IN VITRO COMPARISON OF TWO METHODS OF DETECTION OF THE FUNCTIONAL HIP CENTER VS. ANATOMICAL HIP CENTRE IN COMPUTER ASSISTED SURGERY

INTRODUCTION In orthopedic surgery, the lower limb alignment defined by the HKA parameter i.e. the angle between the hip, knee and ankle centers, is a crucial clinical criterion used for the achievement of several surgeries. It can be intraoperatively determined with Computer Assisted Orthopedic Surgery (CAOS) systems by computing the 3D location of these joint centres. The hip centre used for the computation of the HKA is defined by the experts as the anatomical centre of the femoral head. However, except for Total Hip Replacement procedure, the hip joint is not accessible and the hip center is computed using functional methods. The two most common are the Least Moving Point (LMP) and the Pivoting (PIV). MATERIALS AND METHODS We have analysed on six cadaveric lower limbs the intra-observer variability of both the anatomical and the functional hip centres. The differences between the HKAs angle obtained with the anatomical hip centre ( HC ANAT ) and those obtained with the functional hip centres coming from the LMP ( HC LMP ) and the PIV ( HC PIV ) algorithms have also been analysed. RESULTS The intra-observer variability was on average (standard deviation) 0.9(0.6)mm, 9(5.2)mm and 7.5(4)mm for respectively the HC ANAT , HC LMP and the HC PIV variations. The average impact on the HKA was 1° and 0.8 ° respectively for HC LMP and HC PIV with a maximum of 4°. DISCUSSION Several papers in the literature have studied the accuracy and the robustness of methods allowing CAOS systems to determine the functional hip centre. All of these studies have been performed with simulated data. This study shows results coming from in-vitro data. The results concerning the intra-observer variability shows that the procedure is very robust and reproducible for the determination of HC ANAT . However, functional methods are much less reproducible even if the Pivoting method seems to be a little better. Given these results, the impact of the functional methods on the HKA has been analysed. We have therefore compared the HKA obtained with HC ANAT with those obtained with HC LMP and HC PIV . The results are extremely encouraging since, despite the intra-observer variability, the differences between the anatomical and the functional HKAs are, on average, less than 1° with a maximum inferior to 4°. The impact on the HKA is therefore limited and the accuracy of the functional methods to assess the HKA are sufficient regarding the clinical needs.