Seven fresh frozen shoulder cadavers were imaged with radiographs, magnetic resonance imaging (MRI), computed tomography (CT), and 3-dimensional CT (3-D CT). Native shoulders were imaged, and 3 sequential anterior-inferior glenoid defects were created, measured, and reimaged. Defect sizes were <12.5 % , 12.5 % to 27 % , and >27 % . Four blinded evaluators (2 musculoskeletal radiologists, 2 shoulder fellowship-trained surgeons) reviewed the 112 image sets and estimated the percentage of glenoid bone loss. Images were scrambled and rereviewed by the same observers 2 months later to determine intraobserver reliability.
Pearson correlation coefficients between predicted vs true bone loss across all 4 raters were 0.875 (3-D CT), 0.831 (CT), 0.693 (MRI), and 0.457 (x-ray imaging). Prediction errors (PE) were (mean ¡À SD in percentages) 3-D CT (?3.3 ¡À ?6.6), CT (?3.7 ¡À ?8.0), MRI (?2.75 ¡À ?10.6), and x-ray images (?6.9?¡À ?13.1). Mean PE values were not significantly different among 3-D CT, CT, and MRI; however, the PE SDs were similar among the 4 evaluators for 3-D CT and lower than all other imaging techniques. Prediction based on x-ray images had the largest PE and SD. Covariance parameters revealed large variances for shoulders for MRI and x-ray imaging. The intraobserver intraclass correlation coefficients were 0.947 (3-D CT), 0.927 (CT), 0.837 (MRI), and 0.726 (x-ray image).
The most accurate imaging modality in predicting glenoid bone loss among the 4 blinded independent evaluators was 3-D CT.
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址:北京市海淀区学院路29号 邮编:100083 电话:办公室:(+86 10)66554848;文献借阅、咨询服务、科技查新:66554700 |