Three-dimensional Computed Tomography for Determination of Femoral Anteversion in a Cerebral Palsy Model

BACKGROUND:Previous investigation has proven 3-dimensional (3D) computed tomography (CT) to be a poor method of assessing femoral anteversion in patients with cerebral palsy. However, new advancements in CT software yield the potential to improve upon those dated results. METHODS:CT was performed on 9 femoral models with varying amounts of anteversion (20 to 60 degrees) and varying neck-shaft angles (120 to 160 degrees). Each model was scanned in 2 holding devices. One holder placed the femur in an ideal position relative to the gantry. The other placed the femur in flexion, adduction, and internal rotation simulating a common lower extremity posture in cerebral palsy. Femoral anteversion was measured on 3D reconstructions by 4 observers on 2 separate occasions. Interobserver and intraobserver reliability, accuracy, and the effect of increasing neck-shaft angle of the measurements were examined and compared with previously published data using the same models. RESULTS:Pearson correlation coefficients between first and second measurements by the same examiner were all above 0.96 regardless of positioning of the femur in the gantry. The correlation coefficients among all examiners were 0.97 regardless of positioning of the femur in the gantry. Accuracy in measurements was comparable using 3D CT techniques with mean differences between the normal and cerebral palsy-positioned models of