Automated computation of femoral angles in dogs from three-dimensional computed tomography reconstructions: Comparison with manual techniques

The aim of this ex vivo study was to test a novel three-dimensional (3D) automated computer-aided design (CAD) method (aCAD) for the computation of femoral angles in dogs from 3D reconstructions of computed tomography (CT) images. The repeatability and reproducibility of three manual radiography, manual CT reconstructions and the aCAD method for the measurement of three femoral angles were evaluated: (1) anatomical lateral distal femoral angle (aLDFA); (2) femoral neck angle (FNA); and (3) femoral torsion angle (FTA). Femoral angles of 22 femurs obtained from 16 cadavers were measured by three blinded observers. Measurements were repeated three times by each observer for each diagnostic technique. Femoral angle measurements were analysed using a mixed effects linear model for repeated measures to determine the levels of intra-observer agreement (repeatability) and inter-observer agreement (reproducibility). Repeatability and reproducibility of measurements using the aCAD method were excellent (intra-class coefficients, ICCs ≥ 0.98) for all three angles assessed. Manual radiography and CT exhibited excellent agreement for the aLDFA measurement (ICCs ≥ 0.90). However, FNA repeatability and reproducibility were poor (ICCs < 0.8), whereas FTA measurement showed slightly higher ICCs values, except for the radiographic reproducibility, which was poor (ICCs < 0.8). The computation of the 3D aCAD method provided the highest repeatability and reproducibility among the tested methodologies.