Reliability of coronal curvature measurements on 3D ultrasound images for AIS

Methods Forty participants (35F, 5M; mean age 14.0±2.0 years) who (1) were diagnosed with AIS, (2) had no prior spine surgical treatment, (3) had radiographs on the study day that were not in-brace, and (4) had the major Cobb angle less than 45°, were recruited consecutively from the local scoliosis clinics. All US images and radiographs were acquired in standing positions within an hour of each other. Three raters, who were blinded to the clinical information, measured the Cobb angles from the US and x-ray images at least 3 days apart. The raters repeated these measurements at least 1 week apart to minimize memory bias. The mean absolute deviation (MAD) and the standard deviation (SD) between the two measurements methods were used to estimate the reliability of the US measurement. The intraand interrater reliabilities were assessed by calculating the interclass correlation coefficients and standard errors of measurement (SEM).