[Measurement of the AT and the CCD angle of macerated cadaver femora: a comparative study between CT and MRI measurements].

Abstract

To compare measurements of the real AT (femoral antetorsion) and CCD (caput-collum-diaphysis) angle by computed tomography (CT) and magnetic resonance imaging (MRI) of hip specimen using direct measurements as reference standard, and to show that measurement by MRI can replace CT measurements and may help avoiding X-ray exposition. CT and MRI measurements were obtained on 25 in water-arrested macerated human femora. Postprocessing was done by 4 independent readers on a workstation using a dedicated 3D-software. Direct measurements of the real AT and the CCD angle were used as reference standard. The analysis included Student's t test for paired values, interobserver variability using intra-class correlation coefficients (ICC), maximum and middle divergence of the angles, and Bland-Altman plots. For determining AT and the CCD angle with CT and MRI, good correlation was found between the 4 readers and with measurements using the reference standard. ICCs were 0.97 and 0.90 for measuring AT and CCD angle with CT, and 0.95 and 0.71 for measurements with MRI, respectively. Mean divergence between CT measurements and those of the reference standard was 0.8 degrees for AT and 0.7 degrees for the CCD-angle. Mean divergence between MRI measurements and those of the reference standard was 0.3 degrees for AT and -0.4 degrees for the CCD-angle. Mean divergences between CT and MRI measurements of AT and CCD-angle were 0.5 degrees. Neither systematic errors nor dependences on the qualitative size of the reference data were evident in the divergences of measurements. Measurements of the real AT and CCD angle by CT and MR imaging revealed a good correlation with direct measurements of the femoral specimen and consequently can be recommended for clinical use. MRI measurements can replace CT measurements, avoiding X-ray exposure especially in young patients undergoing preoperative evaluation for hip dysplasia.