A morphometric assessment of developing permanent dentition for age and sex estimation in a select South African sample

Abstract

Dentition plays a crucial role in the fields of forensic science, forensic odontology and anthropology for age and sex estimation. The aim of this study was to determine the degree of sexual dimorphism and age prediction capabilities of permanent dentition using morphometric analysis. Six hundred digital panoramic radiographs (n = 600), belonging to 300 South African Black and 300 South African Indian, aged between 5.00 and 19.99 years were retrospectively examined using a cross-sectional design. The Pearson Correlation Coefficient tests were conducted to determine if a correlation exist between age, sex and each morphometric parameter. “Stepwise Regression Analysis” and “supervised machine learning classification” with a recursive feature elimination and logistic regression were then utilized to parsimoniously prune the morphometric parameters to determine the best models for age and sex estimation, respectively. Males generally displayed larger tooth dimensions than females, with the South African Black population group having larger tooth dimensions than the South African Indian population group. The morphometric parameters of the dentitions showed little sexual dimorphism, with weak correlations less than 0.1, in this study. However, strong correlations between age and the tooth length of the second and third maxillary and mandibular molars (R2 > 0.89) were recorded for the select South African Black and Indian population groups. In conclusion, the sex estimation formulae generated in this study had low performance accuracies for both population groups. However, the age estimation formulae developed from “Stepwise Regression Analysis” in this study were reliable predictors of age, with the tooth and root lengths displaying the best models for age estimation for the South African Black and Indian sample (R2 > 0.9).