Thomas Henry Magee MD, Presenter: Nothing to Disclose

R. Richard Ramnath MD, Abstract Co-Author: Nothing to Disclose

Marc D. Shapiro, Abstract Co-Author: Nothing to Disclose

David S. Williams MD, Abstract Co-Author: Nothing to Disclose

The sensitivity and specificity of 3.0 Tesla MR imaging of the knee for meniscal tears has not been specifically assessed. We underwent a retrospective review of 100 knee MR exams to examine how sensitive and specific knee imaging for meniscal tears is at 3.0 Tesla.

Two experienced musculoskeletal radiologists retrospectively reviewed MR scans of the knee in one hundred consecutive patients. Interpretations were performed by consensus reading. All patients had complete MR scans of the knee on a 3.0 Tesla General Electric unit. All patients had fast spin echo proton density sagittal imaging performed at 2 mm slice thickness. Fat saturated fast spin echo T2 axial, coronal and sagittal imaging and T1 coronal imaging was also performed. Sixty-three of the 100 patients underwent arthroscopy and results were compared with MR interpretations.

There were seventy-two meniscal tears in sixty-nine patients demonstrated by consensus retrospective reading of the 100 knee MR exams. Of these 100 patients, sixty-three underwent arthroscopy. There were two meniscal tears seen on arthroscopy not seen on MR exam. There were no false positive MR readings of meniscal tears as compared to arthroscopy. In this study MR sensitivity for detection of meniscal tears in the knee was 97% and specificity was 100%

3.0 Tesla MR imaging of the knee with use of fast spin echo proton density imaging at 2 mm slice thickness is very sensitive and specific for detection of meniscal tears as compared with arthroscopy.

Magee, T, Ramnath, R, Shapiro, M, Williams, D, Sensitivity and Specificity of 3.0 Tesla MR Imaging in Detection of Meniscal Tears.  Radiological Society of North America 2004 Scientific Assembly and Annual Meeting, November 28 - December 3, 2004 ,Chicago IL. http://archive.rsna.org/2004/4416180.html