Time-related changes of in vivo projection errors in standardized radiographs.

Subtraction radiography requires radiographs made with identical projection geometry. Recently a new radiographic aiming device and an x-ray beam manipulator made it possible to repeat radiographic exposures in vivo with small angular errors and to measure the angular errors occurring. The purpose of the present investigation was to study in vivo the measuring error and the angular errors produced by this new method at different time points (0, 6, and 12 months) and to assess the effect of different time intervals between exposures (0, 6, and 12 months). Ten sites were investigated. At each time point 2 exposures were made, resulting in 6 radiographs per site. For each pair of radiographs the measuring errors occurring during analysis and the projection errors were determined for each time point and for each time interval. Friedman's test and t-test were used to evaluate the errors. A difference in the measuring error of the system and in the angular errors at the time points 0, 6, and 12 months could not be detected. A time interval between exposures also did not have an effect on the measuring error. Time did not affect the vertical and horizontal angular errors measured between the radiographs producing the smallest oblique error. A combination of radiographs producing the largest oblique angular error showed an increase of the horizontal and vertical angular errors when the time interval increases. These effects, however, were not statistically significant.(ABSTRACT TRUNCATED AT 250 WORDS)