AUTHOR=Brühschwein Andreas , Schmitz Bronson , Zöllner Martin , Reese Sven , Meyer-Lindenberg Andrea TITLE=Introduction of a bone-centered three-dimensional coordinate system enables computed tomographic canine femoral angle measurements independent of positioning JOURNAL=Frontiers in Veterinary Science VOLUME=Volume 9 - 2022 YEAR=2022 URL=https://www.frontiersin.org/journals/veterinary-science/articles/10.3389/fvets.2022.1019215 DOI=10.3389/fvets.2022.1019215 ISSN=2297-1769 ABSTRACT=Radiographic and most computed tomographic (CT) angular measurements are two-dimensional. Radiography is affected by projectional variation and geometric distortion. Three-dimensional CT-based volume rendered (VR) views allow free rotation and choice of perspective, but based on the final selected view, a two-dimensional planar image is created that lacks a third dimension. A source of variation for angular measurements remain according to the finally selected projection, plane or viewing perspective. The resulting variation and potential error are tried to be minimized using visually standardized positioning or VR-viewing perspective, which is difficult or even impossible in severe complex angular and torsional bone deformities. Therefore, goal of the study was the development and description of a CT-based technique to measure angular deformities in the canine femur that is independent from positioning and truly three-dimensional. A CT scan of canine hind limbs and scans of 13 normal canine femoral bones in parallel and in various double oblique positioning on the table and hind limbs of canine patients with presumed normal femurs were used for the development and evaluation of the technique to measure selected clinically relevant morphometric canine femoral angles with a focus on torsional and varus deformities. A three-dimensional femoral bone centered coordinate system was introduced into the CT-scans based on osseous reference points and axes and angular measurements of skew lines were described and enabled by geometric definition of projection planes. Detailed three-dimensional description of the cross-sectional anatomical landmarks and reference points was illustrated with multiplanar and three-dimensional CT images. Mean differences between angular measurement results of various parallel (0°) and varying off-z-axis (15° and 45° deviated) double oblique positioning of the femurs on the CT-table in the gantry during scanning were calculated to test for independency of femoral positioning and intra- and interobserver agreement was determined by calculation of the standard deviation and the coefficient of variation for repeated measurements (%) to test for reproducibility of femoral angular measurement results. The technique was feasible in 13 normal canine femurs, resulted in acceptable intra- and interobserver agreement and proved its independency of femoral positioning by introduction of a bone-centered coordinate system.