A precise receptor-positioning device for subtraction radiography, based on cross-arch stabilization.

OBJECTIVES: To devise and test for reproducibility a new patient-beam alignment device for subtraction radiography.
METHODS: A rigid, cross-arch bar was used with bite-registration material. A modified Rinn XCP rod and ring were firmly attached to the bar and were placed in contact with the X-ray cone. The receptor was held by a slot in the bar. Duplicate sets (n=8 pairs) of digital radiographs were made with a modified, calibrated DenOptix photostimulable phosphor system. Image pairs were registered and compared for geometrical and intensity errors by means of quantitative subtraction radiography. The reproducibility of patient-beam alignment for each pair was determined by measuring spatial errors at alveolar crest edges and by comparing magnifications. Intensity errors measured at two predetermined regions of alveolar bone were used to estimate corresponding bone-mass errors.
RESULTS: Misalignment at the alveolar crest was generally <1 pixel (0.085 mm), and everywhere < or =0.25 mm. The magnification was constant within a relative standard deviation of 0.13% (n=8). Trabecular features were generally invisible or barely visible in subtraction images. The standard deviation of intensity errors was 1.07% (n=16). This corresponds to a change in bone mass of approximately 2% or less.
CONCLUSION: The device has potential for clinical population studies in which the goal is to detect small changes in bone mass and alveolar-crest height. It is convenient to use and comfortable for patients. Because of the cross-arch design, patient-beam alignment is expected to be relatively insensitive to bite-registration errors, tooth movement, and tooth loss.