Evaluation of diagnostic accuracy of physical examination and MRI for ligament and meniscus injuries

Abstract Surgical planning shall be based on thorough physical examination and radiological findings during patient evaluation. The use of computational methods in the evaluation of diagnostic tests may facilitate the decision-making process as well as reduce treatment costs. Conventional X-ray is a modality of choice for bone visualisation however, it is not suitable for soft tissue evaluation. MRI is a tool designated for soft tissue examination however, its accuracy is debated in the literature. This study was performed to evaluate the diagnostic accuracy of MRI in most common knee injuries in comparison to physical examination. The evaluation of selected diagnostic methods used in the evaluation of ligaments and meniscus was carried out based on statistical indicators and ROC (Receiver Operating Characteristic) curves. The indicators were determined using Statistica and Matlab software. 69 patients were enrolled in the study, who were previously scheduled for arthroscopic knee surgery. Each patient underwent a closed protocol physical evaluation prior to surgery. MRI reports were collected and stored. Diagnostic accuracy was evaluated with the use of ROC curves. Physical examination presented greater diagnostic accuracy than MRI in ACL tears. The same dependency was found in lateral and medial meniscal tears. MRI showed AUC of 0.835 for MM, 0.788 for LM, and 0.870 for ACL, respectively. The physical examination had the greatest diagnostic accuracy in comparison with ACL tears, where AUC for the Lachman test reached 0,902. Apley test was most accurate for LM tears with AUC reaching 0,842, while the McMurray test was the most accurate for MM lesions. MRI is a preferable addition to a thorough physical examination.