In vivo determination of radiographic projection errors produced by a novel filmholder and an x-ray beam manipulator.

Subtraction radiography requires radiographs made with identical projection geometry. The purpose of the present investigation was to develop and evaluate a new in vivo method to measure angular radiographic projection error. This included the development of a system for accurate control of the x-ray beam angulation relative to the intraoral film and the evaluation of the in vivo projection errors (equivalent to the "angular difference" between repeated exposures) produced by the new system. A filmholder, an x-ray beam manipulator, and an aiming device were newly constructed. Reference wires were attached for measuring the projection error on the filmholder and on the 22 bicuspids at the sites of investigation in 16 patients. Two exposures using standardized projection conditions were made of each site. The relative position of the reference wires on the duplicate radiographs was used to calculate the angular difference in the direction of the x-ray beam which occurred between the 2 exposures. Triplicate measurements allowed the determination of measurement error. The angular projection errors were measured in a horizontal and vertical direction. The reference wires allowed determination of the angular projection errors relative to the filmholder, relative to the dentition and the angular error due to the repositioning of the filmholder in the patient's mouth. The average projection errors deviated only slightly from zero. The largest deviation registered was -0.11 degrees. This applies to the horizontal and vertical direction and all relationships. The standard deviations varied between 0.403 degrees and 0.697 degrees. This applies to the standard deviation for all relationships in the horizontal, the vertical, and the oblique directions. The measuring error found from the triplicate measurements had a standard deviation of 0.08 degrees to 0.19 degrees depending on the distance between the reference wires. This new system for controlling in vivo projection errors seems to provide highly standardized radiographs in vivo. The new measuring system proved to be accurate enough to measure the in vivo projection errors between repeated radiographs produced with the new filmholder and x-ray beam manipulator. The in vivo errors produced by the new system are markedly smaller than all in vivo and in vitro errors previously reported in the literature.