Hyun Ji Kim MD, Presenter: Nothing to Disclose

Sungjun Kim MD, Abstract Co-Author: Nothing to Disclose

Sung-Ah Lee MD, Abstract Co-Author: Nothing to Disclose

Jihae Lee MD, Abstract Co-Author: Nothing to Disclose

Kyung-Bin Joo MD, Abstract Co-Author: Nothing to Disclose

Tae Soo Park, Abstract Co-Author: Nothing to Disclose

Gwang Hyun Lee, Abstract Co-Author: Nothing to Disclose

Choongki Park MD, Abstract Co-Author: Nothing to Disclose

Dong Woo Park MD, Abstract Co-Author: Nothing to Disclose

Yong-Soo Kim MD, PhD, Abstract Co-Author: Nothing to Disclose

Jeong-Nam Heo MD, Abstract Co-Author: Nothing to Disclose

et al, Abstract Co-Author: Nothing to Disclose

To determine if standard MR imaging is suboptimal in the evaluation of rotator cuff tear comparing with MR arthrography or not, through retrospective review of surgically confirmed rotator cuffs.

From January 2003 to April 2006, 41 symptomatic patients (24 male, 17 female) who underwent both MR imaging and shoulder arthroscopy were included. The patient’s mean age was 48.9 years (range, 17-76years). The mean time interval from MR imaging to arthroscopy was 38.5days (range, 1-167 days). MR imagings were performed before and after intraarticular injection with 1.5-T unit. Four independent readers (2 experienced, 2 less-experienced) retrospectively reviewed the 2 sets of MR images (standard MR imaging; NA set, MR arthrography; A set). The reviewers gave scores from 1 to 5 for both sets in the order of the confidence. The 2 sets of MRI were compared for diagnostic performance with ROC analysis for all independent readers. For the evaluation of the influence of the experience on interpretation, ROC comparison was also done for each experienced versus less-experienced readers. These analyses were done under the reference standard of full thickness tear (FT) and partial thickness tear (PT) confirmed on arthroscopy.

Arthroscopy confirmed 23 FT, 7 PT, and 11 normal rotator cuffs. The areas under the ROC curves (AUC) between 2 sets were not significantly different for both FT and PT in all 4 readers (p>0.176). The AUC of one experienced reader was significantly higher than one less-experienced reader in the evaluation of FT on NA set (p=0.031). There was no other significant difference in AUC in terms of experience.

Although MR arthrography is a useful tool for the evaluation of more details of shoulder pathologies, this technique is inherently invasive. Standard MR imaging is not suboptimal modality for the evaluation of rotator cuff.

For the evaluation of rotator cuff, it is not mandatory to perform invasive MR arthrography.

Kim, H, Kim, S, Lee, S, Lee, J, Joo, K, Park, T, Lee, G, Park, C, Park, D, Kim, Y, Heo, J, et al, , Is Standard MR Imaging Suboptimal for the Evaluation of Rotator Cuff Tear? Comparison of Diagnostic Performance between MR Arthrography versus Standard MR Imaging in 41 Patients with Arthroscopic Confirmation.  Radiological Society of North America 2006 Scientific Assembly and Annual Meeting, November 26 - December 1, 2006 ,Chicago IL. http://archive.rsna.org/2006/4440553.html