Anatomical reliability of two fundamental radiological and clinical pelvic parameters: incidence and thickness

Reliability and reproducibility of two radiological pelvic parameters are tested: thickness (length of the segment defined by the middle of the upper endplate of the sacrum to the middle of the bi-coxo-femoral axis) and pelvic incidence (angle defined by the perpendicular line to the centre of the upper endplate of the sacrum and the thickness line). These two parameters provide a pelvis description and assess the relation between pelvis and spinal curves. The anatomical reliability of these radiological parameters was not achieved. The values of these two parameters from X-ray versus direct measurement on 12 anatomical specimens are compared. The direct measurement was performed by means of an electromagnetic Fastrak system (Polhemus society) providing 3D position of anatomical landmarks and allowing to measure the incidence and the thickness. These parameters were also measured from sagittal X-ray. Their values were compared. Incidence: the paired t-test and the variance ratio test were not statistically significant and a highly significant positive correlation existed between anatomical and radiological values (r=0.98; P<0.001). Thickness: the paired t-test was significant (P<0.01). There was a negative correlation between anatomical values of incidence and thickness (r= −0.54; P<0.05) but not for radiological values. A strong correlation exists which validates the radiological measurement of an angle, i.e. incidence, although there is a lack of reliability of the X-ray measurement of a distance, i.e. thickness, which is due to technical conditions of the X-ray examination. The results of this study suggest that in daily practice the X-ray measurement of the incidence only may be considered as an accurate indicator of pelvis morphology for the study of relations between pelvis anatomy and spinal curves (e.g., lordosis, scoliosis, spondylolisthesis).