Acetabular orientation variability and symmetry based on CT scans of adults

Abstract

Purpose: Understanding acetabular orientation is important in many orthopaedic procedures. Acetabular orientation, usually described by anteversion and abduction angles, has uncertain measurement variability in adult patients. The goals of this study are threefold: (1) to describe a new method for computing patient-specific abduction/anteversion angles from a single CT study based on the identification of anatomical landmarks and acetabular rim points; (2) to quantify the inaccuracies associated with landmark selection in computing the acetabular angles; and (3) to quantify the variability and symmetry of acetabular orientation.

Methods: A total of 25 CT studies from adult patients scanned for non-orthopaedic indications were retrospectively reviewed. The patients were randomly selected from the hospital's database. Inclusion criteria were adults 20-65 years of age. Acetabular landmark coordinates were identified by expert observers and tabulated in a spreadsheet. Two sets of calculations were done using the data: (1) computation of the abduction and anteversion for each patient, and (2) evaluation of the variability of measurements in the same individual by the same surgeon. The results were tabulated and summary statistics computed.

Results: This retrospective study showed that acetabular abduction and anteversion angles averaged 54 degrees and 17 degrees, respectively, in adults. A clinically significant intra-patient variability of >20 degrees was found. We also found that the right and left side rim plane orientation were significantly correlated, but were not always symmetric.

Conclusion: A new method of computing patient-specific abduction and anteversion angles from a CT study of the anterior pelvic plane and the left and right acetabular rim planes was reliable and accurate. We found that the acetabular rim plane can be reliably and accurately computed from identified points on the rim. The novelty of this work is that angular measurements are performed between planes on a 3-D model rather than lines on 2-D projections, as was done in the past.