Detection of the glenoid bare spot by non-arthrographic MR imaging, conventional MR arthrography, and 3D high-resolution T1-weighted VIBE MR arthrography: comparison with CT arthrography

Abstract

ObjectivesTo determine the diagnostic accuracy of non-arthrographic MR imaging, conventional MR arthrography, and 3D T1-weighted volumetric interpolated breath-hold examination (VIBE) MR arthrography sequences as compared with a CT arthrography in the diagnosis of glenoid bare spot.MethodsA retrospective study of 216 patients who underwent non-arthrographic MR imaging, conventional MR arthrography, VIBE MRI arthrography, and CT arthrogram between January 2011 and March 2022 was conducted. The diagnostic accuracy of non-arthrographic MR imaging, direct MR arthrography, and VIBE MRI arthrography in the detection of glenoid bare spot was compared with that of CT arthrography. All studies were reviewed by 2 MSK radiologists. Interobserver agreement for MR imaging and MR arthrographic findings was calculated.ResultsSixteen of 216 patients were excluded. Twenty-three of 200 shoulders had glenoid bare spot on CT arthrographic images. The glenoid bare spot was detected in 11 (47.8%) and 7 (30.4%) patients on conventional non-arthrographic MR images and in 18 (78.3%) and 16 (69.6%) patients on conventional MR arthrograms by observers 1 and 2, respectively. Both observers separately described the bare spot in 22 of 23 patients (95.7%) on 3D volumetric MR arthrograms. Interobserver variabilities were fair agreement for conventional non-arthrographic MR imaging (kappa = 0.35, p < 0.05), moderate agreement for conventional MR arthrogram (kappa = 0.50, p < 0.05), and near-perfect agreement for 3D volumetric MR arthrogram reading (kappa = 0.87, p < 0.05).ConclusionsA 3D high-resolution T1-weighted VIBE MR arthrography sequence may yield diagnostic performance that is comparable with that of CT arthrography in the diagnosis of glenoid bare spot.