OBJECTIVE: The objective of this study was to check the reproducibility and influencing parameters of landmarks on three-dimensional images of 4- to 6-year-old Caucasian children. PROBANDS AND METHODS: We examined the reproducibility of 22 landmarks on six three-dimensional facial scans of 4- to 6-year-old randomly-selected kindergarten children. The 3D scans were taken by the digital system faceSCAN II under standardized conditions. One specially-trained investigator marked all landmarks on each of the six 3D scans a total of ten times. Ten different orthodontic residents also placed all landmarks on each of these scans. Standard deviations from the mean were then calculated from the data for each individual landmark on the x-, y- and z-axes. RESULTS: The specially-trained investigator who had placed all landmarks on each of the six 3D scans a total of ten times found 15 of 22 landmarks to be reproducible well to within a standard deviation of less than 1 mm on each spatial plane. For the ten orthodontic residents, only three landmarks were found to be reproducible to within a standard deviation of under 1 mm for each spatial plane. CONCLUSION: Familiarity with 3D facial scans and their corresponding software programs, together with good image quality, improve the reproducibility of the analysis. Landmarks revealing poor reproducibility should be used, if at all, with due caution for analysis. It remains to be seen whether they can be omitted altogether, or whether they should be re-set.
OBJECTIVE: The objective of this study was to check the reproducibility and influencing parameters of landmarks on three-dimensional images of 4- to 6-year-old Caucasian children. PROBANDS AND METHODS: We examined the reproducibility of 22 landmarks on six three-dimensional facial scans of 4- to 6-year-old randomly-selected kindergarten children. The 3D scans were taken by the digital system faceSCAN II under standardized conditions. One specially-trained investigator marked all landmarks on each of the six 3D scans a total of ten times. Ten different orthodontic residents also placed all landmarks on each of these scans. Standard deviations from the mean were then calculated from the data for each individual landmark on the x-, y- and z-axes. RESULTS: The specially-trained investigator who had placed all landmarks on each of the six 3D scans a total of ten times found 15 of 22 landmarks to be reproducible well to within a standard deviation of less than 1 mm on each spatial plane. For the ten orthodontic residents, only three landmarks were found to be reproducible to within a standard deviation of under 1 mm for each spatial plane. CONCLUSION: Familiarity with 3D facial scans and their corresponding software programs, together with good image quality, improve the reproducibility of the analysis. Landmarks revealing poor reproducibility should be used, if at all, with due caution for analysis. It remains to be seen whether they can be omitted altogether, or whether they should be re-set.