Shoulder MRI refinements: differentiation of rotator cuff tear from artifacts and tendonosis, and reassessment of normal findings.

One of the difficulties with rotator cuff imaging lies in the normal variability of the tendon's signal. There may be intermediate signal present within the tendon because of magic-angle phenomenon, muscle and tendon fiber interdigitation, or tendinopathy related to degenerative changes or overuse injury. Partial and complete rotator cuff tears should be distinguishable from these causes of intermediate signal if water signal is reliably identified. This article reviews the important issue of distinguishing between rotator cuff tear and other causes of high signal in the rotator cuff, including artifacts and tendonosis. We include a review of the literature and a brief report of a study we conducted on 20 shoulders of 14 asymptomatic, young volunteers. In this study, the rotator cuff tendons were evaluated for abnormal signal at different TE values to determine at what TE the interpreters were able to confidently distinguish the high-signal intensity of a tear (water) from the intermediate signal intensity associated with artifact and tendinopathy. Readers were able to distinguish water and tendon signal in 70% to 100% of fast-spin echo (FSE) fat-saturated images with TE of 66, but there was interobserver variability at this TE, suggesting that it is less reliable than 88 ms in the identification of rotator cuff tears. By using FSE fat-saturated sequences with TE of 88 and fast spin echo inversion recovery (FSEIR) sequences, readers at all levels of experience were able to differentiate water signal intensity from tendon signal intensity in 100% of cases. Therefore, we suggest that either FSEIR images or FSE fat-saturated images with TE greater than 66 be used to facilitate the differentiation of fluid signal from intermediate increased signal intensity in rotator cuff imaging. Additionally, this article reviews the normal findings of shoulder magnetic resonance imaging (MRI) as revealed by the asymptomatic subjects included in our study, and assesses these findings in respect to previous publications. The normal features reviewed include the subacromion-subdeltoid (SA/SD) bursa, the biceps tendon sheath, the acromioclavicular (AC) joint, and the greater tuberosity of the humerus. A small amount of fluid was commonly seen in the SA/SD bursa, as well as the biceps tendon sheath. Subjective down-sloping of the acromion in the coronal plane, mild degenerative change of the AC joint, and undersurface spurring of the AC joint were uncommon in our normal subjects. Cystic change limited to the posterior aspect of the greater tuberosity was identified in 15% to 45% of shoulders.