Published online before print June 26, 2006, 10.1148/radiol.2401050077
(Radiology 2006;240:482-488.)
© RSNA, 2006
Cartilaginous Defects of the Femorotibial Joint: Accuracy of Coronal Short Inversion Time Inversion-Recovery MR Sequence1
Karl-Peter Jungius, MD,
Marius R. Schmid, MD,
Marco Zanetti, MD,
Juerg Hodler, MD,
Peter Koch, MD and
Christian W. A. Pfirrmann, MD
1 From the Departments of Radiology (K.P.J., M.R.S., M.Z., J.H., C.W.A.P.) and Orthopedic Surgery (P.K.), University Hospital Balgrist, Forchstrasse 340, CH-8008 Zurich, Switzerland. Received January 17, 2005; revision requested March 22; revision received May 19; accepted June 20; final version accepted September 8.
Address correspondence to M.R.S.
 |
ABSTRACT
|
|---|
Purpose: To retrospectively assess the diagnostic performance of the short inversion time inversion-recovery (STIR) magnetic resonance (MR) sequence for depiction and classification of articular cartilaginous lesions in femorotibial joint, with arthroscopy as reference standard.
Materials and Methods: Institutional review board did not require approval and informed consent for review of patients' records or images. All patients (and parents of underage patients) agreed to use of their data. Two musculoskeletal radiologists independently analyzed femorotibial cartilage on coronal STIR images from 84 knee MR examinations in 83 patients (48 male patients [49 knees], 35 female patients; mean age, 39.5 years). Slightly modified Outerbridge classification was used: grade 0, normal cartilage; grade 1, softening or swelling; grade 2, partial-thickness defect; grade 3, fissuring to the level of the subchondral bone; and grade 4, exposed subchondral bone. Arthroscopy performed within 15 weeks was the standard of reference. Classification for arthroscopy differed only in definition of grade 1 (softening or swelling of cartilage). Sensitivity, specificity, accuracy, positive and negative predictive values, and weighted 
values were calculated to assess interobserver reliability.
Results: At arthroscopy, 212 (63%) of 336 surfaces were classified as grade 0 (normal); 37 (11%), as grade 1 abnormalities; 30 (9%), as grade 2 lesions; 25 (7%), as grade 3 lesions; and 32 (10%), as grade 4 lesions. Grades 0 and 1 were considered normal; grades 2–4, as abnormal. For detection of contour defects of the cartilaginous surface, coronal STIR MR imaging had sensitivity values of 77% and 76%, specificity values of 96% and 89%, accuracy values of 91% and 85%, positive predictive values of 86% and 70%, and negative predictive values of 92% and 91% for readers 1 and 2, respectively. Weighted value was 0.63.
Conclusion: Contour defects of femorotibial cartilage can be detected with reasonable accuracy with routine STIR sequence.
© RSNA, 2006
|
INTRODUCTION
|
|---|
Standard techniques for imaging the morphologic features of cartilage include fast spin-echo and spoiled gradient-echo imaging (1,2). Other methods for faster or higher-spatial-resolution morphologic imaging include techniques based on steady-state free precession imaging (1,3). Defects of the articular cartilage are common, and their description is an important part of every report for magnetic resonance (MR) imaging examinations of the knee. The use of a large number of dedicated sequences has been investigated for the improvement of the diagnosis of cartilaginous abnormalities. Typical examples are fat-saturated spoiled gradient-recalled acquisition in the steady state (SPGR) (4) and double-echo steady-state sequences (5). For routine imaging, standardized protocols with a limited number of sequences are commonly applied to reduce imaging time, obtain consistent imaging quality and image characteristics, and include all relevant potential diagnoses. In many instances, patients who are suspected of having meniscal abnormalities on the basis of clinical findings do have cartilaginous defects at arthroscopy (6).
Our standard MR imaging protocol includes a coronal short inversion time inversion-recovery (STIR) sequence, which is useful for the detection of bone bruises, fractures, avascular necrosis, and abnormalities of the collateral ligaments. According to our observations in daily practice, cartilaginous defects are depicted with the STIR sequence quite consistently, although this sequence usually is not included in the list of dedicated MR sequences for imaging of cartilage.
The purpose of our study was to retrospectively assess the diagnostic performance of the STIR sequence for depiction and classification of articular cartilaginous lesions in the femorotibial joint, with arthroscopy as the reference standard.
|
MATERIALS AND METHODS
|
|---|
Patients
All 2115 patients who underwent MR examinations of the knee with a single imager between July 2001 and December 2003 were identified from the hospital's electronic patient charts. For 86 patients, arthroscopy was performed in 87 knees at our hospital within 15 weeks after the MR imaging examination (mean interval in the study group, 8 weeks). The study group included only patients with an arthroscopic and open surgical assessment and a description of the articular cartilage in the surgical report; 21 patients who previously had undergone surgery were excluded. All patients underwent arthroscopy of the knee because they were suspected of having internal derangement. The procedure was performed by one of three experienced orthopedic surgeons; two of them had 15 and 7 years of experience, and one (P.K.) had 9 years of experience in arthroscopy of the knee. All three surgeons used the Outerbridge classification (7–10) to assess articular cartilage. The Outerbridge classification is the standard system for classifying articular cartilaginous findings at arthroscopy in our orthopedic surgery department.
Three knees in three patients had to be excluded because of incomplete description of the cartilage in the arthroscopy report. Thus, our study included 84 knees in 83 patients; 48 patients were male (mean age, 44.8 years; range, 16–68 years), and 35 patients were female (mean age, 32.3 years; range, 13–58 years). In one man, both knees were examined almost simultaneously. The differences among the patients in regard to age and sex were not statistically significant. The indications for surgery included meniscal lesions (26 knees), cruciate ligamentous tears (16 knees), meniscal lesions and cruciate ligamentous tears (five knees), meniscal lesions and cartilaginous defects (eight knees), cruciate ligamentous tears and cartilaginous lesions (three knees), only cartilaginous lesions (11 knees), diagnostic arthroscopic findings (seven knees), plica syndrome (three knees), and other indications (five knees).
Our institutional review board did not require approval and informed consent for the review of patients' records or images. Patients' rights, however, are protected by a law that requires that patients be informed of the possibility that their charts and imaging studies may be reviewed for scientific purposes and grants them the opportunity to forbid such use of their data if they wish. All patients included in our study had agreed to the use of their data. For underage patients, parents had given approval.
Imaging Protocol
All examinations were performed with a 1.5-T MR imaging unit (Magnetom Symphony; Siemens Medical Solutions, Erlangen, Germany) and a dedicated send-receive extremity coil. A fast spin-echo STIR sequence (5550/35/180 [repetition time msec/echo time msec/inversion time msec]; turbo factor, seven; section thickness, 3 mm; intersection gap, 0.7 mm; field of view, 170 mm; matrix size, 512 x 512; number of signals acquired, one; and imaging time, 3 minutes 5 seconds) was performed in the coronal plane.
Other sequences included in our routine protocol but not evaluated in this investigation were sagittal intermediate-weighted (2700/15 [repetition time msec/echo time msec]) and T2-weighted turbo fast-echo (3500/88) sequences, a coronal T1-weighted spin-echo sequence (450/14), and a transverse T2-weighted multiecho data image combination sequence (468/26; flip angle, 30°).
Image Analysis and Classification System
In the first part of the readout, all images from 84 MR imaging examinations were independently reviewed on a picture archiving and communication system workstation (Image Devices, Idstein, Germany) by two radiologists (C.W.A.P. and M.R.S., with 6 and 5 years of experience in musculoskeletal MR imaging, respectively). Both readers were unaware of the reports of arthroscopic findings or other clinical data. Only the images obtained with the coronal STIR sequence were analyzed; images obtained with other sequences from the routine MR imaging protocol were not made available for evaluation. At arthroscopy, the Outerbridge classification (7–10) was used to classify cartilaginous defects as follows: Grade 0 indicated normal cartilage; grade 1, softening or swelling of the cartilage; grade 2, partial-thickness defect of the cartilage; grade 3, fissuring of the cartilage to the level of the subchondral bone; and grade 4, exposed subchondral bone. For the MR imaging evaluation, the classification system was slightly modified. Because softening of cartilage cannot be seen directly on MR images, grade 1 lesions on STIR images were defined as signal intensity alterations (higher or lower signal intensity), with an intact surface of the articular cartilage compared with the surrounding normal cartilage. For all other grades, identical definitions were used for findings at MR imaging and at arthroscopy.
The femorotibial cartilaginous surfaces were subdivided into four areas: lateral femoral, lateral tibial, medial femoral, and medial tibial cartilaginous areas. Altogether, 336 surfaces were evaluated by each of the two readers who were blinded to clinical data and findings in surgical reports. Since the STIR sequence was performed only in the coronal plane, patellar cartilage was not assessed.
At arthroscopy, grades were assigned to cartilaginous findings according to the five-grade scale (grades 0–4) described previously. In knees with multiple cartilaginous defects in one area (documented in surgical reports and on MR images), the score of the worst cartilaginous abnormality was used for this study.
In the second part of the MR readout, a third radiologist (M.Z., with 12 years of experience in musculoskeletal MR imaging) who was not involved in the assessment of the cartilage analyzed all images from MR examinations for the presence of hyperintense subchondral bone marrow. This third reader was blinded to surgical, clinical, and imaging data. Only areas with hyperintense bone marrow that directly bordered the associated cartilaginous defect were rated as positive. If normal bone marrow was identified between the cartilaginous defect and the bone marrow signal intensity alteration, the hyperintense bone marrow area was rated as negative.
Statistical Analysis
Sensitivity, specificity, and accuracy values for detection of contour defects of the articular cartilage for both readers of the first MR readout were calculated. Because normal cartilage (grade 0) and signal intensity alterations within the cartilage without any contour defects (grade 1 lesions) are almost indistinguishable, the threshold value for the analysis was set between grades 0–1 (normal cartilage and signal intensity alterations or swelling of the cartilage classified as negative) and grades 2–4 (contour defects of the cartilage classified as positive) for both arthroscopy and MR imaging. Sensitivity, specificity, diagnostic accuracy, positive predictive value, and negative predictive value were calculated separately for each reader. A value for interobserver variability was calculated for the five-point classification data. Values of 0.41–0.60 were considered to represent moderate agreement; values of 0.61–0.80, substantial agreement; and values of 0.81–1.00, almost perfect or perfect agreement (11).
The presence of hyperintense subchondral bone marrow was compared with the presence and the grade of cartilaginous lesions (comparison of the second part of the MR readout with the first part). For statistical analysis, software (SPSS 10; SPSS, Chicago, Ill) was used. Data clustering was evaluated with other software (Stata, version 8.2 for Macintosh; Stata, College Station, Tex). No cluster effect was found.
|
RESULTS
|
|---|
Classification at Arthroscopy
Of 336 classified surfaces, 212 (63%) were normal (grade 0) and 124 (37%) were abnormal during arthroscopy (Table 1). Abnormal surfaces were classified as follows: 37 surfaces (11%) were grade 1; 30 surfaces (9%), grade 2; 25 surfaces (7%), grade 3; and 32 surfaces (10%), grade 4. Contour defects of the cartilage (grades 2–4) were localized as follows: 14 (16%) of 87 in the lateral femoral compartment, 12 (14%) of 87 in the lateral tibial compartment, 37 (43%) of 87 in the medial femoral compartment, and 24 (28%) of 87 in the medial tibial compartment.
Classification of Articular Cartilage
In the first part of the MR readout, 96% (204 of 212) of normal surfaces, were correctly classified by reader 1, and 78% (165 of 212) of normal surfaces were correctly classified by reader 2. Classification was identical to that of the surgical reports in regard to 73% (244 of 336) of surfaces for reader 1 and in regard to 64% (215 of 336) of surfaces for reader 2 (Table 2). Differences of one grade between the classification in arthroscopic surgical reports and that at MR readout were observed in regard to 18% (62 of 336) of surfaces for reader 1 and in regard to 24% (79 of 336) of surfaces for reader 2 (Table 3). Differences between the classification in surgical reports and that at MR readout of two or more grades were observed in regard to 9% (30 of 336) of surfaces for reader 1 and in regard to 12% (42 of 336) of surfaces for reader 2. For reader 1, overclassification in regard to 9% of cartilaginous areas (29 of 336) and underclassification in regard to 19% (63 of 336) of these areas was observed, whereas for reader 2, overclassification in regard to 19% (65 of 336) of cartilaginous areas and underclassification of 17% (56 of 336) of these areas was observed.
View this table:
[in this window]
[in a new window]
|
Table 3. Results of Analysis of Distribution of Aberration of One Grade between MR Classification and Arthroscopic Classification
|
|
Diagnostic Performance
For the first part of the MR readout, when grade 0 (Fig 1) for normal cartilage and grade 1 (Fig 2) for softening or swelling of the cartilage were combined (cartilaginous surfaces without contour defects), 91% (305 of 336) of the abnormalities were correctly identified by reader 1 and 85% (287 of 336) of the abnormalities were correctly identified by reader 2 (Tables 4, 5). For the diagnosis of contour defects of the cartilaginous surface of grade 2 or higher (Figs 3–5) on images obtained with the coronal STIR sequence, statistical values for readers 1 and 2, respectively, were as follows: sensitivity values, 77% and 76%; specificity values, 96% and 89%; accuracy values, 91% and 85%; positive predictive values, 86% and 70%; and negative predictive values, 92% and 91%.
View larger version (92K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1: Grade 0 finding (no cartilaginous abnormality). Coronal fast spin-echo STIR MR image (5550/35/180) in 32-year-old man with arthroscopically normal articular cartilage of the medial compartment of the left femorotibial joint. MR image shows intact cartilage (arrowheads) of the femur and tibia.
|
|
View larger version (89K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2: Grade 1 lesion (cartilaginous signal intensity alterations). Coronal fast spin-echo STIR MR image (5550/35/180) in 57-year-old man in whom diagnostic arthroscopy of the right knee showed grade 1 lesions in the lateral tibial cartilage. MR image shows signal intensity alterations in lateral tibial cartilage with more hyperintense cartilage at the surface (arrowheads) and hypointense focal spots (arrows) within cartilaginous substance.
|
|
View larger version (84K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3: Grade 2 lesion (partial-thickness cartilaginous defect). Coronal fast spin-echo STIR MR image (5550/35/180) in 15-year-old female patient who underwent open-knee surgery for anterior cruciate ligament repair in the left knee. MR image shows partial-thickness defect (arrowheads) of articular cartilage of lateral femoral condyle; this finding also was seen at surgery.
|
|
View larger version (106K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4: Grade 3 lesion (fissuring of cartilage to level of subchondral bone). Coronal fast spin-echo STIR MR image (5550/35/180) in 39-year-old man who underwent arthroscopic partial lateral meniscectomy. A grade 3 chondral lesion with fissuring to the level of the subchondral bone (arrowheads) in the posterior aspect of the left lateral femoral condyle was reported at arthroscopy. The subchondral bone is not exposed, which discriminates this lesion from a grade 4 lesion, as depicted in Figure 5.
|
|
View larger version (90K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5: Grade 4 lesion (exposed subchondral bone). Coronal fast spin-echo STIR MR image (5550/35/180) in 56-year-old woman who underwent arthroscopic partial medial meniscectomy. MR image shows exposed subchondral bone (arrowheads) of the right medial femoral condyle.
|
|
Interobserver Reliability
For the first part of the MR readout, results of overall interobserver analysis indicated a substantial (11) agreement, with a value of 0.63.
Frequency of Hyperintense Subchondral Bone Marrow
For the second part of the MR readout, according to the analysis of reader 3, hyperintense subchondral bone marrow was associated with 54 (16%) of 336 arthroscopically proved cartilaginous surfaces. Hyperintense subchondral bone marrow lesions were found in 44% (14 of 32) grade 4 lesions. For all other grades, the frequency was lower. Hyperintense subchondral bone marrow was found in 20% (five of 25) of cases with grade 3 lesions, in 23% (seven of 30) of cases with grade 2 lesions, and in 16% (six of 37) of cases with grade 1 lesions. In 10% (22 of 212) of cartilaginous surfaces, the hyperintense area was not associated with a cartilaginous lesion. In 28 (52%) of 54 cartilaginous surfaces with hyperintense subchondral bone marrow, this finding was not associated with a cartilaginous defect (grades 2–4).
|
DISCUSSION
|
|---|
A number of therapeutic decisions depend on the precise diagnosis of cartilaginous abnormalities. Osteotomy, mosaicplasty, intraarticular drugs, or knee replacement surgery may be considered (12). Many MR imaging sequences with reasonable to good diagnostic performance for the characterization of articular cartilage have been described. These sequences include fat-suppressed three-dimensional SPGR (13), T1-weighted fat-suppressed three-dimensional fast low-angle shot (14), three-dimensional multipoint fat-water separation with steady-state free precession (15), and T2-weighted fast spin-echo MR imaging (16).
The results achieved in our study are comparable to those of other MR imaging studies that included assessment of knee cartilage. In our study with coronal STIR imaging for the diagnosis of contour defects of the cartilaginous surface, sensitivity values were 77% and 76%, specificity values were 96% and 89%, and accuracy values were 91% and 85% for readers 1 and 2, respectively. In a study by Bredella et al (16), the sensitivity, specificity, and accuracy of coronal T2-weighted fat-saturated fast spin-echo imaging were 61%, 99%, and 96%, respectively. Bredella et al (16) assessed the accuracy of routine T2-weighted MR imaging for the detection and classification of articular cartilaginous lesions in the knee. Contrary to what we analyzed in our study, they analyzed cartilaginous surface abnormalities to discriminate between low-grade (grades 1 and 2) and high-grade (grades 3 and 4) lesions.
Disler et al (17) retrospectively assessed 43 patients for articular cartilaginous defects of the knee with both standard MR imaging and sagittal fat-suppressed three-dimensional SPGR sequences. One-fourth of the patients had isolated articular cartilaginous lesions that mimicked meniscal tears clinically and were missed on standard MR images. The SPGR sequence had a significantly higher sensitivity than did standard MR sequences for detection of articular cartilaginous defects (75%–85% vs 29%–38%, P < .001 for each comparison), but there was no difference in specificity (97% vs 97%, P > .99) (17).
Recht et al (14) examined 41 patients who were suspected of having internal derangement of the knee by using a T1*-weighted fat-suppressed three-dimensional fast low-angle shot sequence. In comparison with arthroscopy, this sequence had sensitivity, specificity, and accuracy values of 81%, 97%, and 97%, respectively, for the detection of patellar cartilaginous defects. Seventy-seven percent of the lesions were classified identically at MR imaging and at arthroscopy. In the remaining 23%, MR imaging and arthroscopic classifications differed within one grade from each other (14). To obtain high-quality high-spatial-resolution images of articular cartilage with reduced imaging time, Reeder et al (15) combined a technique of multipoint fat-water separation with three-dimensional steady-state free precession imaging. Many of these sequences, however, are not part of routine imaging protocols.
Sonin and co-workers (2) analyzed the accuracy of fast spin-echo intermediate-weighted MR imaging for the assessment of cartilaginous defects in the knee. In comparison with arthroscopic data, the range of sensitivity values for three reviewers was 59%–73.5%, with specificity values of 86.7%–90.5% and accuracy values of 79.6%–86.1%. The value for interobserver variability was 0.63. In comparison with the results in that study, those in our study indicated a slightly better accuracy (91% vs 85%). The performance of the fast spin-echo intermediate-weighted sequence and that of the STIR sequence appear comparable.
The STIR sequence is commonly used in musculoskeletal imaging (18–20) for detection of abnormalities of bone marrow and soft tissue that are related to trauma, inflammation, degeneration, tumors, and other conditions (21). This sequence is robust even in imagers with imperfect field homogeneity. This sequence may produce accurate images when imaging is performed in physically difficult locations, such as the fingers and toes, or when susceptibility artifacts, such as those caused by implants, are present (22). A frequency-selective fat-suppressed T2-weighted spin-echo sequence may alternatively be employed (16). Compared with STIR imaging, the frequency-selective fat-suppressed T2-weighted spin-echo sequence is more likely to be associated with susceptibility artifacts.
In a study of Rubin and co-workers (23), hyperintense focal subchondral areas were seen with fluid-sensitive sequences (STIR or T2-weighted sequences) in 72%–83% of patients with treatable traumatic cartilaginous defects in the knee. Rubin and co-workers concluded that these MR sequences may play an important role in the detection of such lesions. In our series of patients, the frequency of hyperintense areas of subchondral bone marrow was lower (44% of grade 4 lesions). This may be explained by differences in patient selection. Rubin et al included only patients with treatable or potentially treatable cartilaginous defects. In our study, patient selection was consecutive, and, therefore, the cartilaginous defects were often milder. In 51% of hyperintense subchondral bone marrow areas, cartilaginous defects (grade 0 or 1) were not found at arthroscopy.
In our study, the sensitivity for the detection of femoral cartilaginous defects was higher than that for tibial cartilaginous defects for both readers, whereas the specificity and accuracy values for both of these surfaces were comparably high. The reason for this difference is not clear. One possible explanation is that tibial cartilage may be quite thin, especially at the periphery of the tibial plateau (17). Another explanation is the slightly lower frequency of defects in tibial cartilage compared with femoral cartilage (44% femoral vs 29% tibial on the medial side and 17% vs 14% on the lateral side).
Our study had several limitations. Because of the retrospective design of the study, the STIR sequence was acquired only in the coronal plane, and the cartilage of the patella and the trochlea could not be evaluated. The Outerbridge classification system is widely accepted by surgeons for arthroscopic evaluations (7,8,10,24), but it has limitations for MR imaging. Qualitative changes of the cartilage such as grade 1 lesions, which are characterized by softening and swelling at arthroscopy, are difficult to detect with MR imaging. Similarly, at MR imaging, it may be difficult to differentiate between grade 3 lesions (fissuring of cartilage to the level of the subchondral bone) and grade 4 lesions (exposed subchondral bone). Other classification systems, such as that of Noyes, may be more appropriate for MR imaging (14,25) but would not be comparable to the classification at arthroscopy. Both readers were aware that cartilaginous lesions in corresponding cartilaginous surfaces (eg, lateral tibial plateau and lateral femoral condyle) are likely to coexist.
In conclusion, contour defects of femorotibial cartilage can be detected with reasonable accuracy by using a routine STIR sequence.
|
ADVANCES IN KNOWLEDGE
|
|---|
- The coronal short inversion time inversion-recovery MR sequence has accuracy values for the detection of cartilaginous contour defects in the knee joint of between 87% (medial tibial plateau) and 94% (lateral femoral condyle).
- Besides its ability to show bone marrow signal intensity alterations, the short inversion time inversion-recovery sequence has a reasonable accuracy (91% [reader 1] and 85% [reader 2]) also for detection of cartilaginous contour defects.
|
FOOTNOTES
|
|---|
Abbreviations: SPGR = spoiled gradient-recalled acquisition in the steady state • STIR = short inversion time inversion-recovery
Authors stated no financial relationship to disclose.
Author contributions: Guarantors of integrity of entire study, M.R.S., J.H., C.W.A.P.; 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, K.P.J., M.R.S.; clinical studies, K.P.J., P.K.; statistical analysis, K.P.J., C.W.A.P.; and manuscript editing, all authors
|
References
|
|---|
- Gold GE, McCauley TR, Gray ML, Disler DG. What's new in cartilage? RadioGraphics 2003;23:1227–1242.[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]
- 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;49:700–709.[CrossRef][Medline]
- Kshirsagar AA, Watson PJ, Tyler JA, Hall LD. Measurement of localized cartilage volume and thickness of human knee joints by computer analysis of three-dimensional magnetic resonance images. Invest Radiol 1998;33:289–299.[CrossRef][Medline]
- 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]
- McCauley TR, Recht MP, Disler DG. Clinical imaging of articular cartilage in the knee. Semin Musculoskelet Radiol 2001;5:293–304.[CrossRef][Medline]
- Outerbridge RE. The etiology of chondromalacia patellae. J Bone Joint Surg Br 1961;43-B:752–757.
- Cameron ML, Briggs KK, Steadman JR. Reproducibility and reliability of the Outerbridge classification for grading chondral lesions of the knee arthroscopically. Am J Sports Med 2003;31:83–86.[Abstract/Free Full Text]
- Brismar BH, Wredmark T, Movin T, Leandersson J, Svensson O. Observer reliability in the arthroscopic classification of osteoarthritis of the knee. J Bone Joint Surg Br 2002;84:42–47.
- Potter HG, Linklater JM, Allen AA, Hannafin JA, Haas SB. Magnetic resonance imaging of articular cartilage in the knee: an evaluation with use of fast-spin-echo imaging. J Bone Joint Surg Am 1998;80:1276–1284.[Abstract/Free Full Text]
- Kundel HL, Polansky M. Measurement of observer agreement. Radiology 2003;228:303–308.[Abstract/Free Full Text]
- Buckwalter JA, Mankin HJ. Articular cartilage repair and transplantation. Arthritis Rheum 1998;41:1331–1342.[CrossRef][Medline]
- 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]
- 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]
- 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]
- 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]
- 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]
- Tokuda O, Hayashi N, Matsunaga N. MRI of bone tumors: fast STIR imaging as a substitute for T1-weighted contrast-enhanced fat-suppressed spin-echo imaging. J Magn Reson Imaging 2004;19:475–481.[CrossRef][Medline]
- Mayerhoefer ME, Breitenseher M, Hofmann S, et al. Computer-assisted quantitative analysis of bone marrow edema of the knee: initial experience with a new method. AJR Am J Roentgenol 2004;182:1399–1403.[Abstract/Free Full Text]
- Park SW, Lee JH, Ehara S, et al. Single shot fast spin echo diffusion-weighted MR imaging of the spine: is it useful in differentiating malignant metastatic tumor infiltration from benign fracture edema? Clin Imaging 2004;28:102–108.
- Schmid MR, Pfirrmann CW, Hodler J, Vienne P, Zanetti M. Cartilage lesions in the ankle joint: comparison of MR arthrography and CT arthrography. Skeletal Radiol 2003;32:259–265.[Medline]
- Hilfiker P, Zanetti M, Debatin JF, McKinnon G, Hodler J. Fast spin-echo inversion-recovery imaging versus fast T2-weighted spin-echo imaging in bone marrow abnormalities. Invest Radiol 1995;30:110–114.[CrossRef][Medline]
- Rubin DA, Harner CD, Costello JM. Treatable chondral injuries in the knee: frequency of associated focal subchondral edema. AJR Am J Roentgenol 2000;174:1099–1106.[Abstract/Free Full Text]
- Haims AH, Moore AE, Schweitzer ME, et al. MRI in the diagnosis of cartilage injury in the wrist. AJR Am J Roentgenol 2004;182:1267–1270.[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]
TOP
ABSTRACT
INTRODUCTION
