Computer-based evaluation of gender identification and morphologic classification of tooth face and arch forms

Statement of problem. In prosthodontics, sex-related differences in tooth forms are assumed, and optimal esthetics is further assumed to be achieved only if the face, arch, and tooth forms are in harmony. However, there is no universally accepted rule or mathematical method to define these morphologic features. Purpose. The purpose of this study was to evaluate whether anterior tooth from reflected gender and to question the hypothesis that there was a correlation among face, arch, and tooth forms on the basis of dentists' perceptions. Material and methods. Photographic records including full face, dental arch, and anterior tooth images of 60 dental students were made by use of digital photography. A database was arranged as a questionnaire for evaluation by prosthodontists who were asked to define the sex and morphologic features of the subjects by examining stored images of the subjects. Thirteen prosthodontists, all with at least 10 years of experience, served as experts and performed these evaluations twice at 2-week intervals. Kappa statistic was used to measure intraobserver and interobserver agreement. Results. The average correctly identified values in females and males were 53% and 58%, respectively. However, the proportions of females and males who were identified correctly and completely the same in both the first and second evaluations were 36% for females and 41% for males. According to the observations of the experts, face-to-arch form (54%) and face-to-tooth form (51%) correlations were higher than the tooth-to-arch form correlation. The overall correlation among face, arch, and tooth forms was 31%. The experts were in fair agreement in defining tooth forms. Conclusion. The results of gender identification of the subjects showed that the experts were unsuccessful in distinguishing the actual sex by the visual assessment of anterior tooth segment alone, with a fair agreement between 2 evaluation sessions. There were also significant interobserver and intraobserver differences in classifying face, arch, and especially tooth forms. Furthermore, the correlation among these forms was not highly defined.