Errors of computer-assisted migration analysis in conventional radiographs of femoral hip implants--an experimental...

Several methods have been described in the literature in order to analyze migration of femoral hip implants in conventional radiographs. However, no measurements were done regarding the potential errors inherent in such unstandardized radiographs of daily routine. In order to quantify this lack of reproducibility, we carried out experimental examinations with radiographs of a hip revision prosthesis, where different variables of technical X-ray conditions and femoral positions were changed. All radiographs were analyzed with a computer-assisted processing system by different procedures. At first, the radiographs were calibrated geometrically. Then, particular reference distances between defined points of the implant and cortical bone structures were analyzed quantitatively. The error of the axial migration analysis was up to 11.11 mm concerning different femoral positions, up to 8.29 mm in the case of different source-to-film distances and up to 2.21 mm due to different lateral localizations of the central X-ray focus. Significant errors for the migration analysis in the transverse and rotational planes were only found under conditions of varying the femoral position, and not under different technical X-ray conditions. From these experimental results, we derived the following criteria to minimize failure in a quantitative radiographic migration analysis: (1) the patient's leg has to be placed into an antirotation device to ensure identical object position; (2) the same source-to-film distance has to be used; (3) the central X-ray focus has to be localized on the center of the film-cassette; (4) film-screen systems should be of the same type and size; (5) the object has to be placed in the same position as in previous radiographs. As a conclusion, only if these standardization criteria are respected in the daily routine of conventional radiographs will an effective and meaningful use of migration analysis systems be possible to prove or to exclude mechanical failure of femoral hip implants in prospective longitudinal follow-up series.