Pelvic Positioning Creates Error in CT Acetabular Measurements

Abstract

CT allows for accurate measurement of acetabular orientation and shape, but malpositioning of the pelvis may lead to measurement variance. We therefore sought to determine: (1) whether acetabular anteversion measurements using the femoral head centers differed from those using the posterior ischia, and (2) the extent to which changing obliquity, rotation, and tilt of a pelvis in a CT scanner affected the measurement of acetabular variables. A radiopaque human pelvis model with articulated hips was suspended from a plastic sheet as part of an adjustable frame. Changes in the transverse and sagittal planes created rotation and tilt, while rotating the frame in the coronal plane created obliquity. CT scans were obtained, varying the combinations of obliquity, rotation, and tilt by intervals of 5°, up to 20°. Acetabular anteversion (AA), anterior acetabular sector angle (AASA), posterior acetabular sector angle (PASA), and horizontal acetabular sector angle (HASA) were measured. The two methods for measuring AA yielded values differing by 1° to 4° but correlated (r = 0.981) across the spectrum of pelvis positioning. Pelvic obliquity and tilt were linearly associated with changes in the measurements. For each 1°-increase in pelvic obliquity, AA changed -0.4°, and AASA, PASA, and HASA changed 1.93°, 0.99°, and 2.80°, respectively. For each 1°-increase in pelvic tilt, AA changed 0.8°, and AASA, PASA, and HASA changed -1.07°, 0.52°, and -0.51°, respectively. Rotation had no affect on the variables. Small changes in pelvic obliquity and tilt were associated with variances in acetabular measurements. The measured changes were directly proportional to the changes in obliquity and tilt, and were additive. Pelvic rotation created no changes in measurement.