MR imaging of the shoulder: correlation with plain radiography.

Recent advances in surface-coil technology and scanner software have made high-quality MR examinations of the shoulder a practical reality. The high signal-to-noise ratios provided by local coils designed specifically for the shoulder allow imaging with small fields of view and thin sections. New software provides routine oblique imaging with oft-center fields of view and no wraparound artifacts. With these stateof-the-art techniques, several studies have shown that MR is as accurate as, if not more accurate than, arthrography in detecting rotator cuff tears [1 , 2]. In addition, MR can provide an accurate preoperative assessment of the size of a cuff tear and the condition of the remaining tendon. In the absence of a cuff tear, MR can be useful in evaluating impingement, tendinitis, and bursitis; these cannot be assessed with arthrography. MR also can show other abnormalities such as avascular necrosis, labral tears, tumors, osteomyelitis, and arthritis [3]. Although MR is considerably more expensive than arthrography, it is probable that the additional information obtained in a noninvasive fashion will make MR the procedure of choice for the evaluation of shoulder pain. For these reasons, MR is frequently the next study performed after conventional radiography. Comparison with the radiographs is essential for a complete evaluation of the MR images. Each of these techniques has distinct advantages and limitations. In this essay we will illustrate examples of these features for a variety of clinical conditions. Materials and Methods