An accurate method of radiological assessment of acetabular volume and orientation in computed tomography spatial reconstruction

Marek Jóźwiak, Michał Rychlik, Bartosz Musielak, Brian Po-Jung Chen, Maciej Idzior, Andrzej Grzegorzewski

Two-dimensional (2D) measurements of acetabular morphology and orientation are well known; there is less information on these acetabular characteristics in three dimensions. One important reason is the lack of standardized reference planes for the pelvis, especially in relation to the spinopelvic unit; another is that no method precisely assesses the acetabulum in three-dimensional (3D) orientation based on its axis rather than on the directions of the edges of the acetabular rim. We present an objective, highly reliable and accurate, axis-based approach to acetabular anthropometry in the measurement of acetabular volume and spatial orientation in both normal and pathologic hips. This was done using reference planes based on the sacral base (SB) and true acetabular axis in 3D computed tomography (CT) pelvic reconstruction. Radiological examinations of 30 physiologic pelves (60 acetabula) were included in the study. Reliability and accuracy of the method were verified by comparing acetabular angles in 2D pelvic scans with 3D reconstructions. We also applied the method to two pathologic acetabula. Comparison of axis position in the horizontal plane revealed significant positive correlations between 2D angle measurements (acetabular anteversion angle [AAA] and anterior acetabular index [AAI]) and 3D measurement of anteversion angle (p < 0.001 and p = 0.012, respectively). In the frontal plane, there was no difference between abduction angle, measured on topogram, and inclination angle, obtained from a 3D model (p = 0.517). In the sagittal plane, there was a significant negative correlation between AAA and acetabular tilt (p < 0.001). Inter- and intra-observer reproducibility was excellent for determination of the sacral-base plane and assessment of volume, with Fleiss κ coefficients of 0.850 and 0.783, respectively, and intraclass correlation coefficients of 0.900 and 0.950, respectively. Inter-observer reproducibility for evaluation of acetabular axis ranged from 0.783 to 0.883, and intra-rater reliability ranged from 0.850 to 0.900 for all 3D angles. Our method is a new, reliable diagnostic tool for assessing the acetabula in both normal and pathologic hip joints. The sacral-base plane can be used as a stable reference that takes the relationship of the acetabulum to the spinopelvic unit into consideration.