RÉSUMÉ
Age estimation is an integral part of forensic medical and dental practice. In many countries, the age threshold of 14 years is set to determine the minimum age for criminal responsibility. In the present study, the authors studied the final maturation of the first (IPM1) and second mandibular premolars (IPM2) by Cameriere maturity index and determined cut-offs at the age threshold of 14 years, and validate on the test sample. Orthopantomograms of 960 healthy south Indian children and sub-adults (480 boys and 480 girls) aged between 10 and 18 were analysed, 640 as training sample and 320 as a test sample. The results of logistic regression analysis with age (≥14 years) as a dependent variable and IPM1, IPM2, and sex as predictor variables on training sample showed that both IPM1 and IPM2 significantly related to the legal age of 14 years but not the sex (p = 0.052). A receiver operating characteristic (ROC) curve analysis was performed to evaluate the specific cut-off values of IPM1 and IPM2 for predicting 14 years. A cut- off value of IPM1 < 0.01 and IPM2 < 0.02 were obtained using the highest Youden index value. Our results on test sample showed that the combined predictor, IPM1 + IPM2 < 0.02, showed the highest specificity (97.5% in boys and 92.5% for girls), better post-test probability, 97.2% in boys and 91.9% for girls, and a smaller number of false positives (6.8%). In conclusion, the combination of IPM1 and IPM2 could be useful in determining the age of over 14 years in south Indian children.
Sujet(s)
Adolescent/législation et jurisprudence , Détermination de l'âge dentaire/méthodes , Prémolaire/anatomie et histologie , Mandibule , Apex de la racine de la dent/anatomie et histologie , Facteurs âges , Asiatiques , Enfant , Femelle , Humains , Inde , MâleRÉSUMÉ
This paper concerns a method to estimate age in children using third molar maturity index (I3M) as independent variable and age as the dependent variable. Total of 1283 digital orthopantomograms (OPTs) of south Indian children aged between 7 and 22â¯years was included. Training sample of 933 OPTs was used to derive regression formulae, and validation sample of 350 OPTs used to verify accuracy of newly derived regression models. Regression analysis performed revealed that cubic function gave the best correlation between the studied variables. The models exhibited a high correlation coefficient for third molar development in both sexes (0.81 for boys and 0.84 for girls). When new models were verified, slight overestimation of dental age i.e., 0.2â¯years in boys and 0.13â¯years in girls was observed. Therefore, we conclude by recommending newly derived regression models as they provided better age predicting performances.