Inter-Rater and Intra-Rater Repeatability and Reliability of EOS 3-Dimensional Imaging Analysis Software.

Abstract

Quantifying ideal component position for the acetabulum and stem during total hip arthroplasty (THA) has been described by many methods. A new imaging method using low-dose digital stereoradiography, the EOS imaging system, is a biplanar low-dose X-ray system that allows for 3-dimensional modeling of lower limbs and semiautomated measurement of pelvic parameters and implant alignment. Twenty-five patients who underwent primary THA by a single surgeon between October 2014 and December 2014 were retrospectively selected. Only patients with unilateral THA without associated spine pathologies were included, totaling 16 right hips and 9 left hips. There were 8 men and 17 women in the cohort, with a mean age of 67 years (range, 53-82). Three individuals performed measurements of pelvic parameters and implant alignment on 3 separate occasions. An interclass correlation of >0.75 was accepted as evidence of excellent agreement and a confirmation of measurement reliability. Before reviewing patient radiographs, 4 pelvic phantom models were analyzed using the EOS 3-dimensional software to verify accuracy. All anatomic and implant measurements performed by the 3 independent reviewers showed interobserver and intraobserver agreement with interclass correlation >0.75. Three-dimensional modeling of hip implants with the EOS imaging system is a reasonable option for the evaluation of component position after THA.