Magnetic resonance imaging is comparable to computed tomography for determination of glenoid version but does not accurately distinguish between Walch B2 and C classifications

Background: Computed tomography (CT) scan is the standard for the preoperative assessment of glenoid version and morphology before total shoulder arthroplasty. However, the capacity of magnetic resonance imaging (MRI) to visualize bone morphology has improved with advancing technology. The purpose of this study was to compare the accuracy of MRI to CT for assessment of glenoid version and Walch classification. Methods: Three fellowship-trained shoulder surgeons assessed glenoid version and Walch classification of 30 patients with primary shoulder osteoarthritis who received both CT and MRI scans before total shoulder arthroplasty. Version measurements, Walch classification, and observer agreement were compared. Results: Mean glenoid version was -15.5° and -18.6° by CT and MRI, respectively (P = .17). Interobserver reliability coefficients were good for both imaging modalities (CT, 0.73; MRI, 0.62). Intraobserver coefficients were good to excellent for CT (range, 0.76-0.87) and good for MRI (range, 0.75-0.79). For Walch classification, interobserver reliability for both modalities was merely fair, whereas intraobserver reliability was moderate to good. Although identification of type A1, A2, and B1 was nearly identical between CT and MRI, there was observer disagreement on type B2 (P = .001) and C glenoids (P = .03). Specifically, MRI underidentified type B2 and overidentified type C compared with CT. Conclusions: MRI is largely comparable to CT scan for evaluation of the glenoid, with similar measurements of version and identification of less extreme Walch glenoids. However, MRI is less accurate at distinguishing between type B2 and C glenoids.