| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiology, Indiana University Medical Center, Indianapolis, IN
46202-5253.
2 Present address: Department of Radiology, University of Wisconsin Hospital,
Clinical Science Center, E3/311, 600 Highland Ave., Madison, WI
53792-3252.
3 Department of Radiology, Thomas Jefferson University Hospital, Philadelphia,
PA 19107.
Received August 3, 2004; accepted after revision October 6, 2004.
Address correspondence to R. Kijowski.
OBJECTIVE. This study was performed to determine whether a modified version of the classic STIR sequence provides similar information about the integrity of the rotator cuff tendon as the commonly used fat-suppressed T2-weighted fast spin-echo sequence.
SUBJECTS AND METHODS. Sixty-one consecutive MRI examinations of the shoulder in 57 patients were performed using a coronal oblique T1-weighted spin-echo sequence, a modified version of the STIR sequence, and a fat-suppressed T2-weighted fast spin-echo sequence. Three reviewers independently assessed the rotator cuff tendon using the coronal oblique modified inversion recovery sequence and T1-weighted spin-echo sequence. After a minimum of 4 weeks, reviewers assessed the rotator cuff tendon using the fat-suppressed T2-weighted fast spin-echo sequence and T1-weighted spin-echo sequence. The kappa statistic was used to measure the degree of concordance between interpretations when each sequence was used independently. The conditional probability that a full- and a partial-thickness tear would be diagnosed on both sequences was calculated. Image quality was assessed in a side-by-side comparison.
RESULTS. The overall weighted kappa score was 0.82, which indicates excellent concordance between the two sequences. If a full-thickness tear of the rotator cuff tendon was found on the fat-suppressed T2-weighted fast spin-echo sequence, there was a 94.1% probability that the same conclusion would be reached using the modified inversion recovery sequence. If a partial-thickness tear was found on the fat-suppressed T2-weighted fast spin-echo sequence, there was an 80.3% probability that the same conclusion would be reached with the modified inversion recovery sequence. Fat suppression in the modified inversion recovery sequence was superior to that in the T2-weighted fast spin-echo sequence in 26–39% of the examinations.
CONCLUSION. The modified inversion recovery sequence and fat-suppressed T2-weighted fast spin-echo sequence provide similar information about the integrity of the rotator cuff tendon.
Fast spin-echo MRI is widely used in the evaluation of the rotator cuff tendon. Fast spin-echo images provide contrast similar to that of conventional spin-echo images and have the advantage of a shorter acquisition time. Various studies have shown the ability of fast spin-echo sequences to detect full- and partial-thickness tears of the rotator cuff tendon with sensitivity and specificity similar to those of conventional spin-echo sequences [1–3].
One inherent disadvantage of the fast spin-echo technique is that fat remains bright on these images, which makes it difficult to detect subtle defects of the rotator cuff tendon and to identify small amounts of fluid within the subacromial and subdeltoid bursa. For this reason, chemical fat suppression often is applied to fast spin-echo sequences. The addition of chemical fat suppression increases the conspicuousness of fluid within small defects of the rotator cuff tendon. Fat suppression also reduces artifacts from respiratory motion and chemical shift misregistration that may degrade image quality. Various studies have shown that fat-suppressed fast spin-echo sequences are as good as or better than conventional and fast spin-echo sequences without fat saturation in the detection of rotator cuff tendon tears [2, 4–6].
However, fat-suppressed fast spin-echo sequences have some disadvantages. Chemical fat suppression results in images with lower signal-to-noise ratios than non-fat-suppressed images. In addition, uneven fat suppression is common because of inherent inhomogeneities in both the static and the radiofrequency magnetic fields. Chemical fat suppression also is difficult to perform on low-magnetic-field-strength scanners because of the closer resonance frequencies of fat and water on these systems [7, 8].
Inversion recovery sequences can be used to suppress the signal intensity of fat while avoiding some of the disadvantages associated with chemical fat saturation. These sequences are performed with short inversion times to eliminate signal from adipose tissue and are commonly referred to as STIR sequences. STIR sequences have the advantage of increasing the relative signal intensity of fluid as a result of the additive T1 and T2 contrast effect. This allows STIR images to have greater contrast between fluid and other tissues than fat-suppressed T2-weighted fast spin-echo images. A disadvantage of STIR sequences is that the images tend to have relatively poor signal-to-noise ratios.
A modified version of the classic STIR sequence is used at our institution to evaluate the rotator cuff tendon. In this modified inversion recovery sequence, the inversion time is decreased from 150 to 110 msec at a magnetic field strength of 1.5 T. This modification increases the signal intensity of adipose tissue and ultimately leads to an increased signal-to-noise ratio of the final image. The increased signal-to-noise ratio allows better visualization of anatomic details, which is important when evaluating subtle abnormalities of the rotator cuff tendon. The lower inversion time also results in a shorter acquisition time for the modified inversion recovery sequence relative to the classic STIR sequence.
The modified inversion recovery sequence also uses a higher TR and TEeff than the classic STIR sequence, which increases the T2 weighting of the final image. The additive T1 and T2 contrast effect increases the conspicuousness of fluid in the modified inversion recovery sequence and potentially improves visualization of pathologic abnormities in the rotator cuff tendon. We performed this study to determine whether the modified inversion recovery sequence can provide similar information about the integrity of the rotator cuff tendon as the fat-suppressed fast spin-echo sequence in patients presenting with shoulder pain.
Subjects and Methods
The study group consisted of 57 consecutive patients who underwent MRI of the shoulder for the evaluation of shoulder pain. The examinations were performed between February 1, 2003, and April 4, 2003. The age of the patients ranged from 22 to 77 years with an average age of 59 years. The study group was composed of 28 men and 29 women. Four patients had bilateral shoulder examinations performed during a single imaging session. Approval from our institutional review board was not required to perform the study.
MRI of the shoulder was performed on a 1.5-T field strength magnet (LX Horizon, GE Healthcare) using a dedicated receive-only phased-array shoulder coil. A T1-weighted spin-echo sequence, a modified inversion recovery sequence, and a fat-suppressed T2-weighted fast spin-echo sequence were performed on all 61 shoulders as part of the routine shoulder examination. All sequences were performed in the coronal oblique plane with a 15-cm field of view, a 256 x 192 matrix size, a 3.5-mm slice thickness, a 0.5-mm gap between slices, and 2 excitations.
The T1-weighted spin-echo sequence was performed with a TR of 400 msec and a TEeff of 20 msec. The modified inversion recovery sequence was performed with a TR of 3,500 msec, a TEeff of 50 msec, an inversion time of 110 msec, and an echo-train length of 8. The fat-suppressed T2-weighted fast spin-echo sequence was performed with a TR of 3,000 msec, a TEeff of 80 msec, and an echo-train length of 8. A frequency selective chemical presaturation pulse (ChemSat, GE Healthcare) was used in this sequence to suppress signal from adipose tissue. The imaging time for the modified inversion recovery sequence was approximately 3.5 min, and the imaging time for the fat-suppressed T2-weighted fast spin-echo sequence was approximately 3 min excluding shim time.
The T1-weighted spin-echo sequence and modified inversion recovery sequence of all 61 shoulder examinations were reviewed independently by three musculoskeletal radiologists at our institution (reviewers 1, 2, and 3). After an interval of at least 4 weeks to prevent recall bias, the T1-weighted spin-echo sequence and fat-suppressed T2-weighted fast spin-echo sequence of all 61 shoulder examinations were reviewed independently by the same musculoskeletal radiologists. After another interval of at least 4 weeks, the same musculoskeletal radiologists reinterpreted all 61 shoulder examinations to determine the intraobserver variability in evaluating the integrity of the rotator cuff tendon. In one half of the cases, the reviewers used the T1-weighted spin-echo sequence and the modified inversion recovery sequence to evaluate the rotator cuff tendon. In the other half of the cases, the reviewers used the T1-weighted spin-echo sequence and fat-suppressed T2-weighted fast spin-echo sequence to evaluate the rotator cuff tendon.
Throughout the study, the radiologists were unaware of the name of the patient whose examination they were reviewing and of any previous interpretation of the examination. No attempt was made to hide the identity of the imaging sequences used. The modified inversion recovery sequence and fat-suppressed T2-weighted fast spin-echo sequence could easily be distinguished from one another by the reviewers, all of whom are experienced with both imaging sequences. The reviewers used a standardized score sheet to grade the integrity of the rotator cuff tendon. The grading scale that was used in this study to evaluate the rotator cuff tendon (Table 1) is similar to those used in previous studies [1–5].
|
The quality of the fat-suppressed T2-weighted fast spin-echo images and the modified inversion recovery images in each of the 61 shoulder examinations was assessed independently by the same three musculoskeletal radiologists in a side-by-side comparison. They used the following criteria to assess image quality: motion artifact, homogeneity of fat suppression, visualization of the rotator cuff tendon, and visualization of any abnormality seen in the rotator cuff tendon. The modified inversion recovery sequence was scored as inferior to, equal to, or superior to the fat-suppressed T2-weighted fast spin-echo sequence for each criterion. If a reviewer thought that one of the sequences provided better visualization of the rotator cuff tendon or better visualization of any abnormality within the tendon, he was asked to explain why he came to that conclusion.
All examinations in which the three reviewers agreed that both the modified inversion recovery sequence and the fat-suppressed T2-weighted fast spin-echo sequence showed homogeneous fat suppression and minimal motion artifact were identified. Ten of these examinations were chosen at random. The images obtained from those 10 examinations were stored in DICOM format with an associated image viewing program on a CD-ROM disc. This disc was sent to an outside institution where an additional musculoskeletal radiologist evaluated the image quality of both sequences for the 10 examinations. The radiologist was able to adjust the window and level settings of all images stored on the disc.
The following two criteria were used to assess image quality: visualization of the rotator cuff tendon and visualization of any abnormality seen in the rotator cuff tendon. The modified inversion recovery sequence was scored as inferior to, equal to, or superior to the fat-suppressed T2-weighted fast spin-echo sequence for each criterion. The purpose of this portion of the study was to obtain an unbiased opinion about the image quality of the modified inversion recovery sequence and the fat-suppressed T2-weighted fast spin-echo sequence from a musculoskeletal radiologist not associated with our institution. This radiologist routinely uses the fat-suppressed T2-weighted fast spin-echo sequence in his clinical practice to evaluate the rotator cuff tendon.
The kappa statistic was used to measure the concordance or strength of agreement between the tendon score given when the modified inversion recovery sequence was used to evaluate the rotator cuff tendon and the tendon score given when the fat-suppressed T2-weighted fast spin-echo was used to evaluate the rotator cuff tendon. The kappa statistic was used also to measure both interobserver agreement and intraobserver agreement of the tendon scores given when the same sequence was used to evaluate the rotator cuff tendon. Both the unweighted kappa statistic and the weighted kappa statistic were used to measure the concordance or strength of agreement between the tendon scores in this study.
The unweighted kappa statistic is most commonly used to measure the strength of agreement of a binary decision. The weighted kappa statistic is most commonly used to measure the strength of agreement of ordinal data with weights assigned to the magnitude of disagreement. When the unweighted kappa statistic was used, a perfect agreement between the two tendon scores was assigned a weight of 1 and any disagreement between the tendon scores was assigned a weight of 0. The magnitude of disagreement between the two tendon scores was not taken into account when using the unweighted kappa statistic.
For the weighted kappa statistic, perfect agreement between the two tendon scores was assigned a weight of 1, disagreement between a tendon score of zero and 1 was assigned a weight of 1, and disagreement between any other tendon scores was assigned a weight of 0. The magnitude of disagreement between a tendon score of zero and 1 was considered to be less than the magnitude of disagreement between any other tendon scores in the weighted kappa statistic. A statistical software program (SAS/STAT, version 9, SAS Institute) was used to calculate all kappa values with a 95% confidence interval. The agreement between the tendon scores was considered excellent if the kappa value was greater than 0.80, good if it was between 0.61 and 0.80, and moderate if it was between 0.41 and 0.60 [9–12].
The conditional probability, that if a reviewer gave a tendon score corresponding to a full- or partial-thickness tear of the rotator cuff tendon when using the fat-suppressed T2-weighted fast spin-echo sequence, he would give an identical tendon score when using the modified inversion recovery sequence, was calculated. The conditional probability, that if a reviewer gave a tendon score corresponding to a full- or partial-thickness tear of the rotator cuff tendon on the first interpretation of the examination, he would give an identical tendon score when using the same sequence on the second interpretation, was calculated. All conditional probabilities and their SEs were calculated using logistic regression models. The logistic regression models were performed using a statistical software program (SAS/STAT, version 9, SAS Institute). Hypothesis testing was performed to determine how these conditional probabilities differed from 100% [10].
Results
The unweighted and weighted kappa values measuring the degree of concordance between the tendon score given when the modified inversion recovery sequence was used and that given when the fat-suppressed T2-weighted fast spin-echo sequence was used to evaluate the rotator cuff tendon are listed in Table 2. The unweighted and weighted kappa values measuring interobserver agreement and intraobserver agreement of the tendon scores when the same sequence was used to evaluate the rotator cuff tendon are listed in Tables 3 and 4.
|
|
|
The conditional probability, that if a reviewer gave a tendon score corresponding to a full-thickness rotator cuff tear when using the fat-suppressed T2-weighted fast spin-echo sequence, he would give an identical tendon score when using the modified inversion recovery sequence, was 94.1% (Fig. 1A, 1B). This probability is statistically indifferent to 100% with a p value of 0.156. There were only two cases in which a full-thickness tear was diagnosed using one sequence but a partial-thickness tear was diagnosed using the second sequence. These cases are illustrated in Figures 2A, 2B, 2C, 2D and 3A, 3B. In both cases, a full-thickness tear was diagnosed using the modified inversion recovery sequence, but the fat-suppressed T2-weighted fast spin-echo sequence showed a small amount of remaining rotator cuff tendon attached to the humeral head, which led to a diagnosis of a high-grade partial-thickness tear. In one of these cases, significant motion artifact was present on the fat-suppressed T2-weighted fast spin-echo images despite multiple attempts to repeat the sequence. This motion artifact made it difficult to evaluate the rotator cuff tendon using the fat-suppressed T2-weighted fast spin-echo images.
|
|
|
|
|
|
|
|
The conditional probability, that if reviewer gave a tendon score corresponding to a partial-thickness rotator cuff tear when using the fat-suppressed T2-weighted fast spin-echo sequence, he would give an identical tendon score when using the modified inversion recovery sequence, was 80.3% (Fig. 4A, 4B). This probability is statistically less than 100% with a p value of 0.002.
|
|
The conditional probability, that if a reviewer gave a tendon score corresponding to a partial-thickness tear of the rotator cuff tendon on the first interpretation of the examination, he would give an identical tendon score when using the same sequence on the second interpretation, was 90.1%. This probability is statistically less than 100% with a p value of 0.048.
A comparative assessment of the image quality of the modified inversion recovery sequence and the fat-suppressed T2-weighted fast spin-echo sequence for the three reviewers at our institution is summarized in Table 5. Reviewer 1 thought that visualization of the rotator cuff tendon and any abnormality within the tendon was better in the modified inversion recovery sequence because of superior fat suppression in 8% of examinations and less motion artifact in 2% of examinations. Reviewer 2 thought that visualization of the rotator cuff tendon was better in the modified inversion recovery sequence because of superior fat suppression in 8% of examinations, less motion artifact in 16%, and better tissue contrast in 7%. Reviewer 2 thought that visualization of any abnormality in the rotator cuff tendon was better in the modified inversion recovery sequence because of superior fat suppression in 8% of examinations and less motion artifact in 16% of examinations. Reviewer 3 thought that visualization of the rotator cuff tendon and any abnormality in the tendon was better in the modified inversion recovery sequence because of superior fat suppression in 58% of examinations, less motion artifact in 32%, and better tissue contrast in 47%. A comparative assessment of the image quality of the modified inversion recovery sequence and the fat-suppressed T2-weighted fast spin-echo sequence for the reviewer at the outside institution is summarized in Table 6.
|
|
Discussion
MRI of the shoulder is used widely for the evaluation of the rotator cuff tendon. Although imaging protocols may vary from institution to institution, fat-suppressed T2-weighted fast spin-echo images in the coronal oblique plane is one of the most commonly used sequences for the evaluation of the rotator cuff tendon. Various studies have shown the ability of this sequence to reveal partial and complete tears of the rotator cuff tendon with high sensitivity and specificity [2, 4–6]. However, there are some inherent disadvantages of the chemical fat-suppression technique used in the acquisition of the fat-suppressed T2-weighted fast spin-echo sequence [7, 8].
The modified inversion recovery sequence is an alternative sequence that provides uniform suppression of signal from adipose tissue without some of the inherent disadvantages of chemical fat suppression. This study has shown that the modified inversion recovery sequence in the coronal oblique plane provides similar information about the integrity of the rotator cuff tendon as the fat-suppressed T2-weighted fast spin-echo sequence.
The observations made about the integrity of the rotator cuff tendon when using the coronal oblique modified inversion recovery sequence were identical to those made when using the coronal oblique fat-suppressed T2-weighted fast spin-echo sequence for most patients in the study. The kappa score was used as a statistical measure of the concordance of the tendon scores given when the rotator cuff tendon was evaluated using the two different sequences. When the data of all three reviewers in the study were combined, the unweighted kappa score was 0.71. This value indicates good concordance between the fat-suppressed T2-weighted fast spin-echo sequence and the modified inversion recovery sequence. When a weighted kappa statistic was used to analyze the data, the kappa score increased to a value of 0.82, which indicates excellent concordance between the two sequences.
We believe that the weighted kappa statistic is a more clinically relevant method of measuring concordance between the tendon scores in the circumstance when two sequences are used independently to evaluate the rotator cuff tendon. When the weighted kappa statistic was used, the tendon scores zero and 1 were combined to produce a single tendon score. This single tendon score represented all rotator cuff tendons that did not have a partial- or full-thickness tear. All patients with rotator cuff tendons in this category would likely be managed similarly with conservative therapy. Consequently, we believe that distinguishing a normal rotator cuff tendon from a rotator cuff tendon with tendinopathy on an imaging study is not clinically significant. As long as a tear of the rotator cuff tendon is excluded, most patients with the signs and symptoms of rotator cuff disease, at least at our institution, would be treated with conservative therapy regardless of whether an imaging study shows a normal rotator cuff tendon or rotator cuff tendinopathy [13–16].
The modified inversion recovery sequence and the fat-suppressed T2-weighted fast spin-echo sequence had a slightly lower degree of concordance for the diagnosis of a partial-thickness rotator cuff tear than for the diagnosis of a full-thickness tear. This lower degree of concordance may be partially explained by the relatively low inter- and intraobserver agreement that occurs when MRI is used to diagnose partial-thickness rotator cuff tears. One previous study showed interobserver agreement was low when multiple radiologists evaluated the rotator cuff tendon in patients with surgically proven partial-thickness rotator cuff tendon tears [17]. Our study results show that the intraobserver agreement for diagnosing a partial-thickness tear of the rotator cuff tendon on both the modified inversion recovery sequence and the fat-suppressed T2-weighted fast spin-echo sequence was less than the intraobserver agreement for diagnosing a full-thickness tear. If there is less than an excellent agreement between different radiologists and the same radiologist on different occasions in detecting a partial-thickness tear of the rotator cuff tendon using the same sequence, it is unlikely for there to be excellent agreement between a single radiologist using different sequences in the detection of these tears.
This study has clearly shown that the modified inversion recovery sequence and the fat-suppressed T2-weighted fast spin-echo sequence provide similar information about the integrity of the rotator cuff tendon in most patients. However, the modified inversion recovery sequence has certain advantages over the fat-suppressed T2-weighted fast spin-echo sequence. The most striking advantage is that the modified inversion recovery sequence provided more homogeneous fat suppression than the fat-suppressed T2-weighted fast spin-echo sequence on our 1.5-T MRI system. The images obtained with the modified inversion recovery sequence showed homogeneous fat suppression in all examinations performed in the study. In addition, the quality of fat suppression in the modified inversion recovery sequence was judged to be superior to the quality of fat suppression in the fat-suppressed T2-weighted fast spin-echo sequence in 26–39% of examinations. The main reason for this difference is that the two sequences use different methods to suppress signal from adipose tissue. The fat-suppressed T2-weighted fast spin-echo sequence uses chemical fat suppression, which is dependent on the different resonance frequencies for protons in fat and water. Uneven fat suppression is occasionally encountered when using this sequence because of inherent inhomogeneities in both the static and radiofrequency magnetic fields of the MRI scanner. Homogeneous chemical fat suppression is especially difficult to achieve when imaging body parts such as the shoulder that are not located near the isocenter of the scanner. On the other hand, for the modified inversion recovery sequence, suppression of signal from fat is based on relaxation time rather than resonance frequency. As a result, the suppression of fat in the modified inversion recovery sequence is influenced less strongly by inherent inhomogeneities in the magnetic fields of the MRI scanner.
In most cases, the nonuniform fat suppression of the fat-suppressed T2-weighted fast spin-echo images did not reduce the diagnostic quality of the images and merely consisted of poor fat suppression in the subcutaneous tissue of the shoulder. However, the chemical fat suppression failed completely in 8% of the examinations in the study. The absence of fat suppression in the T2-weighted fast spin-echo images in these examinations greatly reduced the reviewer's ability to evaluate the integrity of the rotator cuff tendon [4, 5].
Patient motion also appeared to be less severe on images obtained using the modified inversion recovery sequence. The modified inversion recovery sequence was found to have less motion artifact than the fat-suppressed T2-weighted fast spin-echo sequence in 10–30% of examinations performed in the study. The exact cause of this decreased patient motion artifact is unknown. Both sequences took about the same amount of time to perform. In addition, there are no inherent characteristics of chemical fat suppression that would make the fat-suppressed T2-weighted fast spin-echo sequence more susceptible to motion artifact. A possible explanation is that the modified inversion recovery sequence was performed before the fat suppressed T2-weighted fast spin-echo sequence in all patients. The modified inversion recovery sequence was the third sequence performed during the examination, whereas the fat-suppressed T2-weighted fast spin-echo sequence was the fourth sequence performed. Some patients may have experienced fatigue and had difficulty holding still as the examination progressed.
The modified inversion recovery sequence and the fat-suppressed T2-weighted fast spin-echo sequence appear equal in their ability to provide good visualization of the rotator cuff tendon and good visualization of any abnormality in the tendon. The image quality of the modified inversion recovery sequence was found to be superior in 8–48% of the patients in our study. However, in a large number of these cases, the superior image quality was attributed to the more homogeneous fat suppression and reduced patient motion artifact for the modified inversion recovery sequence. If these factors are eliminated, the quality of the two sequences was considered identical.
The modified inversion recovery sequence may have the slight advantage of having better tissue contrast than the fat-suppressed T2-weighted fast spin-echo sequence. In the modified inversion recovery sequence, the fat within the bone marrow of the humeral head and within the subacromial space is gray in contrast to the dark rotator cuff tendon separating these two structures. For the fat-suppressed T2-weighted fast spin-echo sequence, the fat within the humeral head and subacromial space and the rotator cuff tendon are both of equally low signal intensity. The effect of the greater tissue contrast of the modified inversion recovery sequence on perceived image quality is greatly reviewer-dependent. Two reviewers at our institution did not think that the greater tissue contrast of the modified inversion recovery sequence significantly increased its ability to show the rotator cuff tendon or any abnormality in the tendon. However, one reviewer at our institution and the reviewer at the outside institution thought that the greater tissue contrast of the modified inversion recovery sequence improved overall image quality in a significant number of the examinations when compared with the fat-suppressed T2-weighted fast spin-echo sequence.
To our knowledge, only two previous studies have been performed to investigate the ability of an inversion recovery sequence to reveal the integrity of the rotator cuff tendon. Shellock and associates [18] used a coronal oblique classic STIR sequence as part of their MRI protocol to evaluate the rotator cuff tendon in 47 patients on a 0.2-T imaging system. They found that their imaging protocol had a sensitivity of 89% and a specificity of 100% for detecting surgically proven tears in the rotator cuff tendon. Wright and associates [19] used multiple coronal oblique MRI sequences to evaluate the rotator cuff tendon of asymptomatic individuals on a 1.5-T imaging system. They found that the classic STIR sequence and the fat-suppressed T2-weighted fast spin-echo sequence with a TE value of 66 msec were the best sequences to accurately distinguish fluid signal intensity from intermediate increased signal intensity in the rotator cuff tendon of asymptomatic individuals [19].
One weakness of our study is the lack of surgical correlation, which could not be obtained for most of the patients because of the nature of our clinical practice. However, the main goal of the study was to assess the performance of the modified inversion recovery sequence relative to the fat-suppressed T2-weighted fast spin-echo sequence. The fat-suppressed T2-weighted fast spin-echo sequence was used as the gold standard in this study to assess the ability of the modified inversion recovery sequence to accurately show the integrity of the rotator cuff tendon. This gold standard obviously is not as definitive as the appearance of the rotator cuff at surgery. Nevertheless, various studies have shown the ability of the coronal oblique fat-suppressed T2-weighted fast spin-echo sequence to reveal the integrity of the rotator cuff tendon accurately when surgery was used as a gold standard. The sequence can detect the presence of full-thickness tears of the rotator cuff tendon with specificity values ranging from 87% to 99% and sensitivity values ranging from 85% to 100%. The sequence can depict partial-thickness tears of the rotator cuff tendon with specificity values ranging from 93% to 99%. The sensitivity of the coronal oblique fat-suppressed T2-weighted fast spin-echo sequence in revealing partial-thickness tears of the rotator cuff is more controversial. Although the sensitivity value was as low as 45% in one study, most studies showed sensitivity values ranging from 73% to 92%. When the rotator cuff tendon is being evaluated, the coronal oblique fat-suppressed T2-weighted fast spin-echo sequence, although far from perfect, is clearly the gold standard sequence with which all other sequences must be compared [3–6].
Another weakness of this study is that the fat-suppressed T2-weighted fast spin-echo sequence was performed using a TEeff of 80 msec. Some institutions use a lower TEeff, for example, 60 msec. In addition, the musculoskeletal radiologists were not blinded to the sequence type, which could have introduced bias. Finally, the order of obtaining the sequences was not random. For this reason, the overall image quality of the fat-suppressed T2-weighted fast spin-echo sequence, which was performed later during the course of the examination, could have been degraded by patient motion artifact related to fatigue.
In conclusion, the modified inversion recovery sequence and the fat-suppressed T2-weighted fast spin-echo sequence provided similar information about the integrity of the rotator cuff tendon in most patients in the study. This was especially true for patients with full-thickness tears of the rotator cuff tendon. The modified inversion recovery also had the advantage of more homogeneous suppression of signal from adipose tissue. This study has clearly shown that the coronal oblique modified inversion recovery sequence can be used as an alternative sequence to evaluate the rotator cuff tendon especially on imaging systems in which chemical fat suppression does not provide ideal suppression of signal from adipose tissue.
References
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
