Rotator cuff: evaluation with fast spin-echo versus conventional spin-echo MR imaging.

PURPOSE

To determine if fast spin-echo (SE) magnetic resonance (MR) imaging can provide similar information to that of conventional SE imaging for evaluation of the rotator cuff.

MATERIALS AND METHODS

One hundred twenty-six patients underwent MR imaging with conventional SE and non-fat-suppressed fast SE sequences (65 patients) or conventional SE and fat-suppressed fast SE sequences (61 patients). Two radiologists independently graded the rotator cuff with separate and side-by-side assessment of the fast SE and conventional SE images.

RESULTS

For detection of full-thickness tears, agreement between non-fat-suppressed fast SE and conventional SE images was 93.8% (kappa = 0.78 [good]) and 95.4% (kappa = 0.82 [very good]) for the two readers, respectively, and agreement between fat-suppressed fast SE and conventional SE images was 98.4% (kappa = 0.96 [very good]) and 91.8% (kappa = 0.73 [good]) for the two readers, respectively. Rotator cuff grading was similar for fast SE and conventional SE: weighted kappa = 0.77 (good) and 0.68 (good) for non-fat-suppressed and weighted kappa = 0.83 (very good) and 0.67 (good) for fat-suppressed fast SE images for the two readers, respectively.

CONCLUSION

Fast SE sequences yield similar interpretations as those obtained with a conventional SE sequence for evaluation of the rotator cuff.