Abstract
PURPOSE: To evaluate and compare the diagnostic accuracy of magnetic resonance imaging (MRI) with intraoperative knee arthroscopic findings for identifying or diagnosing meniscal tears. METHODS: We conducted a retrospective study of patients who underwent MRI and knee arthroscopy showing either medial meniscus (MM) or lateral meniscus (LM) tears at a single university hospital. The preoperative MRI findings of patients were compared with intraoperative arthroscopic findings to determine the presence, location, and morphology of meniscal tears. The results of arthroscopy were considered the definitive diagnosis. RESULTS: A total of 543 patients were initially identified. Of these, 220 met the study eligibility criteria and were included. The sensitivity, specificity, and accuracy of MRI in relation to arthroscopy were 94.29% (95% confidence interval [CI], 89.05%-97.50%), 78.75% (95% CI, 68.17%-87.11%), and 88.64% (95% CI, 83.68%-92.51%), respectively, for MM tears and 76.74% (95% CI, 66.39%-85.18%), 94.03% (95% CI, 88.58%-97.39%), and 87.27% (95% CI, 82.13%-91.37%), respectively, for LM tears. Complex tears were the most common morphology of tears, and the posterior horn was the most common location. CONCLUSIONS: MRI is an effective first-line noninvasive diagnostic tool for investigating meniscal tears, with overall diagnostic accuracies of 88.64% for MM tears and 87.27% for LM tears. MM tears had the highest incidence, particularly in the posterior horn. MRI showed high sensitivity for MM tears, high specificity for LM tears, and substantial agreement with arthroscopy in diagnosing both MM and LM tears. However, MRI's specificity for MM tears and sensitivity for LM tears were lower, suggesting that it may not be as reliable in confirming MM tears or ruling out LM tears. LEVEL OF EVIDENCE: Level III, retrospective cohort study.