Purpose: To investigate the agreement among magnetic resonance (MR) imaging, computed tomography (CT), and arthroscopy in the measurement of glenoid bone loss.
Materials and methods: This study was approved by the institutional ethics committee. One hundred seventy-six patients (158 male and 18 female patients; mean age, 26.8 years ± 12.3) with anterior shoulder dislocation underwent both shoulder MR imaging and CT examination. Anterior straight line length, glenoid width, and best-fit bone loss were measured with MR imaging and CT. Sixty-five patients also underwent arthroscopy, which was used as the standard of reference. Assessment of glenoid bone loss at MR imaging was compared with that at CT and arthroscopy. Inter- and intrareader reproducibility of MR imaging-derived measurements of glenoid bone loss was evaluated.
Results: There was excellent correlation between CT and MR imaging with regard to anterior straight line length (r = 0.97, P < .0001), glenoid width (r = 0.95, P < .0001), and severity of glenoid bone loss-particularly with use of best-fit circle width (r = 0.83, P < .0001) rather than best-fit circle area (r = 0.82, P < .0001). In the assessment of glenoid bone loss, the correlation between CT and arthroscopy (r = 0.91, P < .0001) was marginally better than that between MR imaging and arthroscopy (r = 0.84, P < .0001). The inter- and intrareader correlations of MR imaging-derived measurements of glenoid bone loss were excellent (R = 0.90-0.95).
Conclusion: MR imaging assessment of glenoid bone loss, particularly with use of glenoid width, is almost as accurate as CT assessment.
© RSNA, 2013.