Abstract
Radiostereometric analysis (RSA) has become the gold standard technique for measuring implant migration and wear following joint replacement due to its high measurement precision and accuracy. However, RSA is conventionally performed using two oblique radiographic views with the presence of a calibration cage. Thus, a second set of radiographs must be acquired for clinical interpretation, for example anterior-posterior and cross-table lateral views following total hip arthroplasty (THA). We propose a modification to the RSA setup for examining THA, in which RSA measurements are performed from anterior-posterior and lateral views, with the calibration cage images acquired separately from the patient images. The objective of the current study was to compare the accuracy and precision of the novel technique to the conventional technique using a phantom.
X-ray cassette holders were developed to enable simultaneous acquisition of anterior-posterior and cross-table lateral radiographs with the patient in a supine position in the RSA suite. A Sawbones phantom with total hip implant components was attached to a micrometer-driven stage. The femoral component was translated known distances relative to the acetabular cup in all planes, mimicking head penetration due to wear. Double RSA examinations were acquired for each increment using the traditional and novel radiograph orientations. Translations were measured from the radiographic images using RSA software. For both techniques, accuracy was calculated by comparing the measured translations to the known translation from the micrometer, and reported as the 95% confidence interval. Precision was measured by comparing the measured translations between the double exams, and reported as the standard deviation.
Accuracy was greater for the conventional technique in the inferior-superior axis (p = 0.03), greater for the novel technique in the anterior-posterior axis (p = 0.01), and equivalent in the medial-lateral axis (p = 0.06). Overall accuracy for both the conventional and novel techniques was identical at ±0.022 mm. Precision was equivalent between both techniques for the medial-lateral (p = 0.68), inferior-superior (p = 0.14), and anterior-posterior axes (p = 0.86). Overall precision for the conventional technique was ±0.127 mm and for the novel technique was ±0.095 mm.
Utilising standard clinical radiograph view angles within an RSA exam had no detrimental effect on wear measurement accuracy or precision. This reduces the barriers to implementing RSA imaging in routine follow-up of arthroplasty patients, potentially greatly increasing the numbers of patients that can have quantitative data on implant performance. Future applications can involve applying more clinically relevant radiograph view angles to RSA exams of the knee and shoulder.
