Objectives: Age estimation is an integral part of legal investigations for forensic purposes. When the chronological age of the individual is not documented or he/she is under the conflict of the law, age estimation comes into play. The need for age estimation may arise in various legal incidences involving children, and juveniles, for civil aspects such as adoption, child labor, or other criminal proceedings such as rape, kidnapping, and illegal immigration. An accurate and dependable method, that can estimate age with high probability, can aid in narrowing down the list of possible victims or even play a decisive role in such cases. The reliability of the method and the probability of correct age estimation play a decisive role in the court of law. Bedek et al’s model (2019) was recently developed and tested in Croatian and South Indian populations with satisfactory results. As there is no evidence of study in the Western Indian population, looking into the accessibility of the population group, the present study aims to evaluate the validity and reliability of the Bedek method in the Western Indian population. Materials and Methods: Approval was sought from the Institutional Ethical Committee. Five hundred and twenty-five orthopantomographs (OPG) of patients aged 5–15 were obtained. A double-blinded study was done, where the radiographs were analyzed using ImageJ software, independently by two investigators. The data were tabulated and subjected to statistical analysis for accuracy of age estimation and intra- as well as inter-observer reliability. Results: There was an underestimation by a range of −1.3038 to −0.74536. There was underestimation in all the models of Bedek with P < 0.005, for all the teeth models except, the three- and two-teeth model (P > 0.005). Conclusion: In our study, we found that the accuracy of age estimation increases significantly with the number of teeth used. Seven four-teeth models were the most suitable for age estimation on OPG. All models except the three-teeth model and two-teeth model were found to be more accurate.
Read full abstract