Third molar development in a London population of White British and Black British or other Black ethnicity

Population differences in dental development between Black and White ethnic groups have been debated but not previously studied in the UK. Using inappropriate data for dental age estimation (DAE) could lead to erroneous results and injustice. Data were collected from dental panoramic radiographs of 5590 subjects aged 6–24 years in a teaching hospital archive. Demirjian stages were determined for left-sided teeth and third molars and data collected regarding hypodontia and third molar agenesis. Third molar development in self-assigned Black British, including other self-assigned Black ethnicity, was compared with that of self-assigned White British subjects. Data were compared for males and females in the two ethnic groups using T-tests for Demirjian Stages A–G of third molar development and Mann-Whitney tests for Stage H once a cut-off age at the maximum age for Stage G had been imposed. Third molar development occurred earlier in subjects of Black ancestry compared to those of White ancestry. While both ethnic groups showed large age ranges for every third molar stage, in female subjects these generally occurred at least 1.5 years earlier, and in males at least one year earlier. Hypodontia and third molar agenesis were more prevalent in White British, but the ethnic difference in third molar development persisted in subjects with complete dentitions. This is a large study that confirms ethnic differences in a London population, emphasises the difficulties of establishing the 18-year-old threshold using DAE, and confirms the risk of overestimating the age of individuals of Black ethnicity using White ethnic reference data.     

18岁年龄推断的法医影像学研究进展

18周岁年龄节点在司法审判、移民、竞技体育中均有重大意义,18岁年龄推断是法医实践工作的难点之一,也是法医基础研究应面对的问题。目前常用的指标包括锁骨胸骨端、髂嵴、第三磨牙、四肢骨近关节端。18岁年龄推断的常用影像方法多样,包括平片、CT、超声、MRI,薄层CT是目前主要研究手段,但逐渐向无辐射的MRI发展。统计方法多采用一般描述性分析,但其统计效能低,因此逐渐有作者提出应用ROC、数据挖掘技术等高功效的统计方法推断年龄。既往研究显示上述指标的完全发育成熟在18岁年龄推断中有一定意义,可用于18岁年龄推断,多指标的联合应用及高功效的统计方法将会提高18岁年龄推断的准确性。     

Third Molar Development: Evaluation of Nine Tooth Development Registration Techniques for Age Estimations

Multiple third molar development registration techniques exist. Therefore the aim of this study was to detect which third molar development registration technique was most promising to use as a tool for subadult age estimation. On a collection of 1199 panoramic radiographs the development of all present third molars was registered following nine different registration techniques [Gleiser, Hunt (GH); Haavikko (HV); Demirjian (DM); Raungpaka (RA); Gustafson, Koch (GK); Harris, Nortje (HN); Kullman (KU); Moorrees (MO); Cameriere (CA)]. Regression models with age as response and the third molar registration as predictor were developed for each registration technique separately. The MO technique disclosed highest R² (F 51%, M 45%) and lowest root mean squared error (F 3.42 years; M 3.67 years) values, but differences with other techniques were small in magnitude. The amount of stages utilized in the explored staging techniques slightly influenced the age predictions.     

基于数据挖掘实现成人喉软骨和舌骨CT图像重组的年龄推断

目的运用数据挖掘技术探索喉软骨和舌骨CT图像重组用于成人年龄推断的可行性。方法收集413例颈部薄层CT图像,样本年龄范围18~<80岁,将样本随机分为测试集和训练集。参照TURK等的分级方法,对所有样本进行全方位综合阅片分级,将甲状软骨骨化过程分为6级,环状软骨骨化过程分为5级,舌骨大角和舌骨体的骨性结合分为3级。采用scikit-learn 0.17机器学习工具包(Python语言)建立成人年龄推断的多元线性回归模型、支持向量回归模型和贝叶斯岭回归模型。应用留一交叉验证和测试集评估模型性能。结果所有指标与年龄的相关性均呈中等或较差。男性样本应用支持向量回归模型的准确性最高,平均绝对误差为8.67岁,明显优于多元线性回归模型和贝叶斯岭回归模型;女性应用支持向量回归模型准确性最高,平均绝对误差为12.69岁,但与其他两种模型准确性的差异无统计学意义。结论应用数据挖掘技术有助于提高推断成人年龄的准确性,但基于喉软骨和舌骨的成人年龄推断误差仍较大,实际应用中应结合其他指标综合推断年龄。     

Age Estimation in Living Adults using 3D Volume Rendered CT Images of the Sternal Plastron and Lower Chest

Age estimation is commonly of interest in a judicial context. In adults, it is less documented than in children. The aim of this study was to evaluate age estimation in adults using CT images of the sternal plastron with volume rendering technique (VRT). The evaluation criteria are derived from known methods used for age estimation and are applicable in living or dead subjects. The VRT images of 456 patients were analyzed. Two radiologists performed age estimation independently from an anterior view of the plastron. Interobserver agreement and correlation coefficients between each reader's classification and real age were calculated. The interobserver agreement was 0.86, and the correlation coefficients between readers classifications and real age classes were 0.60 and 0.65. Spearman correlation coefficients were, respectively, 0.89, 0.67, and 0.71. Analysis of the plastron using VRT allows age estimation in vivo quickly and with results similar than methods such as Iscan, Suchey–Brooks, and radiographs used to estimate the age of death.     

Dental Age Estimation: A Test of the Reliability of Correctly Identifying a Subject Over 18 Years of Age Using the Gold Standard of Chronological Age as the Comparator

This study was designed to test the reliability of using the third molars to demarcate between child and adult status. A total of 2000 dental panoramic tomographs were used for assessment of the calculated age using the 8‐stage system of tooth development and applied to all four third molars. The LL8 was also assessed using this 8‐stage system. For each tooth development stage, the Normal distribution and percentile summary data were estimated. The calculated dental age compared with the chronological age was statistically significantly different (p < 0.001) for both females and males giving underestimates of the true age. Comparison of single tooth dental age and chronological age was only slightly different. The most important finding is that the assignment to above or below the 18‐year threshold, in the age range 17 years to 19 years, could be wrong on up to 50% of occasions.     

Pubic Symphyseal Age Estimation from Three‐Dimensional Reconstructions of Pelvic CT Scans of Live Individuals

Skeletal age estimation using the pubic symphysis is a reliable and popular practice among forensic anthropologists. In instances in which accessing the skeleton is impossible, such as in the case of living individuals, three‐dimensional (3D) reconstructions of computed tomography (CT) scans can be created and analyzed for aspects of the biological profile. This study examined 3D reconstructions of pelvic CT scans of a sample of 44 live adults to estimate age using the Suchey–Brooks J Hum Evol 1990;5(3):227‐38 method. Images were created using OsiriX software (Pixmeo, Geneva, Switzerland). Estimated age ranges based on the 3D‐CT images captured the true age of the individual at the time of the CT scan 79.5% of the time with good intra‐observer reliability. The features best visualized on the 3D images included the ossific nodule and the pubic symphyseal rim. This study shows that clinical CT scanning parameters are adequate for creating 3D images to be analyzed for age using the pubic symphysis.     

The third molar as an age marker in adolescents: new approach to age evaluation

Adolescence is a relatively short period between childhood and adulthood. It is very difficult to determine adulthood based on biological indicators. The third molar may be considered a potential age marker for the period between the ages of 16–21. Our study evaluated a set of 1700 panoramic radiographs of individuals aged between 5 and 21 years. Results confirmed the statistically significant difference in the course of third molars development. The mean deviation for individuals with one third molar agenesis is ?0.98 years, for individuals with two third molars agenesis ?1.89 years, and with three molars agenesis ?3.28 years. Thus, the extent of the deviation is directly proportional to the number of unformed third molars. The calculation of age according to the mean of stages of all third molars could lead to the underestimation of age. No intergender differences were found. Age determination using third molars could be used for forensic purposes.     

An Analysis of Hounsfield Unit Values and Volumetrics from Computerized Tomography of the Proximal Femur for Sex and Age Estimation

Sex and age are two elements in the establishment of a biological profile for forensic identification. While the pelvic bones are the most ideal structures for sex estimation, the condition of a body is not always ideal due to the nature of death, such as in mass disasters, or postmortem processes. This study utilized CT scans and resultant 3D models of 100 male and 100 female adults of known ages ranging from 18 to 98 years old to collect volumetric and Hounsfield unit measurements of the proximal femur. Equations were created to establish logistic regression models for sex estimation and linear regression models for age estimation. The resultant sex estimation method had an accuracy of 93.5% and utilized the volume of the proximal femur. This study provides three linear regression models for age with an accuracy range of 86%–92% ±12 years. As imaging technologies are increasingly adopted for forensic purposes, the power of 3D data will provide the opportunity for more quantitative and reproducible analyses. The proposed method for sex and age estimation provides a reliable tool that can be utilized in both day-to-day casework and disaster victim identification.     

Radiographic evaluation of third molar development in relation to chronological age among Turkish children and youth

A sharp increase in forensic age estimation of living persons has been observed in recent years. However, ethnic populations residing in different countries have been insufficiently analyzed. According to 2004 data compiled by the Essen-based Turkey Research Center, there are 3.8 million Turkish people living abroad, and 3.2 million of them reside in European Union countries. Despite the high number of Turks living abroad, little is known about third-molar development for forensic application in this population. Hence, it was considered worthwhile to determine the developmental stages of the third molar in a group of Turkish population, to assess chronological age estimation based on the developmental stages, and to compare third molar development according to sex, age and location. Orthopantomograms of 1134 Turkish patients, ages 4-20 years were examined and third-molar developmental stages were evaluated based on Demirjian's classifications. Orthopantomograms were scored by two different observers, and Wilcoxon matched-pairs signed-ranks test used to test intra- and inter-observer reliability revealed a strong agreement between both intra- and inter-observer measurements. Linear regression analysis was performed to correlate third-molar development and chronological age, and further statistical analysis was performed to determine the relation between sex, age and location. Results showed a strong linear correlation between age and molar development (males: r(2)=0.57; females: r(2)=0.56). Mineralizations of left and right third molars were compared using Wilcoxon tests, and no statistical differences were found. No significant differences were found in third-molar development between males and females. Mandibular third molar crypt formation was observed in 2.4% of patients at age seven and maxillary third molar crypt formation was observed in 1.3% of patients at age seven. A strong correlation was found between third-molar development and chronological age. Among the Turkish population, third molar crypt formation is observable at as early as 7 years in both the mandible and maxilla. Agenesis can be determined conclusively if no radiolucent bud is present by age 14.     

Morphometric analysis of third molar root development by an experimental method using digital orthopantomographs

Abstract:  The aim of the study was to conduct a morphometric analysis of the root development of the third molar, with the purpose of overcoming the limits of an exclusively morphological analysis. The study was divided in two phases. The first one was the verification of the existence of a constant crown-root (C/R) ratio on a sample of 100 third molars, characterized by a complete root development. The value obtained was used in the second phase to predict the final root's length, knowing the crown height. So we have calculated, on a sample of 322 third molars with developing roots, the final ratios between incomplete roots and complete roots. Statistical analysis was then made with 90%, 95%, and 99% confidence intervals. The results showed a significant difference according to the age of the subject and the limit values, lower and upper, for subjects aged 16 and 17 years versus those aged 18 and 19. For each analysis, the width of the class of tolerance and other statistical indicators were calculated. The results we obtained support the advantage of a morphometric study as opposed to an exclusively morphological study, but all the techniques used to determine the age of a living subject can provide only indications of the biological age, but no certainties as regards chronological age.     

The Reliability of Chronological Age Determination by Means of Mandibular Third Molar Development in Subjects in Croatia*

Abstract: The aim of this study was to determine the development of the mandibular third molar and its relation to chronological age. The evaluated material consisted of 979 orthopantomograms of patients of Croatian Caucasian origin aged between 5.7 and 14.6 years. Third molar developmental stages were evaluated according to the stages proposed by Nolla. The frequencies of different stages of mineralization and the mean value of the mineralization of the mandibular third molars with regard to age, gender, and size of the mandible were determined and the coefficient of correlation determined between the age of the subject and the stage of development of the mandibular molars. Statistically significant correlation was determined between mineralization of the mandibular third molar and chronological age of the subjects (boys, mandibular left third molar r = 0.779, right third molar r = 0.793; girls, mandibular left third molar r = 0.746, right third molar r = 0.725). It can be concluded that the accuracy of age estimation based on Nolla’s method is applicable for Croatian children.     

Accuracy of Identifying Juvenile/Adult Status from Third Molar Development using Prediction Probabilities Derived from Logistic Regression Analysis

The use of third molars in predicting juvenile/adult status (</≥ 18 years) has important legal ramifications. Third molar development was assessed using Köhler's grading on 268 orthopantomograms of Indian subjects. Logistic regression analysis was applied to determine allocation accuracy of juvenile/adult status and the level of probability that is “reliable” in predicting juvenile/adult status. Allocation accuracies ranged between 75.8% and 78.2% for the sexes combined, with minimal male‐female differences. Adults were categorized more accurately than juveniles, suggesting that Köhler's grading puts Indian juveniles at greater risk of unwarranted punishment. In both sexes, juvenile/adult status was “reliably” predicted when the probability was >80% using individual third molars (excepting the lower right third molar in males); combining upper and lower third molars on the left/right sides, “reliable” predictions were possible when the probability was >80% and >90% for females and males, respectively. Overall, “reliable” juvenile/adult status prediction was achieved in c. 36% of subjects.     

Adapting and Applying the Phenice (1969) Adult Morphological Sex Estimation Technique to Subadults

This research evaluated whether adult morphological sex estimation methods of the innominate could be adapted and applied to subadults. The subpubic concavity, described by Phenice (1969) and revised by Klales et al. (2012), was modified for use with subadults. Two observers scored radiographic images from the PATRICIA database of 334 individuals of both sexes aged between 1.19 and 20.47 years. Score frequencies shifted from score 2 (straight) to higher frequencies of score 3 (convexity) in males and score 1 (concavity) in females with increasing age. Using ordinal logistic regression, sex classification was highest for the oldest age cohort at 97.2% and then decreased by age cohort. Interobserver error rendered a high level of agreement (0.806) using the intraclass correlation coefficient. Results indicate that the Klales method can be modified and applied to subadults to accurately estimate sex following the onset of puberty with a high degree of reliability and validity.     

Reliability of Third Molar Development for Age Estimation in a Texas Hispanic Population: A Comparison Study*

Abstract: Evaluating third molars from 950 Hispanic individuals aged 12–22 years using Demirjian’s schematic for crown and root formation found that Hispanic third molar development was 8–18 months faster than American Caucasians as reported by Mincer, Harris and Berryman in 1993. This represents a statistically significant increase. Earlier development was more apparent in the later stages F through H. Hispanic males reach developmental stages faster than Hispanic females and maxillary third molars reach developmental stages faster than mandibular third molars in both sexes. The earliest age observed for stages B–H (e.g., Stage H first observed at age 13.92 years in females) and the oldest age observed for Stages B–G were developed to facilitate age prediction of unknown individuals. Prediction tables for minimum and maximum age for an observed stage (e.g., if a female maxillary third molar is stage F it means she is older than 13 years) for each sex‐jaw group were calculated.     

Age estimation using CT imaging of the third molar tooth, the medial clavicular epiphysis, and the spheno-occipital synchondrosis: a multifactorial approach

A multi-factorial method for estimating age was devised based on the development of the 3rd molar tooth, the medial clavicular epiphysis, and the spheno-occipital synchondrosis, using multiple regression as the means to construct age estimation formulae and CT scanning as the imaging modality. The sample consisted of approximately 600 individuals from a contemporary Australian population, between the ages of 15 and 25 years, who were admitted to the Victorian Institute of Forensic Medicine, Melbourne, Australia, for the purposes of medico-legal death investigation. Results show that the spheno-occipital synchondrosis does not contribute to the age estimation model for this age cohort. The regression computation for the 3rd molar tooth and medial clavicle, when combined into a single multiple regression calculation, provides a robust model with tighter age ranges at the 95% confidence interval (CI) than when each age marker is used individually. This research provides a method to estimate age for unknown age Australian individuals in the problematic age group of 15-25 years with greater precision than previously possible.     

Effect of Lower Third Molar Segmentations on Automated Tooth Development Staging using a Convolutional Neural Network

Staging third molar development is commonly used for age estimation in subadults. Automated developmental stage allocation to the mandibular left third molar in panoramic radiographs has been examined in a pilot study. This method used an AlexNet Deep Convolutional Neural Network (CNN) approach to stage lower left third molars, which had been selected by manually drawn bounding boxes around them. This method (bounding box AlexNet = BA) still contained parts of surrounding structures which may have affected the automated stage allocation performance. We hypothesize that segmenting only the third molar could further improve the automated stage allocation performance. Therefore, the current study aimed to determine and validate the effect of lower third molar segmentations on automated tooth development staging. Retrospectively, 400 panoramic radiographs were collected, processed and segmented in three ways: bounding box (BB), rough (RS), and full (FS) tooth segmentation. A DenseNet201 CNN was used for automated stage allocation. Automated staging results were compared with reference stages – allocated by human observers – overall and per stage. FS rendered the best results with a stage allocation accuracy of 0.61, a mean absolute difference of 0.53 stages and a Cohen's linear κ of 0.84. Misallocated stages were mostly neighboring stages, and DenseNet201 rendered better results than AlexNet by increasing the percentage of correctly allocated stages by 3% (BA compared to BB). FS increased the percentage of correctly allocated stages by 7% compared to BB. In conclusion, full tooth segmentation and a DenseNet CNN optimize automated dental stage allocation for age estimation.     

Third molar development according to chronological age in populations from Spanish and Magrebian origin

Spain is frequently the entrance country into the European Union for undocumented immigrants, especially those from the Magreb (Northern Africa). Forensic age estimates for these persons are difficult because systematic studies of dental maturity are lacking. Three different populations were analyzed to determine the pattern of development of third molars as a tool for age estimation in people of different ethnic and geographic origin. Orthopantomograms from two different populations of Spanish origin (Galicia in northwestern continental Spain, and Ceuta, a Spanish province in Northern Africa) were compared to radiographs of molars from a Magrebian population (Northern Africa) resident in Ceuta. Orthopantomograms were obtained from a private dental clinic (n=344) in Galicia and from the Public Oral Health Services (n=228) in Ceuta. We looked for differences in third molar mineralization (determined with the Demirjian scale) that might serve as age indicators (older versus younger than 18 years). Differences in maturation patterns were found between sexes and populations of origin. Mineralization of tooth 38 was more advanced in males than females among subjects 18 years of age and older in all three populations. Moreover, mineralization of tooth 38 in subjects aged 18 years and older was significantly slower in the Spanish-Galicia population than in the Magrebian-Ceuta population. We found no significant differences between Magrebian and Spanish individuals from Ceuta. We conclude that differences in tooth 38 mineralization may be related more with socio-geographical than ethnic origin (ancestry), and thus constitute evidence of the relevance of socio-geographic rather than genetic factors in third molar development. We used ROC analysis to determine the accuracy of the examiner's ability to correctly estimate age as younger or older than 18 years. The results suggest that Demirjian stage for tooth 38 can be considered a good indicator of age in all three populations.     

The Utility of the Proximal Epiphysis of the Fifth Metatarsal in Age Estimation

Radiographs of 277 living individuals were assessed via a numerical scoring system to determine the timing of appearance and degree of fusion between the proximal epiphysis of the fifth metatarsal and its diaphysis. The epiphysis was observed to first appear in females at 8 years and 10 years in males and fuse by 14 years in females and 15 years in males. When assessing the level of agreement of category assignment, inter‐observer agreement was 78% for females and 64% for males whereas intra‐observer agreement was 77% for females and 86.1% for males. These results suggest that the maturation of the proximal epiphysis of the fifth metatarsal may be of value in age estimation in the child and that the scoring system is sufficiently robust to merit continued investigation. Previously this epiphysis has been considered an inconstant feature, but this research confirmed its presence in all individuals studied.