Cephalometric landmark identification within the petrous temporal region.

The literature is reviewed regarding the errors associated with relocating cephalometric petrous temporal landmarks. An experimental cranial fixation device, to be used with cephalostat ear rods that subsequently 'hinge away', is presented. Twenty-five consecutive cephalograms taken by this method (Group B) are compared with another two groups of 25 radiographs. The second group (Group A) constitutes those taken with perspex ear rods in situ, whilst the third group (Group C) constitutes those with wooden ear rods. The variance ratios demonstrate that the error in relocating most of the petrous temporal landmarks is significantly greater in Group B, when compared to either Group A or C. In comparing these latter two groups, the variance in Group A is greater for all these landmarks, significantly so for condylion and porion. Some possible factors contributing to these findings are discussed. It would seem that wooden ear rods produce less error in the relocation of the petrous temporal landmarks; that the absence of ear rods produces greater error; and that the presence of perspex ear rods increases the error in the relocation of anatomical porion.