[Reliability of the Evaluation of MRI Examinations after the Treatment of Chondral Defects in the Knee Joint].

PURPOSE OF THE STUDY This study is a component part of the project focused on cartilage imaging after the treatment of a defect. It aims to compare the evaluation of postoperative status performed by two radiologists with the use of 2D MOCART scoring system and to determine whether this method is a reliable tool for the evaluation of postoperative changes. MATERIAL AND METHODS The study evaluated 78 MRI examinations from 25 patients (one patient had two defects treated), each of whom underwent 3 MRI examinations at 6, 12 and 18 months after surgery. The MRI examinations were performed on Philips Ingenia 3T scanner with 8-channel knee coil, in line with the routine protocol (coronal, sagittal and transversal PD SPAIR, coronal T1, sagittal PD HR, sagittal bFFE). The MRI examinations were evaluated independently by two radiologists using the 2D MOCART (Magnetic Resonance Observation of Cartilage Repair Tissue) score. RESULTS The raters agreed in a total of 592 of 702 evaluations, therefore the inter-rater reliability is high, namely 84.3%. The highest inter-rater agreement was in assessing subchondral lamina and subchondral bone. Whereas the lowest inter-rater agreement was achieved in assessing effusion. The total score showed a very strong and statistically significant correlation (r = 0.893). In eight out of nine questions there was no statistically significant difference between the raters. A significant difference was seen only in the assessment of repair tissue structure. Excellent reliability of the total score was also confirmed by the intraclass correlation coefficient. DISCUSSION The high degree of agreement in assessing the signal intensity of repair tissue was considered very positive as it is generally viewed as the major pitfall in evaluations. On the contrary, subjective perception was confirmed in the evaluation of tissue homogeneity, especially when comparing homogeneity with the adjacent tissue in close vicinity that could have changed already. Surprisingly, the lowest inter-rater concordance was reported in the evaluation of effusion, where in some cases, its volume was underestimated, when traced back retrospectively. CONCLUSIONS The results of this study confirm that despite certain doubts regarding subjective perception of some of the evaluation criteria the 2D MOCART scoring system is a very good and objective tool to evaluate the effects of surgery. Key words: magnetic resonance imaging , hyaline cartilage, classification.