3D simulation of radiographic projections to test and reduce the effect of pelvic tilt on the accuracy of cross-table lateral radiography

Background Cross-table lateral (CL) radiography is a convenient and feasible method to assess cup version angle (VA) after total hip arthroplasty; However, pelvic tilt (PT) may contribute to its measurement inaccuracy. How PT affects CL radiographic measurements have not been well studied. We sought (1) to determine the effect of the PT on cup version measurement on CL radiography and (2) to develop a method for reducing measurement errors caused by the PT. Methods We used 3D technique to construct standard model and capture CL radiography simulation. A linear regression model was created to analyze the relationship between PT and VA. CL radiography and computed tomography (CT) were performed for the enrolled patients after surgery. The consistency between CL and CT measurements were verified by intra-class correlation coefficient (ICC). Results There was a high correlation between the VA and PT. For each 1-degree increased in the PT, the VA decreased by 0.76 degrees (R-2 = 0.995, p < 0.001). Based on the data, we created a corrective formula to convert the radiographic measurements into values approximating the actual VA under a natural pelvic position. The VA measurements corrected by our equation was in high agreement with the CT-measured values with reference to the corresponding PT (ICC = 0.988, p < 0.001), which was in sharp contrast to that without PT control (ICC = 0.454, p = 0.203). Conclusions The PT may contribute to cup version measurement inaccuracies on CL radiography. Our mathematical algorithm can serve as a reliable method to improve the accuracy of CL radiography.