Noncontrast magnetic resonance imaging of superior labral lesions : 102 Cases confirmed at arthroscopic surgery

Previous studies report that noncontrast magnetic resonance imaging is limited in the evaluation of the superior glenoid labrum. From our magnetic resonance imaging database of 2552 patients, we prospectively identified 104 patients with superior labral lesions who subsequently went on to arthroscopic surgery. Magnetic resonance images were assessed to identify fraying, flap tears, bucket-handle tears, or displaced flap of fibrocartilage. The biceps tendon was also evaluated. Patients were categorized according to Snyder's classification, and the findings on the magnetic resonance images were correlated with surgical findings. One hundred of the 104 tears suspected on the images were confirmed at surgery. There were four false-positives and two false-negatives, the former reflecting one normal labrum, two meniscoid-type labra, and one sublabral foramen. With arthroscopic surgery as the standard, magnetic resonance imaging had a sensitivity of 98.0% (100 of 102), a specificity of 89.5% (34 of 38), and an accuracy of 95.7% (134 of 140) for detection of superior labral lesions. We concluded that high-resolution noncontrast magnetic resonance imaging can accurately diagnose superior labral lesions and aid in surgical management.