Comparison of conventional MRI and CT scans for assessing bony parameters and classifying On- and Off-Track lesions in anterior shoulder dislocations

Background

Shoulder dislocation is the most common major joint dislocation, with anterior dislocations accounting for 95% of cases. Accurate assessment of bony lesions, such as glenoid bone loss (GBL) and Hill-Sachs lesions, is critical for treatment planning. While computed tomography (CT) is the gold standard for evaluating bony parameters, magnetic resonance imaging (MRI) may serve as a viable alternative, offering no radiation exposure. This study aims to compare the reliability of conventional 2D (two- Dimensional)-MRI with 2D-CT in measuring bony parameters and classifying lesions as on- or off-track. It was hypothesized that there is no difference in evaluation between MRI and conventional CT scans.

Methods

A retrospective case-control study was conducted on 61 patients (mean age 45 ± 19 years) with anterior shoulder dislocations who underwent both CT and MRI imaging. Radiographic measurements, including glenoid diameter, glenoid defect (in width), Hill-Sachs lesion and bony bridge, were obtained independently from CT and MRI scans. Patients were categorized as on- or off-track based on the glenoid track and Hill-Sachs index. Statistical analyses included correlation tests, Bland-Altman plots, interrater agreement (intraclass correlation coefficient), and sensitivity and specificity analyses for lesion classification.

Results

MRI showed good agreement with CT across most parameters, with mean differences of less than 1 mm for glenoid defect, glenoid diameter, and Hill-Sachs lesions. Correlation coefficients ranged from 0.62 (bony bridge) to 0.93 (glenoid defect). Bland-Altman plots revealed good agreement for glenoid parameters but higher variance for the Hill-Sachs lesion and bony bridge. MRI correctly classified 89% of on-track lesions (sensitivity) and 76% of off-track lesions (specificity). Interrater agreement was excellent for glenoid defect measurements (ICC = 0.962) and lower for the bony bridge (ICC = 0.848).

Conclusion

Conventional MRI demonstrates high reliability in measuring bony parameters and good accuracy in classifying on- and off-track lesions compared to CT. MRI is a viable alternative for preoperative evaluation, particularly in cases with minor bony defects. However, in indeterminate defects, a CT scan is recommended to ensure accurate measurements, classification and treatment planning.

Level of evidence

Level III.