Accuracy of MRI in detecting the ACL lesions remain relatively high and unchanged with age, however when it comes to detecting the type and location of meniscal lesions, accuracy of MRI was found to be lower in subjects older than 40 years. Regarding the chondral lesions, while lower rates were found in older age group, accuracy rates were found to be lower than for other pathologies.
Investigation of age as a factor in accuracy of magnetic resonance imaging (MRI) in various knee joint pathologies.
This prospective cohort study was approved by local ethical committee. The subjects comprise of patients for whom arthroscopic intervention was planned and had undergone standardized MRI between May 2015 and May 2016. The subjects were grouped according to their ages; below 40 (Group 1, n=30) and 40 and above (Group 2, n=30). Sensitivity, specificity, and accuracy rates of MRI in anterior cruciate ligament (ACL), meniscal, and chondral lesions, as well as the type and location of meniscal lesions as verified by arthroscopy, were compared among groups.
The sections were evaluated by a radiologist blinded to the arthroscopy results. The statistical analyses were performed on SPSS 22.0 and Medcalc 14. Fisher exact test, Cohen’s kappa statistics, Monte Carlo method, Fisher’s Z transformation and student t-test were utilized. Significance level was set at p<0.05.
The sensitivity of MRI in detecting ACL lesions was 100% for both groups. Specificity and accuracy rates were 100% in group 1, 96.2% and 96.7% in group 2, respectively. A high degree of agreement was observed between MRI and arthroscopy in both groups (Kappa coefficients: 1.000 in group 1 (p<0.001), 0.870 in group 2 (p<0.001)).
In detection of medial meniscal lesions, MRI exerted sensitivity, specificity, and accuracy rates respectively as 70%, 95%, and 86.7% in group 1, and 81.8%, 87.5%, and 83.3% in group 2 with high degree of agreement in both groups (Kappa coefficients: 0.684 and 0.619 (p<0.001)). Accuracy for the type of medial meniscal tears was 86.7% for group 1, 80% for group 2; for the location of the tear accuracy was 86.7% and 76.7%.
Sensitivity rates were lower for both groups regarding the detection of lateral meniscal lesions (64.3% for group 1, 12.5% for group 2). Specificity rates were 100%, yielding accuracy rates of 83.3% for group 1, and 76.7% for group 2. Degrees of agreement were high for group 1, low for group 2 (Kappa coefficients: 0.658 (p<0.001), 0.173 (p=0.092)). For determining the type of lateral meniscal tears, the accuracy was 80% for both groups; while in determining the location of the tear accuracy rates were 80% and 76.7%.
Chondral lesions were detected with a sensitivity of 57.1% in group 1 and 77.3% in group 2. While specificity was 78.3% and 50%, accuracy rates were 73.3% and 70%. A low degree of agreement was found between MRI and arthroscopy for both groups.
Accuracy of MRI in detecting the ACL lesions remain relatively high and unchanged with age, however when it comes to detecting the type and location of meniscal lesions, accuracy of MRI was found to be lower in subjects older than 40 years. Regarding the chondral lesions, while lower rates were found in older age group, accuracy rates were found to be lower than for other pathologies. In conclusion, age is found to be a determining factor in the accuracy of MRI in detection of pathologies regarding knee joint.