Pelvic rotations: a pelvic phantom study

Abstract

Radiographic measurements made on standard pelvic radiographs are commonly used in studying conditions related to the hip joints. Effects caused by variations in pelvic orientation may be a source of error in comparing measurements between sequential radiographs. To define and characterize parameters able to measure rotational differences separately around two axes and altered radiographic focusing along two axes when sequential standard anteroposterior (AP) pelvic radiographs are compared. A pelvic phantom was constructed based on direct three-dimensional measurements of five defined landmarks in a pelvic model. Two ratios, the vertical and transversal rotation ratios, were defined using radiographs of the phantom. The phantom was radiographed in 33 different orientations and with 16 different radiographic focuses using a specially constructed tilt table. On each radiograph, measurements were made and the two rotation ratios were calculated using a measurement program. Linear correlations between pelvic rotations around one axis and the corresponding rotation ratio were found with almost no influence of simultaneous rotation around the other axis. Also, linear correlations were found between altered radiographic focusing along one axis and the non-corresponding rotation ratio. Rotational differences around two axes or altered radiographic focusing along two axes can be measured independently. Effects caused by rotations cannot be distinguished from effects caused by altered radiographic focusing.