Reproducibility of adult facial morphology

Objective: This study was carried out to assess the reproducibility of 3D facial morphology in a cohort of adults. Methods: This was a quantitative evaluation of 80 (M=40, F=40) adults, mean age 24.5 years, at the University Dental Hospital, Wales College of Medicine. Two Minolta Vivid 900 laser scanning devices were placed as a stereo pair to capture the left and right 3D soft tissue profiles of the subjects. 36 subjects were randomly requested to re-attend at the hospital and the 3D scans were repeated. These images were pre-processed and merged to form a 3D soft tissue reproduction of faces using reverse modelling software. The shell deviations between left and right scans of each patient were recorded and analysed for differences. Once the faces were merged, individual subjects who had repeated scans were overlaid onto one another, using the iterative closest point algorithm to determine the reproducibility of facial morphology. Results: The mean shell deviations of left versus right scans were 0.29±0.07mm and 0.32±0.12mm for males and females respectively. Paired t-tests revealed that the mean difference of 0.03mm was statistically insignificant (p>0.05). The difference in the mean scores of the merged faces was 0.03mm. The reproducibility error was 0.7mm and 0.8mm for males and females respectively when a tolerance of 90% was imposed on the aligned faces. Conclusion: A clinical acceptable reproduction of facial morphology was shown in this cohort of adults. However, the error in reproducibility needs to be taken into account when assessing treatment changes.