| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Musculoskeletal Imaging |
1 From the Departments of Radiology (K.S., M.Z., J.H.) and Orthopedic Surgery (L.N.), Orthopedic University Hospital Balgrist, Zurich, Switzerland. Received November 20, 2002; revision requested January 21, 2003; final revision received May 9; accepted June 16. Address correspondence to K.S., Radiologie, Kantonsspital Luzern, 6000 Luzern 16, Switzerland (e-mail: klaus.strobel@ksl.ch).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: THI and conventional US images were obtained in 50 patients suspected of having rotator cuff lesions. Images were graded for visibility of acromioclavicular (AC) joint surfaces and rotator cuff tendon contours and substances: Grade 1 meant poor; grade 2, moderate; and grade 3, good visibility. Accuracy, sensitivity, and specificity of the diagnosis of AC joint osteoarthritis and accuracy of the diagnosis and rates of underestimation and overestimation of the extent of rotator cuff tears were evaluated. Visibility grade differences were evaluated with the Wilcoxon signed rank test. The McNemar test was used to assess differences in diagnoses. Magnetic resonance (MR) arthrography was the reference standard.
RESULTS: Mean visibility grades for readers 1 and 2, respectively, were as follows: for AC joint bone surfaces, 2.1 and 2.0 with THI and 1.7 (P = .010) and 1.7 (P = .16) with conventional US; for AC joint capsule surfaces, 2.2 and 1.9 with THI and 1.8 (P = .005) and 1.8 (P = .34) with US; for supraspinatus tendon contour, 2.6 and 2.2 with THI and 2.1 (P = .001) and 1.9 (P = .055) with US; for supraspinatus tendon substance, 2.2 and 1.9 with THI and 2.0 (P = .036) and 1.7 (P = .070) with US; for subscapularis tendon contour, 2.4 and 2.1 with THI and 2.2 (P = .07) and 2.0 (P = .25) with US; and for subscapularis tendon substance, 1.8 and 1.7 with THI and 2.0 (P = .86) and 1.7 (P = .91) with US. Diagnostic accuracies for the supraspinatus tendon for readers 1 and 2, respectively, were 84% and 74% with THI and 86% and 70% with US (P > .99 for both readers). Corresponding values for the subscapularis tendon were 78% and 72% with THI and 64% (P = .27) and 52% (P = .006) with US.
CONCLUSION: Joint and tendon surface visibility improves with THI, as compared with the visibility achieved with conventional US. THI is superior to conventional US for diagnosis of subscapularis tendon abnormalities.
© RSNA, 2003
Index terms: Magnetic resonance (MR), arthrography, 41.121411, 41.121415, 41.121416, 41.12143 • Shoulder, arthritis, 41.772 • Shoulder, injuries, 41.4813 • Shoulder, US, 41.12989, 41.1299 • Ultrasound (US), harmonic study, 41.12989, 41.1299
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
The role of THI in imaging the musculoskeletal system is less well known, although it may have the potential to help solve diagnostic problems, such as difficulty differentiatingpartial from full-thickness rotator cuff tears (16–19). The purpose of this study was to compare THI of the shoulder with conventional US in patients suspected of having rotator cuff lesions.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
Thus, a total of 50 patients (23 women with a mean age of 51.4 years [age range, 28–73 years] and 27 men with a mean age of 54.5 years (age range, 29–83 years) were included in the study. In nine of these patients, surgery was performed after imaging. Seven patients with full-thickness tears underwent rotator cuff reconstruction (five arthroscopic and two open procedures), and two patients with partial tears underwent arthroscopic débridement. The study was approved by the institutional review board of our institution, and informed consent was obtained from all patients.
US Examinations
US examinations (ie, conventional US and THI) were performed immediately before MR arthrography by one of two musculoskeletal radiologists (M.Z., J.H.), each of whom has more than 10 years of experience in shoulder US. A 7.5–9.0-MHz multifrequency linear probe (Elegra; Siemens Medical Solutions, Erlangen, Germany) was used. The patient sat on a chair with the examiner standing behind him or her. The US examination was standardized according to a previously described technique (20) and included imaging of the acromioclavicular (AC) joint (with the transducer perpendicular to the joint space), the subscapularis (SSC) and supraspinatus (SSP) tendons (with images obtained on the long and short axes of the tendon), and the infraspinatus tendon (with images obtained on the long axis). The long biceps tendon and the SSP and infraspinatus muscles were imaged but not further evaluated in this study.
The patient was asked to place his or her hand on the ipsilateral thigh for the examination. However, for better exposure of the tendon from underneath the acromion, the SSC tendon was evaluated with the arm externally rotated and the SSP tendon was evaluated with the wrist behind the patient’s back. Each image was obtained by using both pulse-inversion THI and conventional US without moving the transducer between image acquisitions. The order of the image acquisitions—that is, either THI first or conventional US first—was determined randomly. The imaging parameters (eg, field of view and focus position) were kept constant while switching between THI and conventional US, with the exception that the image gain was adjusted individually. The images were saved onto a picture archiving and communications system, or PACS (Image Devices, Idstein, Germany), for further review.
MR Arthrography
Use of MR arthrography with an intraarticular gadolinium-based contrast agent has been accepted by our hospital’s institutional review board and by the responsible state agency. MR arthrography was performed in 32 patients with a 1.5-T MR imaging unit (Symphony; Siemens Medical Solutions) and in 18 patients with a 1.0-T unit (Expert; Siemens Medical Solutions) after fluoroscopically guided injection of 10 mL of gadopentetate dimeglumine (2 mmol/L Magnevist; Berlex, Wayne, NJ). The type of imaging unit used depended on availability. A dedicated receive-only surface coil was used with both magnets.
The following examinations were performed with the 1.5-T MR imaging unit: Oblique-coronal proton density–weighted and T2-weighted fat-saturated turbo spin-echo images were obtained by using a repetition time msec/echo time msec of 3,300/14–95, a 4-mm section thickness, a 16-cm field of view, and a 512 x 256 matrix. Oblique-coronal T1-weighted fat-saturated turbo spin-echo images were obtained by using 777/12, a 3-mm section thickness, a 16-cm field of view, and a 512 x 256 matrix. Oblique-sagittal T1-weighted spin-echo images were obtained by using 600/12, a 4-mm section thickness, a 16-cm field of view, and a 512 x 256 matrix. Transverse T1-weighted spin-echo images were obtained by using 600/12, a 3-mm section thickness, a 16-cm field of view, and a 512 x 256 matrix.
The following examinations were performed with the 1.0-T MR imaging unit: Oblique-coronal intermediate- and T2-weighted turbo spin-echo images were obtained by using 3,500/16–98, a 3-mm section thickness, a 16-cm field of view, and a 256 x 256 matrix. Oblique-coronal T1-weighted fat-saturated turbo spin-echo images were obtained by using 800/20, a 4-mm section thickness, a 16-cm field of view, and a 256 x 256 matrix. Oblique-sagittal T1-weighted turbo spin-echo images were obtained by using 700/12, a 5-mm section thickness, a 16-cm field of view, and a 256 x 256 matrix. Transverse T1-weighted spin-echo images were obtained by using 580/20, a 4-mm section thickness, a 16-cm field of view, and a 512 x 512 matrix.
The original findings on the MR arthrograms were used as reference standards for evaluation of the diagnostic performance of US. The MR images were interpreted by four musculoskeletal radiologists from the same institution, who have at least 5 years (5, 5, 10 [M.Z.], and 14 [J.H.] years) of experience in MR arthrography of the shoulder. The MR images were not evaluated by the same radiologist who performed the US examinations.
Review of US Images
After the end of the data acquisition period, all US images were independently reviewed by both staff radiologists who had performed the US examinations. The PACS workstation (ID.Station Report; Image Devices) was used for this purpose. To blind the readers, the patient data and imaging parameters were masked during the review. In addition, there was a time interval of 2–12 weeks between the US and MR arthrographic examinations on one hand and the readout of the US images on the other hand. One hundred sets of US images (50 harmonic, 50 nonharmonic [ie, conventional]) were presented to the readers; the images were intermixed over several sessions. The readers knew that the patients were suspected of having a rotator cuff tear, but they were otherwise blinded with regard to the clinical data.
To evaluate the quality of images depicting the AC joint bone surface and joint capsule, as well as the contour (ie, demarcation of the tendon from the adjacent tissue layers) and substance (ie, visibility of the fibrillar pattern) of the tendons, a subjective three-grade scale was used: Grade 1 indicated poor visibility—that is, the surfaces were poorly demarcated or the fibrillar pattern was not differentiated; grade 2, moderate visibility—that is, the surfaces were partially demarcated or the fibrillar pattern was partially differentiated; and grade 3, good visibility—that is, the surfaces were well demarcated or the fibrillar pattern was well differentiated throughout the structure.
The following diagnoses were made: AC joint osteoarthritis (ie, osteophytes, capsular hypertrophy, and/or bone irregularities) or SSP tendon or SSC tendon tear (partial or full-thickness tear, depending on the presence of a hypoechoic defect and/or a loss of the outer tendon convexity, or the nonvisualization of the tendon). The confidence in each assigned diagnosis was given one of three ratings: 1, meaning not sure; 2, meaning moderately sure; or 3, meaning sure of diagnosis.
Statistical Analyses
Differences in visibility grades and confidence ratings were evaluated with the Wilcoxon signed rank test. The McNemar test was used to assess differences in diagnoses. P < .05 was considered to indicate a significant difference. To assess interobserver agreement, 
values were calculated.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
|
The contour and substance of the SSC tendons were slightly more visible with THI than with conventional US for both readers, although the differences between the two US techniques were not significant. No significant differences in grades of infraspinatus tendon visibility were observed.
Diagnostic Accuracy and Interobserver Agreement
Diagnostic accuracy results are summarized in Table 2. Thirty-eight (76%) patients had osteoarthritis of the AC joint. The accuracies of conventional US and THI in the diagnosis of AC joint osteoarthritis were nearly the same: For reader 1, the accuracy of both THI and conventional US was 82%. For reader 2, the accuracies of THI and conventional US were 72% and 70%, respectively. The sensitivities and specificities in the diagnosis of AC joint osteoarthritis were as follows: for reader 1, 79% and 92%, respectively, with THI and 82% and 83%, respectively, with conventional US; and for reader 2, 74% and 67%, respectively, with THI and 68% and 75%, respectively, with conventional US. The accuracies of conventional US and THI did not differ significantly (reader 1, P = .79; reader 2, P > .99).
|
|
|
|
|
|
|
|
|
|
|
|
|
values were as follows: for AC joint assessment, 0.61 with THI and 0.46 with conventional US; for SSP tendon assessment, 0.67 with THI and 0.58 with conventional US; and for SSC tendon assessment, 0.51 with THI and 0.73 with conventional US (21). Agreement data for the infraspinatus tendons were not calculated because there were only six infraspinatus tendon lesions—three full-thickness and three partial tears. |
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
The use of THI in abdominal US leads to altered diagnostic accuracy and therapeutic consequence: In a study involving the depiction of liver lesions, THI yielded additional information for 14 (29%) of 48 patients and its use led to altered treatment for five patients (10%) (1). In gallbladder and bile duct US, use of THI leads to improved visibility of the normal anatomy and improved rates of detecting intraluminal masses (9,11). In obese patients and other patients who are difficult to image with conventional techniques, use of THI results in markedly improved contrast resolution (2). In vascular US, the intima-media of the internal carotid artery is better visualized with THI.
The inter- and intraobserver variability encountered with THI is significantly reduced compared with the variability encountered with conventional US (14). In breast US, use of THI improves lesion contrast and evaluation of margins (13). In one utility study of phase-inversion THI, only 13 of 200 examinations were examinations of superficial vascular or musculoskeletal structures (7). Otherwise, published experiences with THI of the musculoskeletal system appear to be limited, and so far, to our knowledge, in no formal study has it been determined if the accuracy or image quality is improved with this technique.
MR arthrography may not be a perfect reference standard, but it is a reasonable one. MR arthrographic investigations have revealed sensitivities in the range of 84%–100% and specificities in the range of 93%–99% for the diagnosis of full-thickness tears (22–25). Use of MR arthrography leads to improved depiction of partial articular cuff tears compared with use of conventional MR imaging (26), and when combined with fat suppression, it provides excellent results in the depiction of both full-thickness and partial tears (27).
In a study including both MR arthrography and US, Ferrari et al (19) found that conventional US yielded reliable results for the detection of large full-thickness SSP tears (in eight of eight cases) but had decreased sensitivity in the detection of smaller lesions (in 27 [61%] of 44 patients). In that study, US led to underestimation of the extent of SSP tendon tears in 13 cases (30%) and to overestimation in one case (2%). In that same study, the MR arthrographic diagnoses were correct in 43 cases (98%) (19). In two other studies in which findings were correlated with surgical results, the sensitivities and specificities of US were 57% and 95%, respectively, and 76% and 94%, respectively (28,29).
When our results are compared with those of the previously published investigations, two factors that influence diagnostic performance should be taken into account: (a) The previous evaluations were based on a selection of static images. In contrast, US examinations are performed dynamically, and this may help improve diagnostic accuracy. (b) The relatively large proportion of partial tears in this study may have negatively influenced diagnostic performance.
Our study results support the tendency that US findings are underestimations of the extent of rotator cuff tears: Underestimation occurred in 10%–36% of cases, and overestimation in 4%–12% of cases. The results indicate that subjectively better visibility of the rotator cuff tendons (eg, the SSP tendons seen by reader 1) with THI does not automatically mean higher accuracy in the detection of tendon lesions. On the other hand, significantly higher accuracy in the detection of tendon lesions (eg, the SSC tendon lesions seen by reader 2) is not based on substantially better visibility of the tendon structures.
In conclusion, THI appears to be superior to conventional US in the examination of patients suspected of having rotator cuff tears, although differences between the two techniques are not as impressive as those reported between conventional abdominal US and THI. The visibility of joint and tendon surfaces generally is improved with use of THI as compared with the visibility of these structures at conventional US. THI is superior to conventional US in the diagnosis of SSC tendon abnormalities.
|
FOOTNOTES |
|---|
Author contributions: Guarantors of integrity of entire study, K.S., J.H.; study concepts, M.Z.; study design, K.S., M.Z., J.H.; literature research, K.S, M.Z.; clinical studies, all authors; data acquisition, K.S., M.Z., J.H.; data analysis/interpretation, all authors; statistical analysis, K.S., M.Z., J.H.; manuscript preparation, K.S.; manuscript definition of intellectual content, all authors; manuscript editing, J.H.; manuscript revision/review and final version approval, all authors
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. H. Cha, W. K. Moon, N. Cho, S. M. Kim, S. H. Park, B.-K. Han, Y. H. Choe, J. M. Park, and J.-G. Im Characterization of Benign and Malignant Solid Breast Masses: Comparison of Conventional US and Tissue Harmonic Imaging Radiology, December 1, 2006; 242(1): 63 - 69. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Mesurolle, H. J. S. Bining, M. E. Khoury, A. Barhdadi, and E. Kao Contribution of tissue harmonic imaging and frequency compound imaging in interventional breast sonography. J. Ultrasound Med., July 1, 2006; 25(7): 845 - 855. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Herold, M. Bachthaler, O. W. Hamer, R. Hente, S. Feuerbach, C. Fellner, M. Strotzer, M. Lenhart, and C. Paetzel Indirect MR Arthrography of the Shoulder: Use of Abduction and External Rotation to Detect Full- and Partial-Thickness Tears of the Supraspinatus Tendon Radiology, July 1, 2006; 240(1): 152 - 160. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Papatheodorou, P. Ellinas, F. Takis, A. Tsanis, I. Maris, and N. Batakis US of the Shoulder: Rotator Cuff and Non-Rotator Cuff Disorders RadioGraphics, January 1, 2006; 26(1): e23 - e23. [Abstract] [Full Text]
|
|
|
|
|
|
J. B. Moosikasuwan, T. T. Miller, and B. J. Burke Rotator Cuff Tears: Clinical, Radiographic, and US Findings RadioGraphics, November 1, 2005; 25(6): 1591 - 1607. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
