Comparison of supraspinatus tendon and glenohumeral joint axes in MR imaging of the shoulder

Objective. In magnetic resonance (MR) imaging of the shoulder, oblique coronal images are used for evaluating the supraspinatus tendon (SST) of patients with suspected rotator cuff tear or impingement. This study aimed to compare orientation of the SST long axis with planes perpendicular to the glenohumeral joint (GHJ). Design and patients. The axial scans of 100 consecutive patients referred for MR imaging or MR arthrography of the shoulder were reviewed. Using the electronic cursors of a computer workstation, the angle of the SST long axis was measured and compared with the angle obtained through the GHJ utilizing three different landmarks: perpendicular to the joint (GHJ-90), joint–humeral head center axis (GHJ-H) and joint–scapular body axis (GHJ-S). Results. Differences in angulation between axes of the SST and the three GHJ axes averaged only about 5° [range of means 4.5–5.3°, range of standard deviation (SD) 3.8–4.6°]. In the majority of shoulders, angular differences measured 4 or less for all SST/GHJ comparisons. Similarly, small angular differences in the three GHJ axes were found: 4.5° (SD 3.3°) for GHJ-90/GHJ-S, 5.0° (SD 4.0°) for GHJ-S/GHJ-H and 2.9° (SD 3.0°) for GHJ-90/GHJ-H. Correlation between the GHJ-90 and GHJ-H axes was particularly good, with differences of 4° or less in 84% of shoulders. The orientations of the GHJ axes and that of the SST long axis are comparable. Conclusion. The GHJ may potentially be used as a landmark for obtaining oblique coronal images of the SST. Received: 19 November 1999 Revision requested: 18 January 2000 Revision received: 27 March 2000 Accepted: 5 April 2000