Research Progress on Dental Age Estimation Based on MRI Technology.

Dental age estimation is a crucial aspect and one of the ways to accomplish forensic age estimation, and imaging technology is an important technique for dental age estimation. In recent years, some studies have preliminarily confirmed the feasibility of magnetic resonance imaging (MRI) in evaluating dental development, providing a new perspective and possibility for the evaluation of dental development, suggesting that MRI is expected to be a safer and more accurate tool for dental age estimation. However, further research is essential to verify its accuracy and feasibility. This article reviews the current state, challenges and limitations of MRI in dental development and age estimation, offering reference for the research of dental age assessment based on MRI technology.

Semi-supervised automatic dental age and sex estimation using a hybrid transformer model.

Teeth-based age and sex estimation is an important task in mass disasters, criminal scenes, and archeology. Although various methods have been proposed, most of them are subjective and influenced by observers' experiences. In this study, we aimed to develop a deep learning model for automatic dental age and sex estimation from orthopantomograms (OPGs) and compare to manual methods. A large dataset of 15,195 OPGs (age range, 16 ~ 50 years; mean age, 29.65 years ± 9.36 [SD]; 10,218 females) was used to train and test a hybrid deep learning model which is a combination of convolutional neural network and transformer model. The final performance of this model was evaluated on additional independent 100 OPGs and compared to the manual method for external validation. In the test of 1413 OPGs, the mean absolute error (MAE) of age estimation was 2.61 years by this model. The accuracy and the area under the receiver operating characteristic curve (AUC) of sex estimation were 95.54% and 0.984. The heatmap indicated that the crown and pulp chamber of premolars and molars contain the most age-related information. In the additional independent 100 OPGs, this model achieved an MAE of 3.28 years for males and 3.79 years for females. The accuracy of this model was much higher than that of the manual models. Therefore, this model has the potential to assist radiologists in automated age and sex estimation.

[Establish Assessment Model of 18 Years of Age in Chinese Han Population by Mandibular Third Molar].

OBJECTIVE: To explore the value of estimating chronologic age based on the grades of mandibular third molar development. To evaluate whether mandibular third molar could be used as an indicator for estimating the age under or over 18 years. METHODS: The mineralization status of mandibular third molar of 1 845 individuals aged 10 - 30 was graded and marked based on Demirjian's classification of grades reformed by Orhan. Gender difference was examined by t-test. A cubic regression model was established to analyze the correlation between third molar and chronologic age. Each grade of age cumulative distribution diagram and ROC curve was respectively performed to evaluate the relationship between third molar and the age of 18. Using Bayes discriminant analysis, an equation was established for estimating the age of 18. RESULTS: The inner-rater reliability was 0.903. Statistical analysis showed a moderate correlation between age and grade. Significant differences of both genders were found only in grade D and H (P < 0.05). Males at the grades from 1 to D and females at the grades from 1 to C were under 18 years old, and both males and females at grade H were over 18 years old. The area under the ROC curve was 0.797 (P < 0.05). CONCLUSION: Third molar development shows a high correlation with age, and combined with other indicators, it can be used to estimate the age of 18.

Machine Learning Supported the Modified Gustafson's Criteria for Dental Age Estimation in Southwest China.

Adult age estimation is one of the most challenging problems in forensic science and physical anthropology. In this study, we aimed to develop and evaluate machine learning (ML) methods based on the modified Gustafson's criteria for dental age estimation. In this retrospective study, a total of 851 orthopantomograms were collected from patients aged 15 to 40 years old. The secondary dentin formation (SE), periodontal recession (PE), and attrition (AT) of four mandibular premolars were analyzed according to the modified Gustafson's criteria. Ten ML models were generated and compared for age estimation. The partial least squares regressor outperformed other models in males with a mean absolute error (MAE) of 4.151 years. The support vector regressor (MAE = 3.806 years) showed good performance in females. The accuracy of ML models is better than the single-tooth model provided in the previous studies (MAE = 4.747 years in males and MAE = 4.957 years in females). The Shapley additive explanations method was used to reveal the importance of the 12 features in ML models and found that AT and PE are the most influential in age estimation. The findings suggest that the modified Gustafson method can be effectively employed for adult age estimation in the southwest Chinese population. Furthermore, this study highlights the potential of machine learning models to assist experts in achieving accurate and interpretable age estimation.

DHI-GAN: Improving Dental-Based Human Identification Using Generative Adversarial Networks.

In this work, a novel semisupervised framework is proposed to tackle the small-sample problem of dental-based human identification (DHI), achieving enhanced performance via a "classifying while generating" paradigm. A generative adversarial network (GAN), called the DHI-GAN, is presented to implement this idea, in which an extra classifier is also dedicatedly proposed to achieve an efficient training procedure. Considering the complex specificities of this problem, except for the noise input of the generator, an identity embedding-guided architecture is proposed to retain informative features for each individual. A parallel spatial and channel fusion attention block is innovatively designed to encourage the model to learn discriminative and informative features by focusing on different regional details and abstract concepts. The attention block is also widely applied to the overall classifier to learn identity-dependent information. A loss combination of the ArcFace and focal loss is utilized to address the small-sample problem. Two parameters are proposed to control the generated samples that are fed into the classifier during the optimization procedure. The proposed DHI-GAN framework is finally validated on a real-world dataset, and the experimental results demonstrate that it outperforms other baselines, achieving a 92.5% top-one accuracy rate. Most importantly, the proposed GAN-based semisupervised training strategy is able to reduce the required number of training samples (individuals) and can also be incorporated into other classification models. Our code will be available at https://github.com/sculyi/MedicalImages/.

A fine-grained network for human identification using panoramic dental images.

When accidents occur, panoramic dental images play a significant role in identifying unknown bodies. In recent years, deep neural networks have been applied to address this task. However, while tooth contours are significant in classical methods, few studies using deep learning methods devise an architecture specifically to introduce tooth contours into their models. Since fine-grained image identification aims to distinguish subordinate categories by specific parts, we devise a fine-grained human identification model that leverages the distribution of tooth masks to distinguish different individuals with local and subtle differences in their teeth. First, a bilateral branched architecture is designed, of which one branch was designed as the image feature extractor, while the other was the mask feature extractor. In this step, the mask feature interacts with the extracted image feature to perform elementwise reweighting. Additionally, an improved attention mechanism was used to make our model concentrate more on informative positions. Furthermore, we improved the ArcFace loss by adding a learnable parameter to increase the loss of those hard samples, thereby exploiting the potential of our loss function. Our model was tested on a large dataset consisting of 23,715 panoramic X-ray dental images with tooth masks from 10,113 patients, achieving an average rank-1 accuracy of 88.62% and rank-10 accuracy of 96.16%.

Dental age estimation of Tibetan children and adolescents: Comparison of Demirjian, Willems methods and a newly modified Demirjian method.

Tibetan ethnic group is one of the oldest ethnic groups in China and South Asia. This study set out to analyze the dental development and validate Demirjian method and Willems method in estimating dental age of Tibetan children and adolescents, and to modify Demirjian method based on Tibetan population to provide ethnic-specific reference data and a more reliable method for forensic age assessment in Tibetan ethnic group. In this study, 1951 samples aged between 4 and 15 years were retrospectively collected and analyzed. Multiple linear regression was used to establish relationship between chronological age (CA) and developmental stages of left mandibular permanent teeth. The accuracy of the modified method was tested and compared with that of Demirjian and Willems method. Results showed that dental maturity score (DMS) was significantly greater in girls than in boys in all age groups except for the 4-year age group (p < 0.05). Mean absolute error (MAE) was 0.96 years for both boys and girls by Demirjian method, and 1.06 and 1.16 years for boys and girls respectively by Willems method. Adjusted scores table was established and tested. The age of boys was overestimated by 0.13 years and the age of girls was underestimated by 0.06 years by the adjusted scores table. MAE was lower than that of the other two methods. In conclusion, Demirjian method and Willems method was not sufficiently accurate in estimating dental age of Tibetan population. The modified method was more suitable for dental age estimation of Tibetan population than Demirjian and Willems method.

LCANet: Learnable Connected Attention Network for Human Identification Using Dental Images.

Forensic odontology is regarded as an important branch of forensics dealing with human identification based on dental identification. This paper proposes a novel method that uses deep convolution neural networks to assist in human identification by automatically and accurately matching 2-D panoramic dental X-ray images. Designed as a top-down architecture, the network incorporates an improved channel attention module and a learnable connected module to better extract features for matching. By integrating associated features among all channel maps, the channel attention module can selectively emphasize interdependent channel information, which contributes to more precise recognition results. The learnable connected module not only connects different layers in a feed-forward fashion but also searches the optimal connections for each connected layer, resulting in automatically and adaptively learning the connections among layers. Extensive experiments demonstrate that our method can achieve new state-of-the-art performance in human identification using dental images. Specifically, the method is tested on a dataset including 1,168 dental panoramic images of 503 different subjects, and its dental image recognition accuracy for human identification reaches 87.21% rank-1 accuracy and 95.34% rank-5 accuracy. Code has been released on Github. (https://github.com/cclaiyc/TIdentify).

Automatic human identification from panoramic dental radiographs using the convolutional neural network.

Human identification is an important task in mass disaster and criminal investigations. Although several automatic dental identification systems have been proposed, accurate and fast identification from panoramic dental radiographs (PDRs) remains a challenging issue. In this study, an automatic human identification system (DENT-net) was developed using the customized convolutional neural network (CNN). The DENT-net was trained on 15,369 PDRs from 6300 individuals. The PDRs were preprocessed by affine transformation and histogram equalization. The DENT-net took 128 × 128 × 7 square patches as input, including the whole PDR and six details extracted from the PDR. Using the DENT-net, the feature extraction took around 10 milliseconds per image and the running time for retrieval was 33.03 milliseconds in a 2000-individual database, promising an application on larger databases. The visualization of CNN showed that the teeth, maxilla, and mandible all contributed to human identification. The DENT-net achieved Rank-1 accuracy of 85.16% and Rank-5 accuracy of 97.74% for human identification. The present results demonstrated that human identification can be achieved from PDRs by CNN with high accuracy and speed. The present system can be used without any special equipment or knowledge to generate the candidate images. While the final decision should be made by human specialists in practice. It is expected to aid human identification in mass disaster and criminal investigation.

[Indexes of forensic identification by the digital orthopantomogram of the normal teeth].

OBJECTIVE: To explore more concise and unified forensic identification indexes for people with none dental disease in digital orthopantomogram. METHODS: To select randomly 170 digital orthopantomogram with none dental disease. Then to select indexes for full dentition patterns and dental alignment patterns according to the dental physiological variations and the characters of dental alignment respectively. Finally diversity of the indexes would be evaluated by statistical analysis. RESULTS: The group with none dental disease had 74 kinds of full dentition pattern in 170 samples, thus its diversity was 43.53%. The group had 129 kinds of dental alignment pattern, thus its diversity was 75.88%. The group had 150 kinds of full dentition/dental alignment pattern, thus its diversity was 88.24%. CONCLUSION: The diversity of the full dentition pattern was not very good. So the full dentition coding was not very effective when it was used solely. The diversity of dental alignment pattern was good. So the method of dental alignment coding could be used in the maxillofacial forensic identification. If the group was coded by the full dentition and dental alignment pattern at the same time, its diversity was better than any single pattern. So the method would be valuable in forensic identification.