| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Radiology, University Hospital Balgrist, Forchstrasse 340,
Zurich CH-8008, Switzerland.
2 Present address: Department of Radiology, Kantonsspital Baden, Daettwil
CH-5405, Switzerland.
3 Department of Orthopedics, University Hospital Balgrist, Zurich CH-8008,
Switzerland.
4 Present address: Department of Orthopedic Surgery, Zieglerspital Bern,
Morillonstrasse, Bern 3001, Switzerland.
5 MR Application Development, Siemens AG Medical Solutions, Karl Schall Strasse
4, Erlangen D-91050, Germany.
Received November 25, 2003;
accepted after revision May 20, 2004.
Address correspondence to M. R. Schmid
(marius.schmid{at}balgrist.ch).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. In 50 MR arthrograms of the hip joint obtained in 47 consecutive patients, a sagittal 3D double-echo steady-state sequence (TR/TE, 24/6.5; flip angle, 25°) was compared with a sagittal T1-weighted spin-echo sequence (350/14). Two musculoskeletal radiologists independently evaluated articular cartilage. Sensitivity and specificity for detecting cartilage defects were calculated for those hips that underwent open surgery (n = 21). Lesion conspicuity was retrospectively reviewed and graded between 1 (not visible) and 5 (well defined).
RESULTS. At surgery, a total of 26 lesions of the acetabular
(n = 20) and femoral (n = 6) cartilage were found. For the
3D double-echo steady-state and T1-weighted spin-echo sequences, sensitivities
and specificities for cartilage lesion detection were 58% and 88% and 81% and
81% for reviewer 1 and 62% and 94% and 62% and 100% for reviewer 2,
respectively. Lesion conspicuity was significantly superior (p =
0.036) for the 3D double-echo steady-state sequence (mean grade, 3.4) compared
with the T1-weighted spin-echo sequence (mean grade, 3.0). The kappa value was
fair for the 3D double-echo steady-state sequence (
= 0.40) and
moderate for the T1-weighted spin-echo sequence ( = 0.55).
CONCLUSION. The 3D double-echo steady-state sequence optimized for cartilage imaging improves lesion conspicuity but does not improve diagnostic performance.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MRI with and without intraarticular administration of gadopentetate dimeglumine has been widely used for the assessment of articular cartilage [5–31]. Contrary to many investigations reporting excellent results in imaging cartilage of the knee joint [6–10], evaluation of articular cartilage in the hip joint is more difficult [5, 11]. This difference may be explained by the fact that cartilage of the hip joint is thinner than that in the knee joint and larger coils with lower signal-to-noise ratio are used. Therefore, optimization of MR sequences for cartilage imaging in the hip is relevant.
Several MR sequences have been used for cartilage imaging, usually in the knee joint, including fat-suppressed 3D spoiled gradient-recalled acquisition in the steady state (SPGR) [7, 12–15, 23], fast low-angle shot (FLASH) [8, 18], T2- and intermediate-weighted fat-suppressed fast spin-echo [6, 9], and water-excitation 3D double-echo steady-state [10, 16] sequences. In the detection of cartilage defects in cadaveric knee joints, MR arthrography has been shown to be superior to MRI without intraarticular contrast administration [14]. A combination of MR sequences specific for cartilage and intraarticular gadolinium injection could probably improve imaging of the cartilage in the hip. The 3D double-echo steady-state sequence allows thinner sections to be obtained than conventional spin-echo sequences and provides high contrast between intraarticular structures and injected gadopentetate dimeglumine. The purpose of this study was to compare the diagnostic performance of an MR sequence dedicated to cartilage imaging (3D double-echo steady-state) with a standard MR sequence (T1-weighted spin-echo) in detecting articular cartilage lesions of the hip after intraarticular injection of gadopentetate dimeglumine.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Imaging Protocol
All studies were performed on a 1.5-T MR system (Symphony, Siemens Medical
Solutions) according to a standardized protocol. A flexible wraparound
receive-only surface coil was used. A sagittal 3D double-echo steady-state
sequence (Fig. 1) optimized for
cartilage imaging and a sagittal T1-weighted spin-echo sequence were
performed. In this 3D double-echo steady-state sequence, a 1–2–1
binomial pulse is used for spatial- and frequency-selective excitation of
water isochromats while suppressing signals from fatty tissues. The
double-echo steady-state sequence simultaneously acquires the signals of two
sequences. The first one is the signal of a fast imaging with steady-state
precession (FISP) sequence, also known as coherent steady-state signal. The
second one is the signal of a PSIF sequence (time reversal of FISP). Images
from both echoes are reconstructed and added using a sum-of-squares algorithm.
Adding the FISP and PSIF signals increases the T2-weighting of the FISP images
and allows improved delineation of fat, joint fluid, and cartilage in
comparison with a FISP image. On water-excitation 3D double-echo steady-state
images, both joint fluid and gadopentetate dimeglumine (2 mmol/L) are
hyperintense, whereas articular cartilage has an intermediate signal
intensity.
|
The imaging parameters for the 3D double-echo steady-state sequence in our study were as follows: TR/TE, 24/6.5; section thickness, 1.7 mm; no intersection gap; flip angle, 25°; field of view, 15 cm; matrix, 512 x 512; 1 average; and acquisition time, 4 min 6 sec. For the T1-weighted spin-echo sequence, the following imaging parameters were used: 350/14; section thickness, 4 mm; intersection gap, 0.8 mm; field of view, 16 cm; matrix, 512 x 512; 2 averages; and acquisition time, 3 min 48 sec.
In addition, three other sequences were performed as part of our routine hip MR arthrography protocol: a coronal T1-weighted spin-echo sequence (524/14; section thickness, 3 mm; intersection gap, 0.6 mm; field of view, 16 cm; matrix, 512 x 512; 1 average; and acquisition time, 3 min 25 sec), a coronal proton density–weighted turbo spin-echo sequence with fat saturation (2,500/42; section thickness, 3 mm; intersection gap, 0.6 mm; flip angle, 180°; field of view, 16 cm; matrix 512 x 512; turbo factor, 7; 1 average; and acquisition time, 2 min 50 sec), and an oblique axial fat-suppressed FLASH sequence (600/11.8; section thickness, 3 mm; intersection gap, 0.6 mm; flip angle, 60°; field of view, 16 cm; matrix, 512 x 512; 1 average; and acquisition time, 3 min 52 sec).
Intraarticular contrast injection was performed in a standardized fashion 30 min or less before scanning: 1–2 mL of local anesthetics ([mepivacaine hydrochloride 2%] Scandicain, AstraZeneca), 1 mL of iodinated contrast agent ([iopamidol 200 mg I/mL] Iopamiro 200, Bracco), and a mean of 8 mL (range, 6–10 mL) of a diluted MR contrast agent ([gadopentetate dimeglumine] Magnevist, Schering) at a concentration of 4 mmol/L were injected under fluoroscopic control.
Image Analysis
All 50 MR studies were reviewed on a PACS workstation (ID.Read, Image
Devices) by two musculoskeletal radiologists separately and blinded to
clinical data including surgical reports. Both 3D double-echo steady-state and
T1-weighted spin-echo sequences were reviewed during separate sessions at
least 1 week apart. During an additional session, the diagnostic performance
of the coronal fat-suppressed proton density–weighted turbo spin-echo
sequence was evaluated in the 21 hips that underwent surgery because similar
sequences are often included in MR arthrography protocols. The reviewers were
asked to report whether acetabular, femoral, or both types of cartilage
lesions were present.
Error Analysis
A board of five radiologists including the two initial reviewers
reevaluated all MR studies with surgical proof (n = 21) with regard
to potential sources of error. This evaluation was performed in consensus. A
predefined list of sources of error was used: thin cartilage; no contrast
between the two cartilage layers; anatomic irregularities at the transition
between cartilage and adjacent structures (transitional zones between
cartilage and acetabular labrum, acetabular fossa, fovea of the femoral head,
and border of the femoral neck); partial volume effect artifacts; blurring;
motion artifacts; chemical shift artifacts; noisy images; reviewer error; and
small lesion (limited to a few pixels).
Lesion Conspicuity
Lesion conspicuity was graded between 1 and 5 by the same board of
radiologists. Grades were defined as follows: 1, lesion not visible even in
retrospect; 2, only a few pixels visible in retrospect; 3, lesion
retrospectively visible, prospective error in diagnosis understandable; 4,
lesion clearly visible but not well defined; and 5, lesion clearly visible and
well defined.
Statistics
On the basis of the data of the 21 patients with surgical proof,
sensitivity, specificity, accuracy, and positive and negative predictive
values were calculated for all three evaluated sequences and for both
reviewers. The mean lesion conspicuity grade (all lesions diagnosed by
reviewer 1, reviewer 2, or both reviewers) was calculated for 3D double-echo
steady-state and T1-weighted spin-echo sequence separately. The p
values were calculated using a Wilcoxon's signed rank test. The kappa
statistics were calculated for all 50 examinations to assess the level of
interobserver agreement. According to Landis and Koch
[32], a kappa value of 0.20 or
less indicates poor agreement; 0.21–0.40, fair; 0.41–0.60,
moderate; 0.61–0.80, good; and 0.81–1.0, very good agreement.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Results of both reviewers for all three study sequences are shown in Table 1. In general, all sequences were specific but relatively insensitive. The sensitivity of the 3D double-echo steady-state sequence was identical to that of the T1-weighted spin-echo sequence for reviewer 1, and the sensitivity for both sequences was equal for reviewer 2. The specificity of the 3D double-echo steady-state sequence was slightly lower for reviewer 1 but was higher for reviewer 2. The 3D double-echo steady-state sequence was less accurate than the T1-weighted spin-echo sequence for reviewer 1 but was almost identical for reviewer 2. The coronal proton density–weighted turbo spin-echo sequence was slightly more sensitive but was less specific and did not improve accuracy compared with the other two study sequences.
|
The results of the error analysis are shown in Table 2. Small lesions and partial volume effect artifacts were the most common presumed reasons for false-negative diagnoses. The remaining possible sources of errors were either rare or not considered to be the main reason for false diagnoses at all.
|
Lesion conspicuity as determined by the panel is summarized in
Table 3. Significantly higher
(p = 0.036, Wilcoxon's signed rank test) average lesion conspicuity
grade of the 3D double-echo steady-state sequence was found. Interobserver
agreement for the detection of cartilage lesions in acetabular and femoral
cartilage was fair ( = 0.40) for the 3D double-echo steady-state images
and moderate ( = 0.55) for the T1-weighted spin-echo images.
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In our 21 patients with surgical confirmation, the accuracy of the 3D double-echo steady-state and that of the T1-weighted spin-echo sequences were comparable for reviewer 2 (74% vs 76%, respectively). For reviewer 1, the 3D double-echo steady-state sequence was unexpectedly inferior to the T1-weighted spin-echo sequence (69% vs 81%, respectively).
T1-weighted images are still included in many MR joint protocols. They often provide excellent anatomic information and show abnormalities of subchondral bone that cannot necessarily be seen on gradient-echo images. This last aspect may explain the somewhat unexpected high sensitivity of the T1-weighted sequence in detecting cartilage lesions, although section thickness, intersection gap, and field of view favor the 3D double-echo steady-state sequence. Schmid et al. [5] found that secondary signs of osteoarthritis (e.g., subchondral sclerosis, subchondral cysts, and osteophytes) may be misleading. However, our data may still direct the reviewer to adjacent minor cartilage abnormalities (Figs. 2A, 2B and 3A, 3B). The coronal fat-suppressed proton density–weighted turbo spin-echo sequence yields images with characteristics similar to the T1-weighted fat-suppressed spin-echo sequence, which is preferred by many radiologists for MR arthrography. However, this sequence was not superior to the other two sequences used in our study (Table 1).
|
|
|
|
Contrary to the diagnostic performance that showed a slight advantage for the T1-weighted sequence, lesion conspicuity for the 3D double-echo steady-state sequence was significantly superior to the T1-weighted spin-echo sequence when reviewed retrospectively (Figs. 2A, 2B, 3A, 3B, 4A, 4B). Although this fact is not relevant with regard to the diagnostic performance of the two reviewers involved in this investigation, the 3D double-echo steady-state sequence may be valuable for less experienced radiologists or for discussion with clinicians and patients less accustomed to the relatively subtle signal differences seen on T1-weighted spin-echo images. In addition, the smaller possible slice thickness associated with the 3D acquisition used for the double-echo steady-state sequence allows reduction of partial volume effect artifacts, which may be more relevant in small bones with thin-cartilage surfaces such as in the wrist and foot than in large bones with thick-cartilage surfaces.
|
|
The diagnostic performances of all the investigated sequences are comparable to the results of previously published studies of the hip and other joints. In an MR arthrography study of the hip, accuracies of 78% (reviewer 1) and 69 % (reviewer 2) for detecting cartilage defects were found. A 1.0-T scanner and three different imaging planes (a sagittal T1-weighted turbo spin-echo, an oblique coronal T1-weighted turbo spin-echo and oblique coronal T1-weighted fat-suppressed spin-echo, and an oblique axial T1-weighted fat-suppressed FLASH sequence) were used [5]. Nakanishi et al. [15] investigated articular cartilage of the hip without intraarticular contrast administration. They used a dedicated fat-suppressed 3D SPGR sequence during longitudinal leg traction. In 14 of their 15 patients with osteonecrosis (n = 5), acetabular dysplasia (n = 5), or osteoarthritis (n = 5), cartilage abnormalities corresponded well to macroscopic or arthroscopic findings. The MR appearance of articular cartilage was normal in all eight young volunteers included in that study [15]. Hodler et al. [28] reported a reasonable correlation (r = 0.25–0.58) between anatomic and MR measurements of cartilage thickness in both the acetabulum and femur. A fat-suppressed SPGR sequence and a T1-weighted spin-echo sequence were used in that investigation.
Several studies performed in the knee with fat-suppressed 3D SPGR [7, 12–15, 23], FLASH [8, 18], T2- and intermediate-weighted fat-suppressed fast spin-echo [6, 9], and 3D double-echo steady-state [10, 16] sequences have reported sensitivities, specificities, and accuracies of 43–94%, 88–100%, and 83–98%, respectively. These values cannot directly be compared with values obtained in the hip, however. Hip cartilage is as thin as 1–2 mm (mean) [28, 33, 34] versus knee cartilage that is reported to be up to 7 mm thick at the patella [33]. In addition, hip cartilage is less accessible for imaging because of its anatomic location deep within soft tissue. Flexible surface coils or body array coils are commonly used for imaging of the hip instead of dedicated send–receive extremity coils with their more homogeneous field that are commonly used for imaging the knee. Previously published MR studies of joints with relatively thin cartilage have shown sensitivities, specificities, and accuracies similar to our results in the glenohumeral joint (53–100%, 51–87%, 65–75%, respectively) [20] and ankle joint (59–80%, 76–97%, 75–88%, respectively) [24].
Interobserver agreement was fair to moderate in our study ( = 0.40
for 3D double-echo steady-state and = 0.55 for T1-weighted spin-echo
sequences). Similar or inferior kappa values have previously been reported for
patellar cartilage ( = 0.45 for 3D double-echo steady-state images and
= 0.46 for T1-weighted turbo spin-echo images
[16]), for the hip ( =
0.20–0.31 for MR arthrography and standard spin-echo and 2D
fat-suppressed FLASH sequences
[5]), and for the glenohumeral
joint ( = 0.20–0.27 for T1-weighted spin-echo, intermediate-, and
T2-weighted fast spin-echo sequences
[20]).
The performance of standard MRI and MR arthrography also must be compared with CT arthrography, which may be used as an alternative imaging technique for cartilage imaging. CT arthrography has successfully been used in the knee [14, 17, 18, 27] and ankle [24] joints. Vande Berg et al. [17] compared the performance of CT arthrography and nonarthrographic MRI in the detection of cartilage lesions in cadavers. They found sensitivities and specificities of 80–88% and 78–85%, respectively. In a study of cadaveric specimens, Rand et al. [14] found that the sensitivity of CT arthrography and that of standard MR sequences (T1- and proton density–weighted) were identical in the detection of different degrees of abnormalities of the cartilage surface (86% for all methods) and were slightly lower than in MR arthrography (90% for SPGR sequence). However, CT arthrography failed to depict intrachondral lesions. Gagliardi et al. [27] reported sensitivities, specificities, and accuracies for MR arthrography (80–100%, 98–100%, 94–100%, respectively) that are superior to those of CT arthrography (73–100%, 100%, 92–100%, respectively) in the diagnosis of chondromalacia patellae (grade 2 or higher) in vivo. In another study comparing CT arthrography and MR arthrography of the ankle, similar sensitivities, specificities, and accuracies were found for the two methods [24] (tibia and talus: 60–81%, 89–97%, 90–94% vs 59–80%, 76–97%, 75–88%, respectively). In the hip, CT arthrography is associated with a relevant amount of radiation exposure. This is important because cartilage damage in the hip joint occurs in young individuals. Therefore, MR examinations should be favored in this region.
The comparison of two sequences in only one imaging plane as used in our investigation is somewhat artificial because consultation of coronal images may prevent false results that are due to partial volume effect artifacts or problems relating to anatomic transitional zones (e.g., the acetabular cartilage abutting the labrum). However, in a study design with additional imaging planes, any differences between the two evaluated types of sequences might have become less evident.
In conclusion, for MR arthrograms, both a standard T1-weighted spin-echo and a water-excitation 3D double-echo steady-state MR sequence are specific but insensitive to cartilage lesions of the hip joint. The 3D double-echo steady-state sequence optimized for cartilage imaging does not improve the diagnostic performance in this setting, although conspicuity of cartilage lesions was markedly better than on T1-weighted spin-echo sequence.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  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]
|
|
|
|
 |
|
 G. Andreisek, S. R. Duc, J. M. Froehlich, J. Hodler, and D. Weishaupt MR Arthrography of the Shoulder, Hip, and Wrist: Evaluation of Contrast Dynamics and Image Quality with Increasing Injection-to-Imaging Time Am. J. Roentgenol., April 1, 2007; 188(4): 1081 - 1088. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. W. A. Pfirrmann, B. Mengiardi, C. Dora, F. Kalberer, M. Zanetti, and J. Hodler Cam and Pincer Femoroacetabular Impingement: Characteristic MR Arthrographic Findings in 50 Patients Radiology, September 1, 2006; 240(3): 778 - 785. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
