| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Radiology, University Hospital Balgrist, Forchstrasse 340,
Zurich CH-8008, Switzerland.
2 Department of Orthopedic Surgery, University Hospital Balgrist, Zurich
CH-8008, Switzerland.
3 MR Applications Development, Siemens Medical Solutions, Erlangen,
Germany.
Received July 7, 2004;
accepted after revision August 14, 2004.
Address correspondence to M. R. Schmid.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Our study included 52 consecutive patients who had knee surgery within 4 months of undergoing an MRI examination including an axial 2D MEDIC (TR/TE, 884/26; flip angle, 30°) sequence. Cartilage was surgically graded on a 5-point scale: 0, normal; 1, softening or swelling; 2, partial thickness defect; 3, fissuring to the level of the subchondral bone; or 4, exposed subchondral bone. Cartilage was graded on MRI according to a scale that was almost identical to the surgical scale except that grade 1 lesions were defined as signal alteration or swelling of cartilage. Two blinded reviewers independently analyzed patellar cartilage. Sensitivity, specificity, accuracy, and weighted kappa values for interobserver variability were calculated.
RESULTS. Low-grade cartilage lesions predominated in our study
group. When grade 2 or higher was considered the threshold for relevance, the
sensitivity, specificity, and accuracy for the MEDIC sequence was as high as
79%, 82%, and 81%, respectively. Increasing the threshold of relevance to
grade 3 increased the sensitivity, specificity, and accuracy to as high as
83%, 91%, and 90%, respectively. Interobserver agreement for the MEDIC
sequence was good (weighted 
= 0.68).
CONCLUSION. The 2D MEDIC sequence performs comparably to previously described sequences optimized for cartilage imaging such as the 3D double-echo steady-state or 3D spoiled gradient-recalled sequences with good interobserver agreement, high sensitivity, and excellent specificity for revealing low- to intermediate-degree cartilage defects.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Standard radiographs and clinical findings [7, 8] do not correlate well with the degree of osteoarthritis. MRI with and without intraarticular contrast administration is currently the most reliable imaging method for detecting articular cartilage defects in the knee joint [9-23], with slightly less impressive results for other joints whose cartilage is thinner [24-29].
Standard MRI sequences such as T2-weighted and proton-density-weighted spin-echo and turbo spin-echo sequences with and without fat suppression show satisfactory results in cartilage defect detection [9, 12]. Dedicated cartilage sequences may improve the reliability of MRI [10, 11, 13, 15-18]. This group includes 3D spoiled gradient-echo (SPGR) or 3D fast low-angle shot (FLASH) sequences with fat saturation and 3D double-echo steady-state (DESS) sequences, which showed sensitivity and specificity as high as 96% and 95%, respectively [15]. The MEDIC combination sequence is another MRI sequence potentially useful in imaging of articular cartilage. On MEDIC and DESS images, bone is hypointense, articular cartilage is of intermediate to low signal intensity, and joint fluid is hyperintense relative to cartilage, unlike on SPGR or FLASH sequence images obtained with fat suppression, in which joint fluid appears hypointense compared with articular cartilage. The purpose of our study was to evaluate the diagnostic value of the 2D MEDIC in the detection of patellar cartilage defects.
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Imaging
All patients were examined on a 1.5-T MR system (Symphony, Siemens Medical
Solutions). For analysis of the retropatellar cartilage, only the axial MEDIC
sequence (TR/TE, 884/26; flip angle, 30°; bandwidth, 391 Hz/pixel; section
thickness, 3 mm; intersection gap, 1.5 mm; 19 sections; field of view, 15 cm;
matrix size, 240 x 256; one acquisition; acquisition time, 3 min 48 sec)
was evaluated. Six echoes were used in this multiecho implementation with the
following TE: 12.8, 18.0, 23.2, 28.4, 33.6, and 38.8 msec. The following
sequences were included in the routine imaging protocol but were not evaluated
for this study: sagittal proton-density- and T2-weighted turbo spin-echo
sequences, coronal T1-weighted spin-echo sequence, and coronal T2-weighted
turbo spin-echo sequence with fat saturation.
MEDIC Sequence Description
In orthopedic imaging, T2*-weighted gradient-echo sequences are
typically acquired with a single echo per line in k-space (e.g., in fast
imaging with steady-state free precession [FISP] or FLASH sequences with long
TE). To compensate for the low signal-to-noise ratio inherent in such
sequences, one must use a low receiver bandwidth. This, however, compromises
spatial resolution because the shape of the gradient echo is significantly
impaired by the T2* decay of the transversal magnetization. The
MEDIC sequence provides a potential solution to this problem. The sequence is
schematically shown in Figure
1. A series of identically phase-encoded gradient echoes is
sampled per line in k-space. Unipolar readout gradients are used to avoid
off-resonance effects and to achieve flow compensation in the readout
direction. Magnitude images are reconstructed for each echo. The resulting
images are then combined using a sum of squares algorithm. The combination of
multiple echoes improves the signal-to-noise ratio. Receiver bandwidth can
then be increased, and the readout duration reduced. As a consequence,
T2* effects and impairment of the spatial resolution are reduced
compared with a low bandwidth gradient-echo sequence acquiring a single
echo.
Surgical Grading of Cartilage Lesions, Image Analysis, and Statistics
The grading used by the surgeons involved in this study is a modification
of the system originally described by Outerbridge
[30]. The following degrees of
cartilage damage are differentiated on a 5-point scale: 0, normal; 1,
softening or swelling; 2, partial thickness defect; 3, fissuring to the level
of the subchondral bone; or 4, exposed subchondral bone. In the original
Outerbridge classification system
[30], grades 2 and 3 lesions
are defined as fragmentation or fissuring with a diameter of less (grade 2) or
more (grade 3) than 1 inch (
2.5 cm). In cases of multiple cartilage
defects within the retropatellar cartilage surface, only the highest grade of
cartilage defect was described in surgical reports. The axial 2D MEDIC was
retrospectively reviewed separately on a PACS workstation (ID.Read, Image
Devices) by two musculoskeletal radiologists who were blinded to clinical data
including surgical reports. Their experience with MRI of the joints was 10 and
6 years, respectively. Grade 1 lesions were diagnosed in the presence of
signal alterations without cartilage deformity or when cartilage swelling was
seen. Grades 2-4 were graded according to the same criteria as the ones used
during surgery (Figs. 2,
3,
4). According to the cartilage
description in the surgical reports, the retropatellar cartilage (medial and
lateral facets) was analyzed as a single entity. In cases of multiple
retropatellar cartilage defects, the score of the worst cartilage abnormality
was used for this study.
|
|
Sensitivity, specificity, and accuracy of the MEDIC sequence compared with the surgical reports were calculated. Weighted kappa values [31, 32] were calculated to assess interobserver agreement. According to Landis and Koch [33], a kappa value of 0.20 or less indicated poor agreement; 0.21-0.40, fair; 0.41-60, moderate; 0.61-0.80, good; and 0.81-1.0, very good agreement. For statistical analysis, Statistical Package for the Social Sciences software (version 10, SPSS) was used.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
value for multiple categories) was good (weighted
= 0.68).
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To our knowledge, the diagnostic performance of the 2D MEDIC sequence in detecting lesions of articular cartilage has not been evaluated in the peer-reviewed literature. Similar to the 3D DESS sequence [13, 17, 18], our study sequence is a "bright-fluid" sequence (Figs. 2, 3, 4), showing hyaline cartilage as relatively dark in comparison to the hyperintense joint fluid. Proton-density-weighted spin-echo sequences show similar relative signal intensities. However, the signal-difference-to-noise ratio between cartilage and joint fluid is inferior to 3D DESS and MEDIC images. The MEDIC sequence is a T2*-weighted gradient-echo sequence optimized for orthopedic imaging. The MEDIC sequence addresses the main limitations of other gradient-echo sequences such as low signal-to-noise ratio, image degradation due to susceptibility gradients, and chemical shift. This is achieved by acquiring and combining multiple echoes with a high receiver bandwidth instead of using a single low-bandwidth echo. Our results show that the diagnostic performance of the 2D MEDIC sequence is comparable to the established 3D DESS sequence that has already been described by several authors [13, 17, 18, 23]. The best diagnostic performance of the 3D DESS sequence has been reported by Murphy [13] with a sensitivity, specificity, and accuracy of 80%, 95%, and 91%, respectively, in detecting cartilage lesions. However, only grades 3 and 4 of cartilage lesions were included in that investigation [13], whereas our study population included a larger proportion of low-grade lesions, which are more difficult to detect. The comparison to the 3D DESS sequence is important because this study has been shown to be superior for detection of low-grade or superficial cartilage defects to 3D fat-suppressed FLASH (or SPGR) sequence despite the greater tissue contrast of the FLASH sequence in a phantom study of Mosher and Pruett [23]. In our study population, low-grade chondral defects predominated. Such abnormalities are more difficult to detect than larger defects. This fact has to be taken into consideration when the results are compared with those of other investigations.
Unlike 3D DESS and MEDIC sequences, most other cartilage MRI sequences are "dark-fluid" sequences. In fat-suppressed SPGR or FLASH sequences with flip angles of approximately 60° [10, 11, 14-16, 20], articular cartilage appears relatively hyperintense in comparison to joint fluid. Disler et al. [10] have compared fat-suppressed SPGR images to standard spin-echo MRI sequences and found significantly higher sensitivities for cartilage lesion detection with the fat-suppressed SPGR sequence (75-85% vs 29-38%). The specificities of fat-suppressed SPGR and spin-echo sequences were identical at 97%. According to our results, the 2D MEDIC sequence has sensitivity similar to that of the fat-suppressed SPGR sequence.
Few studies have compared the diagnostic performance of standard MRI and MR arthrography in detecting articular cartilage defects in the knee joint. Rand et al. [16] found that the sensitivity of SPGR with fat saturation (diagnosis correct within one grade) increased from 81% to 90% after intraarticular contrast administration. In an in vitro study comparing major types of cartilage sequences, MR arthrography had a variable advantage (sensitivity, 45% vs 3-38%) in the detection of grade 2 cartilage lesions over 3D FISP, 3D FLASH (or SPGR), and 2D FLASH sequences [22].
CT arthrography has rarely been used in the diagnosis of articular cartilage abnormalities. In lesions involving less than half of the cartilage thickness, sensitivity and specificity of CT arthrography were slightly superior to those of standard MRI (80-88% vs 78-86%) [19]. This method, however, requires an intraarticular contrast injection and is associated with a relevant local and a small effective radiation dose to the patient.
The subjects in our study were relatively young (mean age, 36 years), and low grades of cartilage lesions predominated. This is a clinically relevant patient group. A number of surgical procedures have been introduced for repair of localized chondral defects [2]. Some patients with a lack of cartilage damage in one knee compartment can benefit from unicompartmental knee arthroplasty or tibial osteotomy [4]. When comparing in vivo studies of articular cartilage abnormalities, the reviewers should keep in mind that the standard of reference (arthroscopy) is not perfect. For several classification systems, relevant interobserver variability in the grading of low-grade chondral defects in the knee joint has been shown [35].
One limitation of our study could be the relatively thick slices used in the MEDIC sequence. Thinner slices using the MEDIC sequence are possible. However, for clinical imaging, the coverage in the z-axis with axial images is important because they may not only show retropatellar cartilage but also abnormalities of the femoral groove, the medial and lateral retinaculum, the patellar tendon, and meniscal and synovial cysts. At our institution, the 3-mm slice thickness with an intersection gap of 1.5 mm is considered to represent an optimal compromise with regard to signal-to-noise ratio, z-axis coverage, and examination time. If optimized for cartilage imaging with a smaller slice thickness, a smaller interslice gap, and an increased number of acquisitions, the MEDIC sequence may perform even better than presented in our results. Other limitations are the relatively small sample size including a small number of high-grade cartilage lesions. In addition, the results of our study have not been compared with another imaging method in the same study population. Instead, our results have been compared with surgical findings and to results of other investigations with different reviewers, different patients, and different equipment.
In conclusion, the 2D MEDIC sequence performs comparably to previously described sequences optimized for cartilage imaging such as the 3D DESS or 3D SPGR sequences with good interobserver agreement, high sensitivity, and excellent specificity in detecting low to intermediate degrees of cartilage defects.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  G. Andreisek, J. M. Froehlich, J. Hodler, D. Weishaupt, V. Beutler, C. W. A. Pfirrmann, C. Boesch, and D. Nanz Direct MR Arthrography at 1.5 and 3.0 T: Signal Dependence on Gadolinium and Iodine Concentrations--Phantom Study Radiology, June 1, 2008; 247(3): 706 - 716. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. R. Duc, C. W. A. Pfirrmann, M. R. Schmid, M. Zanetti, P. P. Koch, F. Kalberer, and J. Hodler Articular Cartilage Defects Detected with 3D Water-Excitation True FISP: Prospective Comparison with Sequences Commonly Used for Knee Imaging Radiology, October 1, 2007; 245(1): 216 - 223. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. R. Duc, P. Koch, M. R. Schmid, W. Horger, J. Hodler, and C. W. A. Pfirrmann Diagnosis of Articular Cartilage Abnormalities of the Knee: Prospective Clinical Evaluation of a 3D Water-Excitation True FISP Sequence Radiology, May 1, 2007; 243(2): 475 - 482. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
