Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Introduction: In forensic context, age estimation in adolescents and young adults is essential to prove whether or not the individual is a minor or an adult. The study aim is to systematically review the literature regarding the age prediction performances of dental age estimation methods based on the development of one or more third molars combined. The set research question is: “What is the most accurate dental age estimation method based on third molar(s) development? Materials and methods: This systematic review was performed according to criteria set forward by PRISMA statement. An electronic search until 03/03/2020 was performed in 4 databases: MEDLINE/PubMed, Cochrane library, EMBASE and Web of Science. Results: 113 studies were considered eligible for inclusion after review of title/abstract, full-text and reference list. Since we only discussed the point prediction and related uncertainty aspect, the age prediction performance and outcomes reported as the error, mean error, mean absolute error, RMSE (root mean squared error), R2 and others were included and classified. The ultimate data extracted for analysis were classified in 3 tables (respectively based on a validation sample): a) mean error (chronological age- estimated age) +- standard deviation, mean absolute error and R*2 number=6, b）regression formula +- standard deviation and R*2 number=11. c) regression model + root mean squared error and R*2 number=5. Discussion: The staging technique turned to be the most accepted registration predictor status compared to others, maybe due to the early origin and many developed branches. And it’s obvious that Demirjian et al. 1973 and Köhler Gleiser and Hunt 1955(modified) outperformed the others. Correspoundingly, the regression formula, linear regression models and Bayesian models were broad used to calculate the estimated ages as age estimation methods. a) The Demirjian et al. method performed better (José 2005) compared to others, especially with low ME(0.07-0.1)+S.D(1.22-1.23), MAE(0.074-0.106) and high R*2(0.45-0.54) and QA score (10) in a large Spanish subadult population. b) The Iranians turned out to have better result in 4TM and 2TM. By the Köhler Gleiser and Hunt 1955(modified) staging technique, they gained a low standard deviation and a high R*2 together with high QA score 9 for the tooth position examined in the study. And in general, the Belgian population performed best in 3TM examined. While for 1 TM, Thais population gained the lowest S.D (M: 1.29-1.3 F: 1.47-1.48) and a extremely high R*2 (M: 0.892 F: 0.85-0.86) using Demirjian technique. c) The Brazilians were found having better scores in RMSE 0.74-1.56 for both genders in linear regression models. Conclusion: In forensic practice, it is difficult to select a method for age estimation based on the development of third molars. In current systematic review, the Demirjian et al.1973(both technique and method)provided the most accurate point prediction of age in Spanish population, Köhler Gleiser and Hunt 1955(modified) technique + regression formula method did the most accurate prediction in Iranians population(4TM and 2TM), Demirjian technique + regression formula outperformed others in Thais population (1TM), and Köhler Gleiser and Hunt 1955(modified) technique + linear regression model method performed best in Brazilians, while more research is needed to verify these methods in the future.

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Age is an important feature of a person. Sometimes this information is not available. This is possible when you deal with migration. Because The official documentation is not available. Or a corpse where there is no identification available. In this case, a method is needed to determine the age. There exist methods to determine the age based on teeth information. However, these method’s rely on manual staging of the teeth which is time consuming. In this thesis, an automatic approach is researched to determine the age of a person. In this approach, you try to predict the age based on dental X-ray images. The proposed method tries to solve this problem using convolutional neural networks. This neural network is trained on a dataset of 4000 pictures of people between the ages of 8 and 25 year. The thesis focuses on preprocessing the data, selecting the best possible neural net architecture and finetuning the hyperparameters to get the best result possible. In the end, the proposed method was evaluated using 5-fold cross validation. The result is an MAE is 089 years and the MSE is 1.45 years. These results are very competitive towards the already available methods.

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Dental identification relies on comparing dental treatment features, odontogenic morphological features or a combination of both. Treatment features are usually the first line of assessment used. However, factors such as incomplete dental records, ongoing decline in number of dental restorations has made morphological features become more important. There is at the moment, no conclusive evidence that the human dentition is indeed unique, which is the basis in being able to use morphological features for identification. The aim of this study is to identify unique tooth dimensions on panoramic radiograph for human identification purposes. Sixty-two digital panoramic radiograph from 31 females and 31 males (age ranged from 13 to 56 years old) were collected retrospectively from a private dental clinic in Belgium. Six tooth measurements were performed on all seven mandibular left permanent teeth (excluding third molars): tooth length (TL), crown length (CL), root length (RL), crown width (CW), cervical width (CEJW) and root width (RW). 9 length-width ratios were then calculated using these measurements: CL/CW, CL/CEJW, CL/RW, RL/CW, RL/CEJW, RL/RW, TL/CW, TL/CEJW, TL/RW. Analyses for the ratios were based on log-transformed values. Intra and inter observer reliabilities for each parameter were assessed in terms of intra-class correlation (ICC), standard error of measurement (SEM) and reproducibility coefficient (RC). Sex differences in measurements/ratios per tooth position were evaluated using two-way linear models and correlations between measurements/ratios per tooth position were verified with Spearman correlations. Univariate and multivariate analyses were performed to calculate ‘potential set’ for an indication of uniqueness. Combining all measurements for tooth 31 produced the lowest potential set, which indicates the most unique tooth for identification purposes. Compared to single parameters, combining measurements or ratios substantially improve the mean potential set.

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Background During human identification, forensic experts often encounter decomposed and fragmentary human remains. Poorly preserved bodies, as well as fragmented remains, limit considerably the methods used for identification and significantly increase the complexity of the identification procedures. Aim The aim of the current study is to register and evaluate measurements between mandibular landmarks and to quantify their uniqueness and their potential for identification of human remains. Materials and methods One hundred eighty-five digital panoramic radiographs (94 males and 91 females; age range 16-66 years) were retrospectively collected from the dental clinic files of the Audacia Dental Clinic, Brussels, Belgium, and imported in image enhancement software. Fourty-one parameters were measured (5 linear, 6 angles and 10 ratios) on both sides of the mandible. Univariate and multivariate analyses were performed to quantify the potential for identification of each parameter and their combinations. Results Measurements E and R 3 proved to be the most useful in identification. The best identifying measurement combination uses all the ratios parameters. Discussions and Conclusions The method described in this study allows for human identification by narrowing the AM search field based on the available PM panoramic radiograph.