RESUMO
PURPOSE: To categorise discrepancies in findings of the menisci and anterior cruciate ligament (ACL) between arthroscopy and MRI. MATERIALS AND METHODS: The MRIs of 236 patients were retrospectively analysed by an experienced radiologist without knowledge of clinical and/or operative findings. Discrepancies in arthroscopic findings were reevaluated together with the arthroscopist to determine their cause of error. RESULTS: The diagnostic accuracies for injuries of the medial and lateral meniscus and the ACL were 92.4%, 92.4%, and 94.1%, respectively. For the menisci, causes for discrepancies in findings (n = 31) were: overinterpretation of central signal intensities with contact to the meniscal surface but without disturbance of the meniscal contour as a tear (n = 12), insufficient arthroscopie evaluation of the knee joint (n = 11), overlooked tears on MR imaging (n = 6), misinterpretation of normal anatomic structures (n = 1), "magic angle" phenomenon (n = 1), and missed tears at MRI (n = 1). Causes for discrepancies for the ACL (n = 18) were: nearly complete versus complete rupture either at MRI or arthroscopy and vice versa (n = 9), insufficient arthroscopic evaluation (n = 6), insufficient MRI technique (n = 2), and overlooked tear on MR imaging (n = 1). CONCLUSIONS: Discrepant findings between MRI and arthroscopy may be also due to an insufficient arthroscopic evaluation in clinical routine. The close cooperation between surgeons and radiologists improves the understanding of the methods of each other.