RSNA 2010 

Abstract Archives of the RSNA, 2010


LL-MKS-TU1A

SLAP Tears at 3.0 T Shoulder MR Arthrography: Diagnosis with 3D Isotropic Turbo Spin-Echo SPACE Sequence versus 2D Conventional Sequences

Scientific Informal (Poster) Presentations

Presented on November 30, 2010
Presented as part of LL-MKS-TU: Musculoskeletal

Participants

Joon-Yong Jung MD, Presenter: Nothing to Disclose
Won-Hee Jee MD, Abstract Co-Author: Nothing to Disclose
Michael Y. Park MD, Abstract Co-Author: Nothing to Disclose
So-Yeon Lee, Abstract Co-Author: Nothing to Disclose
Yang-Soo Kim MD, Abstract Co-Author: Nothing to Disclose

PURPOSE

To prospectively determine the reliability and accuracy of shoulder MR arthrography with 3D isotropic fat-suppressed (FS) turbo spin-echo sequence (TSE-SPACE) for diagnosing SLAP lesions compared with 2D conventional sequences at 3.0 T.

METHOD AND MATERIALS

Eighty-seven patients who underwent shoulder MR arthography using 3.0-T imagers and subsequent arthroscopy were included. In addition to 2D conventional sequences, 3D FS TSE-SPACE was obtained to produce multiplanar intermediate-weighted images with 0.6-mm isotropic resolution. Each set of MR imaging from two different methods was independently scored for the presence of SLAP lesion by two reviewers. Type I SLAP lesions were regarded as negative as they most often are not clinically relevant. Diagnostic performances between readers were compared with ROC curves. SLAP type in each case was correlated with arthroscopic reports and accuracies in predicting SLAP type between two methods were compared.

RESULTS

At arthroscopy, there were 30 SLAP tears: type II (n = 19), type III (n = 2), and type IV (n = 2), V (n = 7). Sensitivity, specificity and accuracy on 2D conventional sequences of each reader were 90%, 84% and 86%, and 89%, 84% and 86%, respectively, whereas on TSE-SPACE, sensitivity, specificity and accuracy of each reader were 79%, 86% and 83%, and 83%, 86% and 85%, respectively. There were no statistical differences in sensitivity and specificity between two methods (McNemar). Interobserver agreement was substantial (κ=0.76 with 2D conventional sequences; κ =0.68 with TSE-SPACE). The areas under the ROC curves for each reader were 0.93 and 0.91 for 2D conventional sequences and 0.89 and 0.90 for TSE-SPACE, which were not significantly different. SLAP types were correctly categorized in 79% of tears with 2D conventional sequences, whereas 66% of tears was correctly cocategorized with FS intermediate-weighted TSE-SPACE sequence.

CONCLUSION

Although 3.0 T shoulder MR arthrography using 3D FS intermediate-weighted TSE-SPACE sequence displayed a competitive accuracy for diagnosing SLAP tears compared with 2D conventional sequences, the accuracy in determining the SLAP types was decreased with TSE-SPACE sequence.

CLINICAL RELEVANCE/APPLICATION

Although 3.0 T shoulder MR arthrography using 3D isotropic FS TSE-SPACE displayed a competitive accuracy for diagnosing SLAP tears, the accuracy in determining the SLAP types was decreased.

Cite This Abstract

Jung, J, Jee, W, Park, M, Lee, S, Kim, Y, SLAP Tears at 3.0 T Shoulder MR Arthrography: Diagnosis with 3D Isotropic Turbo Spin-Echo SPACE Sequence versus 2D Conventional Sequences.  Radiological Society of North America 2010 Scientific Assembly and Annual Meeting, November 28 - December 3, 2010 ,Chicago IL. http://archive.rsna.org/2010/9014525.html