Two-dimensional magnetic resonance imaging sequences correlate to three-dimensional computed tomography for evaluation of glenoid bone loss

Background: To correlate two-dimensional (2D) fast-imaging employing steady-state acquisition (FIESTA) magnetic resonance imaging protocols with three-dimensional (3D) computed tomography (CT) for calculation of glenoid bone loss (GBL). FIESTA is a high-resolution protocol with high contrast, high signal-to-noise ratio, and sharp edge definition. Methods: We included patients who underwent a shoulder stabilization procedure from 2010 to 2023 with 2D FIESTA and 2D proton density fat-saturation (PD FS) magnetic resonance arthrogram sequences as well as 3D CT of the shoulder. We excluded patients with history of prior shoulder surgery or posterior instability, if the CT and magnetic resonance imaging were >90 days apart or if there was a documented instability event between imaging studies. Two raters calculated the primary outcome of GBL by the perfect-circle linear method twice for each modality separated by 3 weeks. Intrarater and inter-rater reliabilities were calculated. We analyzed linear regression models and concordance correlation coefficients between modalities. Results: Twelve patients were analyzed. Overall mean GBL by imaging modality was 12.1 ± 11.2% on PD FS, 11.8 ± 10.0% on FIESTA, and 11.5 ± 9.7% on 3D CT. GBL had excellent intrarater and inter-rater reliabilities (>0.9) on all modalities. For GBL, concordance correlation coefficients were 0.982 (95% CI 0.960-0.992) for PD FS vs. CT and 0.998 (95% CI 0.993-0.999) for FIESTA vs. CT. Linear regression models improved both PD FS and FIESTA accuracy to within 0.6% of CT-calculated GBL 95% of the time. Conclusion: 2D FIESTA magnetic resonance arthrogram is a suitable alternative to 3D CT reconstruction for calculation of GBL and slightly outperformed PD FS sequences.