Purpose: To compare the diagnostic accuracy of three-dimensional (3D) isotropic magnetic resonance (MR) arthrography with two-dimensional (2D) MR arthrography for the diagnosis of labral lesions of the shoulder performed by using a 3.0-T imager.
Materials and methods: Institutional review board approval was given for this retrospective study and informed consent was waived. From March 2006 to January 2007, 100 patients underwent 3D isotropic and 2D MR shoulder arthrographic imaging by using a 3.0-T imager and subsequent arthroscopic surgery. The 3D isotropic sequences were performed by using fast gradient-echo imaging with fat suppression (voxel size, 0.6 x 0.6 x 0.6 mm; imaging time, 5 minutes 32 seconds) and were evaluated for the presence of superior, anterior, and posterior labral lesions by using 3D isotropic and 2D MR arthrography. The statistical differences between the sensitivity and specificity for both methods were analyzed by using the McNemar test, with arthroscopic findings regarded as reference standard.
Results: Surgical findings confirmed 53 superior labral anterior posterior (SLAP) lesions, 17 anterior labral lesions, and five posterior labral lesions. Respective sensitivity and specificity were 85% and 96% for SLAP lesions, 100% and 98% for anterior labral lesions, and 80% and 99% for posterior labral lesions by using 2D MR arthrography and 83% and 96% for SLAP lesions, 100% and 96% for anterior labral lesions, and 80% and 99% for posterior labral lesions by using 3D isotropic MR arthrography. There was no significant difference in sensitivities and specificities of both methods; however, the power of this study was limited and larger comparison is needed.
Conclusion: Isotropic 3D shoulder MR arthrography combined with a multiplanar reconstruction technique can help in the diagnosis of shoulder labral lesions as does 2D MR arthrography, but with shorter imaging times.