RESUMEN
OBJECTIVE: The aim of this study was to compare in vitro accuracy of femoral measurements obtained from different radiographical positions with actual femoral anatomical dimensions in dogs. MATERIALS AND METHODS: Craniocaudal projections of the femur in ventrodorsal position, standard craniocaudal radiographical (SR) and craniocaudal radiography with horizontal radiographical beam (HR), in addition to anatomical macroscopic measurements (A), were obtained from 45 femurs from 23 canine cadavers, for calculation of the canal flare index (CFI). The differences between the radiographical positions and 95% limits of agreement (LOA) between the CFI-SR, CFI-HR and CFI-A were evaluated by the Bland-Altman method. RESULTS: The standard craniocaudal radiograph provided an approximate dimension of the proximal femoral morphology (mean difference: 0.121 ± 0.391) with a different CFI value (p = 0.0341) from CFI-A. The craniocaudal radiograph taken using a horizontal beam provided greater accuracy (mean difference: 0.087 ± 0.42) in our study. CLINICAL SIGNIFICANCE: The craniocaudal projection using a horizontal radiographical beam was more accurate than the standard craniocaudal projection for measurement of the true anatomical dimensions of the canine femur, minimizing the influence of the technique on the CFI values.
