Reproducibility of maxillofacial landmark identification on three-dimensional cone-beam computed tomography images of patients with mandibular prognathism: Comparative study of a tentative method and traditional cephalometric analysis.

Objective: To clarify the reproducibility of a tentative method for identifying maxillofacial landmarks on three-dimensional (3D) images obtained with cone-beam computed tomography (CBCT) for dental use in patients with mandibular prognathism. Also, the influence of level of experience of dentists applying the method was investigated by dividing them into two groups according to experience. Materials and Methods: Dentists with less (group A) or more (group B) than 3 years of experience of cephalometry and 3D image manipulation analyzed CBCT data from 10 patients. using two different landmark identification methods: method 1 used conventional cephalometric definitions and method 2 used detailed landmark identification definitions developed for each cross-sectional plane. The plotting of nine landmarks was performed twice, and 10 coordinate values were obtained for each landmark. To assess reproducibility, the 95% confidence ellipse method was used. Results: Comparative analysis showed that method 2 was highly reproducible. Group B subjects attained smaller ellipsoid volumes than group A subjects, regardless of the landmark identification method used. With method 1, except for condyle and coronoid process, all landmarks showed a higher level of reproducibility in group A subjects than in group B subjects. With method 2, however, five landmarks showed no differences between the methods. Conclusion: The method proposed here may be highly reproducible regardless of the evaluators' experience.