HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by De Smet, A. A. Articles by Tuite, M. J. Search for Related Content PubMed PubMed Citation Articles by De Smet, A. A. Articles by Tuite, M. J. Social Bookmarking                      What's this? Hotlight (NEW!) What's Hotlight?

Use of the "Two-Slice-Touch" Rule for the MRI Diagnosis of Meniscal Tears

¹ Both authors: Department of Radiology, University of Wisconsin Medical School and Hospital, Radiology - E3/311, 600 Highland Ave., Madison, WI 53792.

Received August 4, 2005; accepted after revision August 31, 2005.

 
Address correspondence to A. A. De Smet.

Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
OBJECTIVE. We reviewed our experience with knee MRI to determine if using the "two-slice-touch" rule increased our positive predictive value for diagnosing meniscal tears. This rule classifies a meniscus as torn if there are two or more MR images with abnormal findings and as possibly torn if there is only one MR image with an abnormal finding. We also compared our sensitivity and specificity using fast spin-echo imaging with previously reported studies.

MATERIALS AND METHODS. We reviewed the medical records of 174 patients who had knee MR examinations and correlative knee arthroscopy to determine our accuracy when we diagnosed menisci as torn, possibly torn, or intact using knee arthroscopy as the gold standard.

RESULTS. Our positive predictive value increased from 91% to 94% (p = 0.37) for medial meniscal tears and from 83% to 96% (p = 0.02) for lateral meniscal tears when using the two-slice-touch rule compared with the standard criterion of diagnosing a meniscus as torn if one or more images are abnormal. When using the standard criterion, we had 95% sensitivity and 85% specificity for diagnosing medial meniscal tears and 77% sensitivity and 89% specificity for diagnosing lateral meniscal tears.

CONCLUSION. When using the two-slice-touch rule, we had an increased positive predictive value for diagnosing menisci as torn, which was statistically significant for the lateral meniscus. Our accuracy using the standard criterion with fast spin-echo imaging was comparable to that reported in previous studies with spin-echo imaging.

Keywords: biomedical statistics • knee • meniscal tears • MRI • musculoskeletal imaging

Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
In the past 20 years, many authors have studied the accuracy of MRI for diagnosing meniscal tears [1]. However, only one previous study has reported that false-positive diagnoses of a meniscal tear could be reduced by noting the number of abnormal findings on meniscal images on the MR examination [2]. In that study, a meniscus was found to be highly likely torn if two or more MR images of a meniscus had distortion or signal to the meniscal surface, and considerably less likely to be torn if only one MR image of a meniscus showed an abnormal finding [2]. These findings have been presented as the "two-slice-touch" rule at national meetings (De Smet AA, presented at the 2005 annual meeting of the Radiological Society of North America). However, this concept of diagnosing a meniscal tear as torn or possibly torn based on the number of abnormal MR images has not been used in any other study on the MR diagnosis of meniscal tears and has not been validated when used prospectively.

We have used the two-slice-touch rule for diagnosing meniscus tears on knee MRI for the past 10 years. Four years ago, we changed our MRI sequences from conventional spin echo to fast spin echo when we assessed patients for meniscal tears. Although several authors have reported comparable accuracy for diagnosing meniscal tears when using conventional and fast spin-echo imaging, a recent study found a significantly lower accuracy when using fast spin-echo imaging [3-5].

Because of this controversy regarding fast spin-echo imaging and the lack of a prospective study confirming the usefulness of the two-slice-touch rule, we reviewed our experience in diagnosing meniscal tears on MRI. We now report our results when using the two-slice-touch rule to diagnose a meniscus as torn, possibly torn, or not torn. We also compare our results using fast spin-echo imaging with those reported using both conventional and fast spin-echo imaging.

Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
The study was performed in compliance with Health Insurance Portability and Accountability Act (HIPAA) regulations and with approval from our institutional review board. A waiver of informed consent was obtained from our institutional review board before performing the study.

Patient Selection and Medical Records Review
We reviewed the medical records of 561 consecutive patients who had knee MRI examinations at our institution from July through October 2003. We noted in how many menisci we diagnosed a possible tear. We then determined the accuracy of our MR interpretations using the results of knee arthroscopy as the gold standard in all 174 patients who had correlative arthroscopic surgery at our institution after their knee MRI examination and who had not had prior meniscal surgery. The knee arthroscopies were performed by any one of three fellowship-trained orthopedic surgeons specializing in sports medicine and arthroscopy. Each arthroscopist had 7 or more years of experience as a faculty member at an academic medical center.

MRI
The MRI examinations were interpreted by any one of seven fellowship-trained musculoskeletal radiologists. The MRI experience of these observers ranged from 1 to 20 years. By prior agreement within the musculoskeletal division, each meniscus was evaluated prospectively using the following rules: A meniscus was diagnosed as torn if meniscal distortion or intrameniscal signal contacting the meniscal articular surface was noted on two or more images for a given meniscus. These two images need not be consecutive and could be two coronal images, two sagittal images, or one coronal image and one sagittal image. A meniscus was diagnosed as possibly torn if only one MR image was abnormal and as not torn if no image showed distortion or definite signal contacting the surface. Identification of a displaced meniscal fragment was also considered to be a definite sign of a tear, although in such patients there is almost invariably meniscal distortion or meniscal signal to the surface on more than one image.

The interpreters of these MRI studies used the two indirect signs of a meniscal tear—disruption of the popliteomeniscal fascicle and posteromedial tibial bone bruise—to look more carefully for an adjacent meniscal tear but diagnosed a definitely torn or possibly torn meniscus on the basis of the two-slice-touch rule [6, 7].

All knee MRI examinations were performed using our standard protocol on a 1.5-T MR magnet (Signa, GE Healthcare) using a knee phased-array cylindric coil. All studies consisted of coronal fast spin-echo T1-weighted, coronal fat-saturated fast spin-echo proton density-weighted, sagittal fast spin-echo proton density-weighted, and sagittal fat-saturated fast spin-echo T2-weighted images.

A field of view of 14 cm, slice thickness of 3 mm with a 1.5-mm interslice gap, and a matrix of 256 x 192 were used for all four sequences, except that a matrix of 256 x 224 was used for the coronal T1-weighted images. The parameters were as follows for the coronal T1-weighted images: TR range/TE, 600-700/17; 1 signal average; and echo-train length, 3; for the coronal fat-saturated proton density-weighted images: 1,800-2,000/17; 1 signal average; and echo-train length, 4; for the sagittal proton density-weighted images: 2,000-2,200/17; 1 signal average; and echo-train length, 4; and for the sagittal T2-weighted fat-saturated sequences: 3,000-3,400/60; 1 signal average; and echo-train length, 6. The scanning time for each of the four sequences was between 2 and 3.5 minutes.

Data Analysis
Sensitivity, specificity, positive predictive value, and negative predictive value were calculated for the medial and lateral menisci in two ways: first, by considering all MRI diagnoses of a tear or possible tear as positive for a tear; and second, by excluding the cases of possible tears. By calculating our accuracy using both methods, we could compare changes in accuracy when the two-slice-touch rule was used. Chi-square analysis was used to compare the data for these two sets of accuracy calculations. A difference with a p value of less than 0.05 was considered statistically significant.

Our results were then compared with those of prior studies reporting the accuracy of diagnoses of meniscal tears using conventional spin-echo and fast spin-echo imaging. We calculated our accuracy considering menisci diagnosed as torn or possibly torn as indicating a tear because those prior studies used the presence of one or more MR images showing meniscal distortion or intrameniscal signal contacting the meniscal surface as evidence of a meniscal tear. We selected for comparison only studies that reported at least 100 patients with arthroscopic correlation. For series of fewer than 100 patients, significant variation can occur by random chance in the calculated accuracy values [8].

Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
For the 174 patients who underwent correlative knee arthroscopy, the MRI diagnoses are compared with the arthroscopic findings in Table 1. The MRI diagnosis of a possible tear was made in seven (4.0%) of 174 medial menisci. Tears were found at arthroscopy in three (43%) of these seven medial menisci diagnosed as possible tears. The MRI diagnosis of a possible tear was made in 11 (6.3%) of the 174 lateral menisci. Tears were found at arthroscopy in two (18%) of these 11 lateral menisci diagnosed as possible tears.

The data in Table 1 were used to calculate accuracy as given in Table 2. Because more than half of the menisci diagnosed as possible tears were found to be intact at arthroscopy, the specificity and positive predictive value increased for both medial and lateral menisci when only menisci with two or more abnormal images were considered to be torn. However, the differences were statistically significantly different only in the lateral meniscus.

In Table 3, our results are compared with those of nine prior studies in which more than 100 patients were evaluated using 1.5-T MR magnets followed by arthroscopic correlation [9-18].

Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

 
This study was undertaken to confirm that using the two-slice-touch rule increased our positive predictive value for diagnosing a meniscal tear, and to determine whether our accuracy using fast spin-echo imaging was comparable to that previously reported.

With regard to our first objective, we found a significantly increased positive predictive value when we diagnosed a lateral meniscus as definitely torn on the basis of two or more abnormal images when compared with diagnosing a tear on the basis of one or more abnormal images (Table 2). Although the finding was not statistically significantly different, we also had fewer false-positive diagnoses of medial meniscal tears when we diagnosed a medial tear as torn only in the presence of two or more abnormal MR images (Table 2).

Our results confirm a prior study in which the authors retrospectively calculated a positive predictive value of 94% for medial menisci and 90% for lateral menisci when menisci were diagnosed as torn only when two or more images showed abnormal findings [2]. We had similar results, with a positive predictive value of 94% for tears of the medial meniscus and 96% for tears of the lateral meniscus (Table 2). In the earlier study, 55% of medial menisci and 30% of lateral menisci with one abnormal MR image were found to be torn at arthroscopy [2]. We had similar results in that a tear was found at arthroscopy in only 43% of the medial and 18% of the lateral menisci that had one abnormal MR image.

We emphasize that the use of the two-slice-touch rule does not change the criteria for the MR diagnosis of a meniscal tear. The criteria for diagnosing meniscal tears with knee MRI have not changed since the original large patient series reported in the late 1980s [9, 10]. As described in a recent review, these criteria are intrameniscal signal contacting the articular surface of the meniscus, distortion of meniscal shape, and identification of a displaced meniscal fragment [1].

The advantage of the two-slice-touch rule is that it identifies a small subset of patients with signal to the surface on only a single image in whom a meniscal tear is less likely. We diagnosed 4% of the medial and 6.3% of the lateral menisci as possibly torn. If we had used the standard criterion of a single abnormal image for diagnosing a meniscus as torn and diagnosed these menisci as torn, our positive predictive value would have been lower. Only one retrospective study before ours has used the criterion that a definite tear is diagnosed only when two or more MR images are abnormal [11]. That study considered a tear to be present when there were two or more consecutive abnormal images, but it did not discuss how menisci with one definitely abnormal image were classified [11]. We cannot compare our findings in detail with those of that study because its focus was the diagnosis of displaced fragments, and there are limited data on the accuracy of diagnosing tears without displaced fragments [11].

The sports medicine physicians at our institution have expressed how much they value our high positive predictive value for the diagnosis of a meniscal tear when using the two-slice-touch rule. Also, by using the category of a possible tear, we give them more flexibility in managing these patients than if we diagnosed their menisci as definitely torn. Because these patients have an MRI abnormality on only one image, they often have only a small tear if a tear is present. Therefore, conservative treatment rather than surgical treatment can be used if clinically appropriate.

With regard to the second purpose of this study, we found that our sensitivity and specificity for diagnosing meniscal tears were comparable to those previously reported using both conventional and fast spin-echo imaging (Table 3). Our sensitivity of 95% for diagnosing medial meniscal tears is at the upper end of reported values for nine large series in which patients had MRI studies using high-field-strength 1.5-T magnets. Our sensitivity of 77% for diagnosing lateral meniscal tears was in the middle of the reported range of 70-92% (Table 3).

Although a previous article comparing conventional spin-echo and fast spin-echo imaging concluded that fast spin-echo imaging is inferior to conventional spin-echo imaging for the diagnosis of meniscal tears [19], later studies have found no significant difference in accuracy when comparing fast and conventional spin-echo imaging [3, 4]. More recently, a sensitivity of 80% was reported for the MRI diagnosis of meniscal tears when using fast spin-echo imaging compared with 93% when using conventional spin-echo imaging [5]. Those authors strongly advocated that fast spin-echo imaging not be used for diagnosing meniscal tears [5]. They based this recommendation on their study and a review of the literature in which fast spin-echo imaging had an average sensitivity of approximately 80% for meniscal tears whereas conventional spin-echo imaging had a sensitivity of more than 90% [5]. However, Vande Berg et al. [11] reported a sensitivity of 96% for diagnosing meniscal tears using fast spin-echo imaging.

Fast spin-echo imaging has the potential for decreased accuracy in the diagnosis of meniscal tears because of the increased blurring seen in fast spin-echo sequences, especially when a short TE is used [5, 19]. However, other studies have noted that the short TE blurring can be reduced by using a shorter echo-train length and higher performance gradients, which reduce blurring by permitting shorter interecho spacing, and by using the second echo as the effective TE [20]. During the time of the MR scans analyzed in this study, we used an echo-train length of 4 for our sagittal proton density-weighted images but did not have the higher-speed gradients on our MR unit to allow shorter interecho spacing. Despite the lack of high-speed gradients, our sensitivity of 95% for medial meniscal tears was near the upper end of the range of reported sensitivities in other studies. Our sensitivity of 77% for lateral meniscal tears, although lower than we would like, is comparable to other studies.

In our opinion, the optimal MRI parameters for diagnosing a meniscal tear have not been defined. Probably there are important interactions between each of the multiple imaging parameters used in MRI. Another variable that has not been fully studied is the value of adding fat saturation to the proton density-weighted sagittal sequence. Sagittal fat-saturated proton density-weighted images were used in the two recent studies that reported a sensitivity of 93% for diagnosing meniscal tears using conventional spin-echo imaging and 96% when using fast spin echo [5, 11].

We reviewed the studies listed in Table 3 to determine the range of variation in imaging parameters in these studies. The variations in pixel size used in these studies are significant, with the field of view varying from 14 to 18 cm and the matrix size varying from 256 x 192 to 512 x 346. In those same studies, slice thickness varied from 3 to 5 mm and interslice gaps varied from none to 1.5 mm. Similarly, the types of surface coils and the number of excitations also varied considerably, which would probably give differing signal-to-noise ratios in the MR images. We believe that with all the variables that can be adjusted, it is possible that many different combinations of imaging parameters will give comparable accuracy for the diagnosis of meniscal tears.

Limitations of our study include the lack of direct comparison of conventional and fast spin-echo imaging. It would have been preferable if we had imaged all 561 patients with both conventional and fast spin-echo imaging and performed independent interpretations of each imaging set. Such a comparison would not have been possible in our busy clinical practice because the time to perform each MRI examination would be more than doubled.

A second limitation is a possible selection bias. We selected the 174 patients for this study from a consecutive series of 561 knee MRI examinations in order to use arthroscopy as the gold standard. However, the decisions of the surgeons to operate on a given patient depend on many variables, including the patients' symptoms and physical findings and the surgeons' review of our interpretations of the MR examinations. This limitation is unavoidable because it would be inappropriate for every patient undergoing MRI of the knee to have subsequent knee arthroscopy.

In addition, the size and type of the tear may also influence the surgeon's decision to operate on a given patient. We would expect that patients with large or displaced tears would be more likely to have knee arthroscopic surgery than patients with small tears. Thus, studies using operative findings as the gold standard for MRI accuracy may have more of the larger or displaced tears that are easier to diagnose on MRI.

In summary, we have shown that by prospectively diagnosing menisci as torn when there are two or more MR images with meniscal signal to the surface or distortion, we had a high positive predictive value for a tear to be found at arthroscopy. In contrast, menisci with only one abnormal MR image were considerably less likely to be found to be torn at arthroscopy. We also found an accuracy for diagnosing meniscal tears using fast spin-echo imaging comparable to that reported in previous studies using either conventional or fast spin-echo imaging.

References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Rubin DA, Paletta GA Jr. Current concepts and controversies in meniscal imaging. Magn Reson Imaging Clin N Am2000; 8:243 -270[Medline]
2. De Smet AA, Norris MA, Yandow DR, Quintana FA, Graf BK, Keene JS. MR diagnosis of meniscal tears of the knee: importance of high signal in the meniscus that extends to the surface. AJR1993; 161:101 -107[Abstract/Free Full Text]
3. Escobedo EM, Hunter JC, Zink-Brody GC, Wilson AJ, Harrison SD, Fisher DJ. Usefulness of turbo spin-echo MR imaging in the evaluation of meniscal tears: comparison with a conventional spin-echo sequence. AJR 1996; 167:1223 -1227[Abstract/Free Full Text]
4. Anderson MW, Raghavan N, Seidenwurm DJ, Greenspan A, Drake C. Evaluation of meniscal tears: fast spin-echo versus conventional spin-echo magnetic resonance imaging. Acad Radiol1995; 2:209 -214[CrossRef][Medline]
5. Blackmon GB, Major NM, Helms CA. Comparison of fast spin-echo versus conventional spin-echo MRI for evaluating meniscal tears. AJR 2005; 184:1740 -1743[Abstract/Free Full Text]
6. Blankenbaker DG, De Smet AA, Smith JD. Usefulness of two indirect MR imaging signs to diagnose lateral meniscal tears. AJR 2002; 178:579 -582[Abstract/Free Full Text]
7. Kaplan PB, Gehl RH, Dussault RG. Bone contusions of the posterior lip of the medial tibial plateau (contrecoup injury) and associated internal derangements of the knee at MR imaging. Radiology1999; 211:747 -753[Abstract/Free Full Text]
8. De Smet AA, Norris MA, Yandow DR, Graf BK, Keene JS. Diagnosis of meniscal tears of the knee with MR imaging: effect of observer variation and sample size on sensitivity and specificity. AJR1993; 160:555 -559[Abstract/Free Full Text]
9. Crues JV III, Mink J, Levy TL, Lotysch M, Stoller DW. Meniscal tears of the knee: accuracy of MR imaging. Radiology1987; 164:445 -448[Abstract/Free Full Text]
10. Mink JH, Levy T, Crues JV III. Tears of the anterior cruciate ligament and menisci of the knee: MR imaging evaluation. Radiology 1988;167 : 769-774[Abstract/Free Full Text]
11. Vande Berg BC, Malghem J, Poilvache P, Maldague B, Lecouvet FE. Meniscal tears with fragments displaced in notch and recesses of knee: MR imaging with arthroscopic comparison. Radiology2005; 234:842 -850[Abstract/Free Full Text]
12. Fischer SP, Fox JM, Del Pizzo W, Friedman MJ, Snyder SJ, Ferkel RD. Accuracy of diagnoses from magnetic resonance imaging of the knee: a multicenter analysis of one thousand and fourteen patients. J Bone Joint Surg Am 1991; 73:2 -10[Abstract/Free Full Text]
13. Quinn SF, Brown TF. Meniscal tears diagnosed with MR imaging versus arthroscopy: how reliable a standard is arthroscopy? Radiology 1991;181 : 843-847[Abstract/Free Full Text]
14. De Smet AA, Graf BK. Meniscal tears missed on MR imaging: relationship to meniscal tear patterns and anterior cruciate ligament tears. AJR 1994; 162:905 -911[Abstract/Free Full Text]
15. Justice WW, Quinn SF. Error patterns in the MR imaging evaluation of menisci of the knee. Radiology 1995;196 : 617-621[Abstract/Free Full Text]
16. Kojima KY, Demlow TA, Szumowski J, Quinn SF. Coronal fat suppression fast spin-echo images of the knee: evaluation of 202 patients with arthroscopic correlation. Magn Reson Imaging1996; 14:1017 -1022[CrossRef][Medline]
17. Rubin DA, Kettering JM, Towers JD, Britton CA. MR imaging of knees having isolated and combined ligament injuries. AJR1998; 170:1207 -1213[Abstract/Free Full Text]
18. Cheung LP, Li KCP, Hollett MD, Bergman AG, Herfkens RJ. Meniscal tears of the knee: accuracy of detection with fast spin-echo MR imaging and arthroscopic correlation in 293 patients. Radiology1997; 203:508 -512[Abstract/Free Full Text]
19. Rubin DA, Kneeland JB, Listerud J, Unerberg-Davis SJ, Dalinka MK. MR diagnosis of meniscal tears of the knee: value of fast spin-echo vs conventional spin-echo pulse sequences. AJR1994; 162:1131 -1135[Abstract/Free Full Text]
20. Kowalchuk RM, Kneeland JB, Dalinka MK, Siegelman ES, Dockery WD. MRI of the knee: value of short echo time fast spin-echo using high performance gradients versus conventional spin-echo imaging for the detection of meniscal tears. Skeletal Radiol 2000;29 : 520-524[CrossRef][Medline]

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				M. Englund, A. Guermazi, D. Gale, D. J. Hunter, P. Aliabadi, M. Clancy, and D. T. Felson Incidental Meniscal Findings on Knee MRI in Middle-Aged and Elderly Persons N. Engl. J. Med., September 11, 2008; 359(11): 1108 - 1115. [Abstract] [Full Text] [PDF]

				S. Y. Lee, W.-H. Jee, and J.-M. Kim Radial Tear of the Medial Meniscal Root: Reliability and Accuracy of MRI for Diagnosis Am. J. Roentgenol., July 1, 2008; 191(1): 81 - 85. [Abstract] [Full Text] [PDF]

				A. A. De Smet, D. H. Nathan, B. K. Graf, B. A. Haaland, and J. P. Fine Clinical and MRI Findings Associated with False-Positive Knee MR Diagnoses of Medial Meniscal Tears Am. J. Roentgenol., July 1, 2008; 191(1): 93 - 99. [Abstract] [Full Text] [PDF]

				A. A. De Smet and R. Mukherjee Clinical, MRI, and Arthroscopic Findings Associated with Failure to Diagnose a Lateral Meniscal Tear on Knee MRI Am. J. Roentgenol., January 1, 2008; 190(1): 22 - 26. [Abstract] [Full Text] [PDF]

				C.-H. Lee, C.-Y. Tzeng, and Y.-M. Lin Specificity Gains at the Expense of Sensitivity and Vice Versa Am. J. Roentgenol., June 1, 2007; 188(6): W574 - W574. [Full Text] [PDF]

				A. De Smet and M. Tuite Reply Am. J. Roentgenol., June 1, 2007; 188(6): W575 - W575. [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by De Smet, A. A. Articles by Tuite, M. J. Search for Related Content PubMed PubMed Citation Articles by De Smet, A. A. Articles by Tuite, M. J. Social Bookmarking                      What's this? Hotlight (NEW!) What's Hotlight?