Justice W W, Quinn S F
Department of Radiology, Oregon Health Sciences University, Portland, USA.
Radiology. 1995 Sep;196(3):617-21. doi: 10.1148/radiology.196.3.7644620.
To categorize errors in interpretation of magnetic resonance (MR) images of the knee and explain the discrepancy between MR imaging and diagnostic arthroscopic findings.
Five hundred sixty-one patients underwent arthroscopy and MR imaging of the knee. Prospective and retrospective readings were used to categorize tear type and location. Sixty-six patients had 68 discrepancies between MR imaging and arthroscopic findings.
Prospective interpretation of MR images of the lateral meniscus had a sensitivity of 82%, specificity of 98%, and accuracy of 93%. For images of the medial meniscus, sensitivity was 96%; specificity, 91%; and accuracy, 95%. The number of diagnostic errors at retrospective analysis was 76% (n = 52) of the number at prospective analysis.
MR imaging is accurate in evaluation for meniscal tears. Most errors in interpretation that occur at prospective evaluation also occur at retrospective evaluation. Some false-positive errors may be related to incomplete arthroscopic evaluation of the meniscus and confusion between what represents fraying and what represents a tear.