Age Estimation from Radiographic Images of the Knee

Medico-legal indicators and cut-offs in different age classes through quantitative analysis of epiphyseal fusion segments on knee CT scans

Scientists are interested in determining age in subadults for several forensic purposes. High- resolution instrumental techniques are being increasingly used for age estimation, driven by the need to minimize errors; in this context, several studies have focused on the knee joint, recognized as a potential site for age examination in late adolescence. We analyzed 200 CT scans performed on Russian subjects (106 males and 94 females) between 13 and 20 years, without growth diseases, endocrine disorders, or osteodystrophy. Each subject underwent two scans, one for each leg. Two indicators were measured for each bone (femur, tibia and fibula): the entire length of the epiphyseal scar and the length of the part/s that is/are fused with metaphysis. Intra class Correlation Coefficient (ICC) was performed to evaluate the intra-operator reproducibility. The ratio between the two lengths was calculated for each bone (FemurR, TibiaR and FibulaR). The first aim was to evaluate a correlation between the ratios of the three bones and the three bones treated as a single parameter (given by the sum of the ratios) versus age. The results showed good correlations in both cases (τ a = 0.74, 0.64, 0.57 and 0.67). The second aim was to estimate the cut-offs derived from the sum of the three ratios respect to four age classes (14-15 years: cut-off ≤ 0.63, 15-16 years: cut-off ≤ 1.19, 16-17 years: cut-off ≥ 0.68 and 17-18 years: cut-off ≥ 1.49. The results from this research encourage further studies of the knee joint as an indicator of legal adult age.

Pediatric age estimation from radiographs of the knee using deep learning

OBJECTIVES

Age estimation, especially in pediatric patients, is regularly used in different contexts ranging from forensic over medicolegal to clinical applications. A deep neural network has been developed to automatically estimate chronological age from knee radiographs in pediatric patients.

METHODS

In this retrospective study, 3816 radiographs of the knee from pediatric patients from a German population (acquired between January 2008 and December 2018) were collected to train a neural network. The network was trained to predict chronological age from the knee radiographs and was evaluated on an independent validation cohort of 423 radiographs (acquired between January 2019 and December 2020) and on an external validation cohort of 197 radiographs.

RESULTS

The model showed a mean absolute error of 0.86 ± 0.72 years and 0.9 ± 0.71 years on the internal and external validation cohorts, respectively. Separating age classes (< 14 years from ≥ 14 years and < 18 years from ≥ 18 years) showed AUCs between 0.94 and 0.98.

CONCLUSIONS

The chronological age of pediatric patients can be estimated with good accuracy from radiographs of the knee using a deep neural network.

KEY POINTS

• Radiographs of the knee can be used for age estimations in pediatric patients using a standard deep neural network. • The network showed a mean absolute error of 0.86 ± 0.72 years in an internal validation cohort and of 0.9 ± 0.71 years in an external validation cohort. • The network can be used to separate the age classes < 14 years from ≥ 14 years with an AUC of 0.97 and < 18 years from ≥ 18 years with an AUC of 0.94.

Contribution of the use of clavicle bone density in age estimation

Age estimation is an essential element in the field of forensics and is also of judicial interest in forensic medicine of the living. Despite all the methods, age estimation is often imprecise with many biases, especially in mature individuals. The main objective of our study is to assess the correlation between age and computed tomography bone mineral density of the medial end of the clavicle. A sample of 180 chest multi-slice computed tomography scans without contrast injection performed in living individuals, with suspected respiratory infection, aged 15 to 100 years was used. Bone density measurement was performed on the medial meta-epiphyseal region of the clavicles. A significant negative correlation was found between bone density and age of individuals (p-value < 0.05). The mean absolute error was calculated for men at 13.4 years and 13.1 years for women, which was associated with an absence of bias. Good precision of the estimate for both sexes was also calculated on a subgroup of individuals whose age was greater than 40 years, with average absolute errors of about 12 years, which was associated with an absence of relative error. In view of our preliminary results, the study of bone density of the medial end of the clavicle could be of interest in the estimation of age. Several studies will be necessary to evaluate the reproducibility of these results on independent bone samples and in the estimation of age in the living individual.

A probability model for assessing age relative to the 18-year old threshold based on magnetic resonance imaging of the knee combined with radiography of third molars in the lower jaw

OBJECTIVE

This study aims to generate a statistical model based on magnetic resonance imaging of the knee and radiography of third molars in the lower jaw, for assessing age relative to the 18-year old threshold.

METHODS

In total, 58 studies correlating knee or tooth development to age were assessed, 5 studies for knee and 7 studies for tooth were included in the statistical model. The relation between the development of the anatomical site, based on a binary system, and age were estimated using logistic regression. Separate meta-populations for knee and tooth were generated from the individual based data for men and women. A weighted estimate of probabilities was made by combining the probability densities for knee and tooth. Margin of errors for males and females in different age groups and knee and tooth maturity were calculated within the larger framework of transition analysis using a logit model as a base. Evidentiary values for combinations of knee and tooth maturity were evaluated with likelihood ratios.

RESULTS

For males, the sensitivity for the method was calculated to 0.78 (probability of correctly classifying adults), the specificity 0.90 (probability of correctly classifying minors), the negative predictive value 0.80 (proportion identified minors are minors) and the positive predictive value 0.89 (proportion identified adults are adults) indicating a model better at identifying minors than adults. The point at which half the female population has reached closed knee lies before the 18-year threshold, adding the knee as an indicator lowers specificity and increases sensitivity. The sensitivity when using tooth as an indicator for females is 0.24 and specificity 0.97, signifying few minors misclassified as adults but also a low probability of identifying adults. The negative predictive value for women when using tooth as the sole indicator is 0.56 and positive predictive value 0.88. Probabilities were calculated for males and females assuming a uniform age distribution between 15 and 21years. The calculated margin of error of minors classified as adults in a population between 15 and 21 years with the model was 11% for males and 12% for females. Further, the evidentiary value as well as margin of error vary for different combinations of knee and tooth maturity.

CONCLUSION

The statistical model based on the combination of MRI knee and radiography of mandibular third molars is a valid method to assess age relative to the 18-year old threshold when applied on males and of limited value in females.

The growth pattern of Neandertals, reconstructed from a juvenile skeleton from El Sidrón (Spain)

Ontogenetic studies help us understand the processes of evolutionary change. Previous studies on Neandertals have focused mainly on dental development and inferred an accelerated pace of general growth. We report on a juvenile partial skeleton (El Sidrón J1) preserving cranio-dental and postcranial remains. We used dental histology to estimate the age at death to be 7.7 years. Maturation of most elements fell within the expected range of modern humans at this age. The exceptions were the atlas and mid-thoracic vertebrae, which remained at the 5- to 6-year stage of development. Furthermore, endocranial features suggest that brain growth was not yet completed. The vertebral maturation pattern and extended brain growth most likely reflect Neandertal physiology and ontogenetic energy constraints rather than any fundamental difference in the overall pace of growth in this extinct human.

Diffusion-Tensor Imaging of the Physes: A Possible Biomarker for Skeletal Growth-Experience with 151 Children

Purpose To determine the changes of diffusion-tensor imaging (DTI) and tractography in the distal femur and proximal tibia related to age, sex, and height. Materials and Methods Following institutional review board approval, with waiver of consent and with HIPAA compliance, the authors retrospectively analyzed DTI images of the knee in 151 children, 73 girls (median age, 14.1 years; range, 6.5-17.8 years) and 78 boys (median age, 16.6 years; range, 6.9-17.9 years), studied from January 2013 to October 2014. At sagittal echo-planar DTI (20 directions, b values of 0 and 600 sec/mm²), regions of interest were placed in the tibial and femoral physes. Using a fractional anisotropy threshold of 0.15 and an angle threshold of 40°, the authors performed tractography and measured apparent diffusion coefficient (ADC) and tract length and volume. Changes related to age, sex, and height were evaluated by using fitted nonlinear polynomial functions on bootstrapped samples. Results Femoral tract volume and length increased and then decreased with age (P < .001); the peaks of femoral tract volume are consistent with the growth spurt, occurring earlier in girls (10.8 years) than in boys (13.0 years) (P < .001). Girls had smaller tract volumes in comparison to boys (P = .013). ADC peaks 2 years earlier than tract volume (girls at 9.3 years, boys at 11.0 years). Girls with greater than 50th percentile of height had longer tracts and greater tract volumes compared with girls with less than 50th percentile (P < .020). DTI parameters of boys do not correlate with percentile of height (P > .300). Conclusion DTI of the physis and metaphysis shows greater tract length and volumes in subjects who are at ages when the growth is fastest. ADC and tract length and volume have an earlier and smaller peak in girls than in boys. Femoral tract length and volume are larger in taller girls. ^© RSNA, 2017.

A Bayesian Approach to Age-at-Death Estimation from Osteoarthritis of the Shoulder in Modern North Americans

Osteoarthritis (OA) is a marker of degeneration within the skeleton, frequently associated with age. This study quantifies the correlation between OA and age-at-death and investigates the utility of shoulder OA as a forensic age indicator using a modern North American sample of 206 individuals. Lipping, surface porosity, osteophyte formation, eburnation, and percentage of joint surface affected were recorded on an ordinal scale and summed to create composite scores that were assigned a specific phase. Spearman's correlation indicated a positive relationship between each composite score and age (right shoulder = 0.752; left shoulder = 0.734). Transition analysis revealed a tendency toward earlier degeneration of the right shoulder. Bayesian statistics generated phase-related age estimates based on highest posterior density regions. Best age estimates were into the seventh decade at the 90th and 50th percentile. The proposed method supplements traditional techniques by providing age estimates beyond a homogenous 50+ age cohort.

Skeletal assessment and secular changes in knee development: a radiographic approach

OBJECTIVES

The purpose of this study is to conduct an analysis of ossification patterns in the distal femoral, proximal tibial, and proximal fibular epiphyses, and the patella. The results generated from this analysis will be compared with previous standards published by Elgenmark () and Garn et al. () to determine if clinical and skeletal age assessment standards should be updated for contemporary Americans.

MATERIALS AND METHODS

Using the Pediatric Radiology Interactive Atlas (Patricia), a total of 1,317 epiphyses were scored for presence or absence from radiographs of 1,056 white individuals born in or after 1990. Statistical modeling of epiphyseal appearance was conducted for all major percentiles, including the 5th and 95th percentiles through logistic regression.

RESULTS

Compared with Elgenmark () and Garn et al. (), our data suggest that the distal femoral and proximal tibial epiphyses show overall earlier ossification, while the proximal fibular epiphysis shows later ossification. When examining the pooled sex 50th percentile for our data, we found that ossification timing differences are 1.2 weeks earlier in the distal femoral epiphysis, 2.1 weeks earlier in the proximal tibial epiphysis, and 1.4 years later in the proximal fibular epiphysis.

DISCUSSION

The epiphyses that appear early in life, for example the distal femoral epiphysis, require gestational age information to accurately estimate appearance times. There are considerable differences between the ossification timing patterns presented in this study and those of previous standards, which did not include gestational ages. Several factors may explain the observed differences in the epiphyses of the knee including: the availability of gestational age information, the analysis of longitudinal versus cross-sectional data, differences in socioeconomic status and prenatal care, and secular change. KEYWORDS age estimation, growth standards, ossification, skeletal maturation, subadult/juvenile growth.

A biochemical approach for assessing cutoffs at the age thresholds of 14 and 18 years: a pilot study on the applicability of bone specific alkaline phosphatase on an Italian sample

BACKGROUND

The majority of age estimation methods analyze morphological changes of specific skeletal (or dental) structures reflecting global bone development (biological parameter) in order to estimate a chronological value. This morphological and structural development is the consequence of a very active tissue metabolism and intensive modeling process which involve both bone formation and bone resorption. Several biochemical markers of bone formation and bone resorption are available, and specific biochemical tests can be performed on blood or urine samples, but such markers of bone turnover have never been employed for age estimation in living individuals for forensic purposes. The aim of this study was to ascertain the applicability of serum bone alkaline phosphatase (BALP) concentration in the age estimation for forensic purposes. We focused on the legal age thresholds of 14 and 18 years (LAT) because, in Italy, the former is considered the minimum age for criminal responsibility and the latter defines adult age and the possibility of applying general criminal laws.

MATERIALS AND METHODS

This study analyzed, from a forensic point of view, BALP and Tanner stages of 202 healthy white individuals (116 females and 86 males) between the ages of 10 and 30 years. We derived a linear logistic model to estimate the probability that an individual was older or younger than LAT using two variables: BALP concentration and Tanner stages. The predictive accuracy of the test was assessed by the determination of the receiver-operating characteristic curve (ROC curve). The test was performed to identify a threshold (cutoff) that could be used to assign an individual to the population of those younger or older than LAT.

RESULTS

ROC curve showed that the use of both serum BALP concentration and Tanner stages has a very good level of reliability in age assessment (the area under the ROC curve, AUC, ranged from 0.918 to 0.962). Best results were obtained in the assessment of male over 18 years of age (sensibility and specificity respectively of 0.90 and 0.93 with an accuracy of 0.92). Worst results were obtained in the assessment of female over 18 years of age (sensibility and specificity respectively of 0.87 and 0.82 with an accuracy of 0.84). We also calculated the probability of the correctness in the age estimation.

CONCLUSION

The results showed that the use of serum BALP concentration in the age assessment could be a promising and integrative method to established ones, but more research has to be done to validate the value of the proposed method in the forensic practice.

Projection radiography of the clavicle: still recommendable for forensic age diagnostics in living individuals?

As superimposition effects often impede the evaluation of the ossification status of the medial clavicular epiphysis in standard posterior-anterior (PA) radiographs, additional oblique images (right anterior oblique, RAO, and left anterior oblique, LAO) are currently recommended to allow for reliable stage assessments. The present study examines the influence of the radiographic projection type on stage determination. To this end, 836 sternoclavicular joints were prospectively obtained during forensic autopsies of bodies aged between 15 and 30 years. Subsequently, three different radiographs (PA, RAO, and LAO) were taken from each specimen and separately evaluated as to the developmental stage of the medial clavicular epiphysis. A forensically established five-stage classification system was used. In 25 % of the cases, the medial clavicular epiphysis depicted in an oblique projection showed a different ossification stage than in the PA projection. In at least 10 % of the cases, a higher ossification stage was observed which would have significant disadvantages in criminal proceedings (ethically unacceptable error). In conclusion, the usage of the current radiographic reference data, which rely upon chest radiographs taken as PA projections, appears to be inadmissible for oblique projections. Projection radiography of the clavicle can therefore no longer be recommended for forensic age estimation practice. As to the question of whether an individual has achieved the age of 18 or 21, computed tomography of the clavicle must be regarded as the exclusive method of choice.