Background: Successful hip reconstruction to restore the normal hip biomechanics requires precise placement of implants. Computer assisted navigation in total hip arthroplasty has been proposed to have the potential to help achieve a high accuracy in implant placement. The goal of the study was to evaluate the accuracy of an imageless computer navigation system on cadavers and to validate a non-invasive computed tomography method for post-operative determination of acetabular cup orientation.
Methods: Total hip arthroplasty was performed on seven cadaver hips with the aid of an imageless computer navigation system. The achieved cup orientation were recorded using three methods, (1) intra-operatively using the imageless computer navigation system, (2) post-operatively with direct bone digitization and (3) with a computed tomography based three dimensional model interpreted by three raters. Measurement from the direct bone digitization was taken as the gold standard to evaluate the other two methods. The intra-rater and inter-rater consistency of the computer tomography-model method were assessed by Cronbach's alpha determination.
Findings: Compared with the cup orientation obtained from the direct bone digitization, the average difference for anteversion and abduction were 3.3 (3.5) degrees (P=0.045) and 0.6 (3.7) degrees , respectively, for navigation reading. The average differences for computer tomography-model for three raters were 0.5 (2.1) degrees , 0.8 (1.5) degrees and 3.2 (3.3) degrees (P=0.043) for anteversion and 0.4 (1.6) degrees , 0.3 (1.6) degrees and 2.1 (2.7) degrees for abduction. The intra-rater consistency ranged from 0.626 for a novice rater to over 0.97 for experience raters. The inter-rater consistency (including novice and experienced raters) was over 0.90.
Interpretation: While the values for cup orientation determined with imageless computer navigation were comparable to those from direct bone and implant digitization, the measurement for anteversion obtained was not as accurate as that for abduction. The proposed computer tomography-model method has an excellent intra-rater consistency for experienced raters, as well as an excellent overall inter-rater consistency. The study confirms that a non-invasive computed tomography based model analysis can be used in clinical practice as a valid method for post-operatively evaluating the orientation of the acetabular component.