New regression models for dental age estimation in children using third molar maturity index: A preliminary analysis testing its usefulness as reliable age marker.

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 22 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., 0.2 years in boys and 0.13 years in girls was observed. Therefore, we conclude by recommending newly derived regression models as they provided better age predicting performances.

Validation of third molar maturity index (I3M) for discrimination of juvenile/adult status in South Indian population.

Deliberate falsification of age was considered to be one of the main reasons for forensic age estimation of the living individuals. This posed to be a challenging task during criminal and legal proceedings, and ultimate care must be taken not to classify juveniles as adults. Third molars are the only developing teeth during late adolescence and early adulthood. Our study was designed to analyze the usefulness of the third molar maturity index (I3M) specific cut-off value (I3M < 0.08) to discriminate adults (≥18 years) and juveniles (<18 years) in South Indian children. 216 panoramic radiographs (114 females and 102 males) of living subjects aged between 14 and 21 years were analyzed. Our results demonstrated high sensitivity (83.3% and 90.2%) and specificity (98.3% and 95.1%) for females and males respectively. The positive likelihood ratios of being adult were 50.00 and 18.35 while the negative likelihood ratios were 0.17 and 0.10 in females and males respectively. The estimated posttest probability was 98.0% in females and 94.8% in males. The obtained results showed that the specific cut-off value of I3M < 0.08 may be a useful additional tool in discrimination of individuals who are around 18 years of age.