using the Wilson score method (15). Inter- and intraobserver agreement values were calculated by using commercially available statistical software (SPSS version 11, 2001; SPSS, Chicago, Ill). The standard errors for the values were calculated. The grading of inter- and intraobserver agreement was performed according to the recommendations of Landis and Koch (16). Significance was set at P < .05.
|
RESULTS
|
|---|
Diagnostic Performance
For the first evaluation, for reader 1 and reader 2, respectively, overall sensitivity (Table 1) was 61% and 62%, overall specificity was 91% and 85%, and overall accuracy was 75% and 73%. Sensitivity was best for the lateral tibial plateau (80% and 80%, respectively) and worst for the medial femoral condyle (45% and 40%, respectively). Specificity was best for the medial femoral condyle (100% and 100%, respectively) and the medial tibial plateau (100% and 92%, respectively) and worst for the patella (64% and 73%, respectively).
In comparison with findings at the first evaluation, findings for reader l at the second evaluation (Table 2) were slightly more specific (overall specificity for the knee joint, 93% vs 91% [second session vs first session]) and less sensitive (overall sensitivity, 56% vs 61%). In contrast with the findings of reader 1, the findings of reader 2 at the second session became slightly more sensitive (overall sensitivity, 66% vs 62% [second session vs first session]) and less specific (overall specificity, 78% vs 85%). For both readers, overall accuracy (for the second session and first session, respectively, 74% and 75% for reader 1 and 71% and 73% for reader 2) remained approximately the same.
Interobserver and Intraobserver Agreement
The overall value for interobserver agreement was substantial for the first session ( = 0.73) and for the second session ( = 0.65) (Table 3). Interobserver values for the different articular surfaces varied between moderate ( = 0.45 for the patella) and almost perfect ( = 0.92 for the lateral tibial plateau).
Overall intraobserver agreement was almost perfect ( = 0.84) for reader 1 and moderate ( = 0.60) for reader 2 (Table 2). Intraobserver values for the individual articular surfaces were between almost perfect ( = 1.00 for the medial femoral condyle) and moderate ( = 0.43 for the medial tibial plateau). All inter- and intraobserver values were statistically significant (P < .05).
Cross-Table Analysis
For reader 1, the MR and surgical grades matched precisely for 90 of 178 (50.6%) surfaces (Table 4). For 140 of 178 (78.7%) surfaces, grades matched within one degree. Overdiagnosis of cartilage lesions (18 of 178 surfaces, 10.1%) was less common than was underdiagnosis (70 of 178 surfaces, 39.3%). For reader 2, there was precise matching for 86 of 178 (48.3%) surfaces. For 138 of 178 (77.5%) surfaces, the grades matched within one degree. Overdiagnosis (26 of 178 surfaces, 14.6%) was less common than was underdiagnosis (66 of 178 surfaces, 37.1%).
Error Analysis
Error analysis is shown in Table 5. The following probable reasons for a false-positive diagnosis were found: thin but normal cartilage, lesion at periphery of cartilage surface, pulsation artifact, partial volume effect, or an unexplained reason (when the lesion was diagnosed both during the blinded evaluation and the retrospective review). The most probable reasons for a false-negative diagnosis were as follows: thin cartilage, lesion at periphery of cartilage surface, partial volume effect, pulsation artifact, reviewer error, superficial lesion or lesion with small transverse diameter, or an unexplained reason. The majority of the false-positive diagnoses were due to the location of the lesion at the periphery of the cartilage surfaces. The majority of false-negative findings were not seen during retrospective review, even with knowledge of their exact location.
|
DISCUSSION
|
|---|
Articular cartilage may be assessed with standard MR sequences such as intermediate- or T2-weighted spin-echo sequences without (17–20) or with (21,22) fat suppression. Other sequences are employed specifically for cartilage imaging, including fast low-angle shot or spoiled gradient-recalled acquisition in the steady state (SPGR) (23–27), double-echo steady state without (28) and with fat suppression or with water excitation (29). Multiple-echo data image combination (30) is another sequence employed for cartilage imaging. The sensitivities and specificities reported for such sequences have been quite variable (23–30), probably depending on patient population, extent of cartilage abnormality, magnet and coil specifications, and details of sequence design. There remains an interest in a fast, routinely applicable, and reliable sequence with good lesion conspicuity.
The suppression of the fat signal increases the contrast between fat and other tissues and thus affects the dynamic range of the image (31). The most widely used method for fat suppression in musculoskeletal imaging is frequency-selective fat saturation (21–23,26). Water excitation by means of a spectral spatial pulse may alternatively be employed. With this technique, only water is excited by using section-selective composite pulses. Lipid spins are left in equilibrium and do not produce signal (32). Short acquisition times and reduced chemical-shift artifacts are major advantages of water excitation (32).
There have been several approaches to achieve fat suppression with balanced SSFP imaging. These include fluctuating-equilibrium MR (9), linear combinations of SSFP (8), intermittent fat-suppressed SSFP (33), Dixon imaging (10), and phase-sensitive SSFP (11). All these techniques have specific advantages and disadvantages, but all share the goal of elimination of the bright lipid signal intensity seen with balanced SSFP sequences.
The image acquisition time with the 3D true FISP sequence with water excitation as used in our investigation was 3 minutes 8 seconds. This acquisition time compares favorably with those reported for dedicated, commonly used cartilage gradient-echo sequences, such as fat-suppressed SPGR (23–25,34), fast low-angle shot (26), or double-echo steady state (28) sequences, with acquisition times of 10 minutes 18 seconds, 9 minutes 40 seconds, and 4 minutes 8 seconds, respectively.
When looking at each articular surface separately, our sagittal 3D water-excitation true FISP sequence performed well with regard to specificity (46%–100%) and accuracy (57%–90%) but was less sensitive (40%–82%) in comparison with previously published data. The results of imaging with standard non–fat-suppressed sequences for the detection of cartilage defects are variable. Reported sensitivities, specificities, and accuracies with T2-weighted sequences are 40%–48%, 58%–100%, and 52%–94%, respectively (22,27). For intermediate-weighted sequences, the corresponding values are 28%–74%, 79%–91%, and 50%–86%, respectively (19,27). Although some of these sequences enable good contrast between articular cartilage and adjacent joint fluid, capsule, muscle, and fat, fat suppression may improve diagnostic performance. Bredella et al (22) evaluated a T2-weighted fast spin-echo sequence with frequency-selective fat saturation for the diagnosis of cartilage lesions of the knee. They reported a sensitivity of 59% for transverse images and 61% for coronal images and a specificity of 99% for both imaging planes. Sensitivity and specificity of fat-saturated SPGR sequences are superior to those of standard MR sequences (24,25) and to those of SPGR sequences without fat suppression (27). Yoshioka et al (23) evaluated sagittal fat-saturated intermediate-weighted and T2-weighted fast spin-echo MR sequences and a fat-saturated 3D SPGR MR sequence. In their study, the SPGR sequence was more specific than were fast spin-echo sequences (85% vs 68%), with a comparable sensitivity (97% for the SPGR sequence, 100% for the fast spin-echo sequence).
The reliable differentiation of early chondral lesions with an intact cartilage surface (grade 1) or with superficial erosions (grade 2) from intact cartilage is known to be difficult (35) as was also demonstrated with our results. Because of the superior spatial resolution and signal-to-noise ratio, 3.0-T magnets appear to improve the diagnostic performance of MR imaging in early chondral lesions (36). Use of thinner partitions is another possible solution. Eckstein et al (37) demonstrated that at 3.0 T, the precision error for cartilage volume and thickness measurement was smaller with 1.0-mm partitions in comparison with 1.5-mm partitions. However, findings of that study are not directly comparable with ours because the focus of that study was cartilage volume, thickness, and surface area assessment and not lesion detection.
In addition to being attributed to superficial and small lesions, false-negative diagnoses at MR imaging were also commonly attributed to partial volume effects. The use of additional planes might have reduced the number of the false-negative findings at the periphery of cartilage surfaces.
There may be a certain inclusion bias as a limitation of our study because cartilage lesions may be more pronounced in patients undergoing surgery than in the general population. This bias cannot be avoided when surgery has to serve as the standard of reference but should be kept in mind when interpreting the study results.
The results of our study show a decreased sensitivity in detection of chondral lesions at the medial femoral condyle and tibial plateau. This problem could at least partially be ameliorated by using additional imaging planes and increasing the sample size. Therefore, further testing of this sequence with larger sample sizes and additional imaging planes would be helpful before its clinical use.
Gradient-echo sequences have many advantages in cartilage MR imaging but also have some limitations. The 3D water-excitation true FISP sequence used in this study demonstrates subchondral bone marrow changes but not as clearly as do fat-suppressed long–repetition time spin-echo sequences. Such abnormalities are an indirect sign of cartilage lesions and may improve the diagnostic performance of MR imaging. Gradient-echo sequences are more sensitive to susceptibility artifacts—notably seen in knees after surgery. On the other hand, gradient-echo techniques tend to provide excellent contrast between cartilage and joint fluid. Three-dimensional acquisition is easily performed with gradient-echo sequences and thus allows one to obtain thin and continuous sections.
In conclusion, the 3D water-excitation true FISP MR sequence allows assessment of the articular cartilage of the knee with moderate-to-high specificity and low-to-moderate sensitivity.
|
ADVANCE IN KNOWLEDGE
|
|---|
- The three-dimensional water-excitation true fast imaging with steady-state precession MR sequence has a sensitivity of 56%–66%, a specificity of 78%–93%, an accuracy of 71%–75%, and a substantial interobserver agreement ( = 0.65–0.73) in the diagnosis of cartilage defects.
|
FOOTNOTES
|
|---|
Abbreviations: FISP = fast imaging with steady-state precession • SPGR = spoiled gradient-recalled acquisition in the steady state • SSFP = steady-state free precession • 3D = three-dimensional
See Materials and Methods for pertinent disclosures.
Author contributions: Guarantor of integrity of entire study, S.R.D.; 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, S.R.D., J.H., C.W.A.P.; clinical studies, all authors; statistical analysis, S.R.D., J.H., C.W.A.P.; and manuscript editing, all authors
|
References
|
|---|
- Peyron JG. Epidemiological aspects of osteoarthritis. Scand J Rheumatol Suppl 1988;77:29–33.[Medline]
- Pendleton A, Arden N, Dougados M, et al. EULAR recommendations for the management of knee osteoarthritis: report of a task force of the standing committee for international clinical studies including therapeutic trials (ESCISIT). Ann Rheum Dis 2000;59:936–944.[Abstract/Free Full Text]
- Jackson DW, Scheer MJ, Simon TM. Cartilage substitutes: overview of basic science and treatment options. J Am Acad Orthop Surg 2001;9:37–52.[Abstract/Free Full Text]
- Scheffler K, Lehnhardt S. Principles and applications of balanced SSFP techniques. Eur Radiol 2003;13:2409–2418.[CrossRef][Medline]
- Ajaj W, Goehde SC, Schneemann H, Ruehm SG, Debatin JF, Lauenstein TC. Dose optimization of mannitol solution for small bowel distension in MRI. J Magn Reson Imaging 2004;20:648–653.[CrossRef][Medline]
- Fuchs F, Laub G, Othomo K. TrueFISP: technical considerations and cardiovascular applications. Eur J Radiol 2003;46:28–32.[CrossRef][Medline]
- Zenge MO, Ladd ME, Vogt FM, Brauck K, Barkhausen J, Quick HH. Whole-body magnetic resonance imaging featuring moving table continuous data acquisition with high-precision position feedback. Magn Reson Med 2005;54:707–711.[CrossRef][Medline]
- Vasanawala SS, Pauly JM, Nishimura DG. Linear combination steady-state free precession MRI. Magn Reson Med 2000;43:82–90.[CrossRef][Medline]
- Vasanawala SS, Pauly JM, Nishimura DG. Fluctuating equilibrium MRI. Magn Reson Med 1999;42:876–883.[CrossRef][Medline]
- Reeder SB, Pelc NJ, Alley MT, Gold GE. Rapid MR imaging of articular cartilage with steady-state free precession and multipoint fat-water separation. AJR Am J Roentgenol 2003;180:357–362.[Abstract/Free Full Text]
- Vasanawala SS, Hargreaves BA, Pauly JM, Nishimura DG, Beaulieu CF, Gold GE. Rapid musculoskeletal MRI with phase-sensitive steady-state free precession: comparison with routine knee MRI. AJR Am J Roentgenol 2005;184:1450–1455.[Abstract/Free Full Text]
- Noyes FR, Stabler CL. A system for grading articular cartilage lesions at arthroscopy. Am J Sports Med 1989;17:505–513.[Abstract/Free Full Text]
- Mintz DN, Hooper T, Connell D, Buly R, Padgett DE, Potter HG. Magnetic resonance imaging of the hip: detection of labral and chondral abnormalities using noncontrast imaging. Arthroscopy 2005;21:385–393.[Medline]
- Gahunia HK, Babyn P, Lemaire C, Kessler MJ, Pritzker KP. Osteoarthritis staging: comparison between magnetic resonance imaging, gross pathology and histopathology in the rhesus macaque. Osteoarthritis Cartilage 1995;3:169–180.[CrossRef][Medline]
- Newcombe RG. Two-sided confidence intervals for the single proportion: comparison of seven methods. Stat Med 1998;17:857–872.[CrossRef][Medline]
- Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33:159–174.[CrossRef][Medline]
- Brown TR, Quinn SF. Evaluation of chondromalacia of the patellofemoral compartment with axial magnetic resonance imaging. Skeletal Radiol 1993;22:325–328.[Medline]
- Yulish BS, Montanez J, Goodfellow DB, Bryan PJ, Mulopulos GP, Modic MT. Chondromalacia patellae: assessment with MR imaging. Radiology 1987;164:763–766.[Abstract/Free Full Text]
- Sonin AH, Pensy RA, Mulligan ME, Hatem S. Grading articular cartilage of the knee using fast spin-echo proton density-weighted MR imaging without fat suppression. AJR Am J Roentgenol 2002;179:1159–1166.[Abstract/Free Full Text]
- Hayes CW, Sawyer RW, Conway WF. Patellar cartilage lesions: in vitro detection and staging with MR imaging and pathologic correlation. Radiology 1990;176:479–483.[Abstract/Free Full Text]
- Rose PM, Demlow TA, Szumowski J, Quinn SF. Chondromalacia patellae: fat-suppressed MR imaging. Radiology 1994;193:437–440.[Abstract/Free Full Text]
- Bredella MA, Tirman PF, Peterfy CG, et al. Accuracy of T2-weighted fast spin-echo MR imaging with fat saturation in detecting cartilage defects in the knee: comparison with arthroscopy in 130 patients. AJR Am J Roentgenol 1999;172:1073–1080.[Abstract/Free Full Text]
- 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]
- Disler DG, McCauley TR, Wirth CR, Fuchs MD. Detection of knee hyaline cartilage defects using fat-suppressed three-dimensional spoiled gradient-echo MR imaging: comparison with standard MR imaging and correlation with arthroscopy. AJR Am J Roentgenol 1995;165:377–382.[Abstract/Free Full Text]
- Disler DG, McCauley TR, Kelman CG, et al. Fat-suppressed three-dimensional spoiled gradient-echo MR imaging of hyaline cartilage defects in the knee: comparison with standard MR imaging and arthroscopy. AJR Am J Roentgenol 1996;167:127–132.[Abstract/Free Full Text]
- Recht MP, Piraino DW, Paletta GA, Schils JP, Belhobek GH. Accuracy of fat-suppressed three-dimensional spoiled gradient-echo FLASH MR imaging in the detection of patellofemoral articular cartilage abnormalities. Radiology 1996;198:209–212.[Abstract/Free Full Text]
- Recht MP, Kramer J, Marcelis S, et al. Abnormalities of articular cartilage in the knee: analysis of available MR techniques. Radiology 1993;187:473–478.[Abstract/Free Full Text]
- Ruehm S, Zanetti M, Romero J, Hodler J. MRI of patellar articular cartilage: evaluation of an optimized gradient echo sequence (3D-DESS). J Magn Reson Imaging 1998;8:1246–1251.[Medline]
- Knuesel PR, Pfirrmann CW, Noetzli HP, et al. MR arthrography of the hip: diagnostic performance of a dedicated water-excitation 3D double-echo steady-state sequence to detect cartilage lesions. AJR Am J Roentgenol 2004;183:1729–1735.[Abstract/Free Full Text]
- 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]
- Gold GE, Hargreaves BA, Reeder SB, Vasanawala SS, Beaulieu CF. Controversies in protocol selection in the imaging of articular cartilage. Semin Musculoskelet Radiol 2005;9:161–172.[CrossRef][Medline]
- Hauger O, Dumont E, Chateil JF, Moinard M, Diard F. Water excitation as an alternative to fat saturation in MR imaging: preliminary results in musculoskeletal imaging. Radiology 2002;224:657–663.[Abstract/Free Full Text]
- Scheffler K, Heid O, Hennig J. Magnetization preparation during the steady state: fat-saturated 3D TrueFISP. Magn Reson Med 2001;45:1075–1080.[CrossRef][Medline]
- Yoshioka H, Alley M, Steines D, et al. Imaging of the articular cartilage in osteoarthritis of the knee joint: 3D spatial-spectral spoiled gradient-echo vs. fat-suppressed 3D spoiled gradient-echo MR imaging. J Magn Reson Imaging 2003;18:66–71.
- 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]
- Link TM, Sell CA, Masi JN, et al. 3.0 vs 1.5T MRI in the detection of focal cartilage pathology: ROC analysis in an experimental model. Osteoarthritis Cartilage 2006;14:63–70.[CrossRef][Medline]
- Eckstein F, Charles HC, Buck RJ, et al. Accuracy and precision of quantitative assessment of cartilage morphology by magnetic resonance imaging at 3.0T. Arthritis Rheum 2005;52:3132–3136.[CrossRef][Medline]
TOP
ABSTRACT
INTRODUCTION
