Photocephalometry: Errors of projection and landmark location

A method called photocephalometry was recently described for the possible soft-tissue evaluation of orthognathic surgery patients by the superimposition of coordinated cephalographs and photographs. A grid analysis was performed to determine the accuracy of the superimposition method. In addition, the reliability of landmark identification was analyzed by the method error of Baumrind and Frantz, using three replicates of twelve patients' photographs. Comparison of twenty-one grid intervals showed that the magnification of the photographic image for any given grid plane is not correlated to that of the radiographic image. Accurate comparisons between soft- and hard-tissue anatomy by simply superimposing the images are not feasible because of the difference in the enlargement factors between the photographs and x-ray films. As was noted by Baumrind and Frantz, a wide range exists in the variability of estimating the location of landmarks. Sixty-six percent of the lateral photographic landmarks and 57% of the frontal landmarks had absolute mean errors for all twelve patients that were less than or equal to 2.0 mm. In general, the envelope of error for most landmarks was not circular. Although the photocephalometric apparatus as described by Hohl and colleagues does not yield the desired quantitative correlation between hard and soft tissues, valuable quantitative information on soft tissue can be easily obtained with the standardization and replication possible with the camera setup and enlarged photographs.