Automatic human identification based on dental X-ray radiographs using computer vision

Automatic personal identification using a single CT image

OBJECTIVES

Computer vision (CV) mimics human vision, enabling computers to automatically compare radiological images from recent examinations with a large image database for unique identification, crucial in emergency scenarios involving unknown patients or deceased individuals. This study aims to extend a CV-based personal identification method from orthopantomograms (OPGs) to computed tomography (CT) examinations using single CT slices.

METHODS

The study analyzed 819 cranial computed tomography (CCT) examinations from 722 individuals, focusing on single CT slices from six anatomical regions to explore their potential for CV-based personal identification in 69 procedures. CV automatically identifies and describes interesting features in images, which can be recognized in a reference image and then designated as matching points. In this study, the number of matching points was used as an indicator for identification.

RESULTS

Across six different regions, identification rates ranged from 41/69 (59%) to 69/69 (100%) across over 700 possible identities. Comparison of images from the same individual achieved higher matching points, averaging 6.32 ± 0.52% (100% represents the maximum possible matching points), while images of different individuals averaged 0.94 ± 0.15%. Reliable matching points are found in the teeth, maxilla, cervical spine, skull bones, and paranasal sinuses, with the maxillary sinuses and ethmoidal cells being particularly suitable for identification due to their abundant matching points.

CONCLUSION

Unambiguous identification of individuals based on a single CT slice is achievable, with maxillary sinus CT slices showing the highest identification rates. However, metal artifacts, especially from dental prosthetics, and various head positions can hinder identification.

CLINICAL RELEVANCE STATEMENT

Radiology possesses a multitude of reference images for a CV database, facilitating automated CV-based personal identification in emergency examinations or cases involving unknown deceased individuals. This enhances patient care and communication with relatives by granting access to medical history.

KEY POINTS

Unknown individuals in radiology or forensics pose challenges, addressed through automatic CV-based identification methods. A single CT slice highlighting the maxillary sinuses is particularly effective for personal identification. Radiology plays a pivotal role in automated personal identification by leveraging its extensive image database.

Accelerating computer vision-based human identification through the integration of deep learning-based age estimation from 2 to 89 years

Computer Vision (CV)-based human identification using orthopantomograms (OPGs) has the potential to identify unknown deceased individuals by comparing postmortem OPGs with a comprehensive antemortem CV database. However, the growing size of the CV database leads to longer processing times. This study aims to develop a standardized and reliable Convolutional Neural Network (CNN) for age estimation using OPGs and integrate it into the CV-based human identification process. The CNN was trained on 50,000 OPGs, each labeled with ages ranging from 2 to 89 years. Testing included three postmortem OPGs, 10,779 antemortem OPGs, and an additional set of 70 OPGs within the context of CV-based human identification. Integrating the CNN for age estimation into CV-based human identification process resulted in a substantial reduction of up to 96% in processing time for a CV database containing 105,251 entries. Age estimation accuracy varied between postmortem and antemortem OPGs, with a mean absolute error (MAE) of 2.76 ± 2.67 years and 3.26 ± 3.06 years across all ages, as well as 3.69 ± 3.14 years for an additional 70 OPGs. In conclusion, the incorporation of a CNN for age estimation in the CV-based human identification process significantly reduces processing time while delivering reliable results.

[Forensic dentistry for identity verification. A survey at the state police level]

BACKGROUND

In Germany, the identification of unknown dead persons is the responsibility of the police. According to INTERPOL standards, primary (e.g., DNA, fingerprints, and teeth) and secondary (e.g., tattoos) characteristics are examined. Forensic dentistry is already used internationally as an efficient method. In this study, the approach of state police in Germany was analyzed. The methods used for identification, the role of forensic dentistry, the cooperation with dentists, and possible optimization approaches are investigated.

METHODS

By means of a digital questionnaire, police officers competent in all federal states for the discovery of unknown dead bodies were asked about identification methods and specifically about the use of forensic dentistry.

RESULTS

Eighty-five officers from at least 11 federal states participated in the survey. The procedure turned out to be department specific. In 72.6% of the cases, different characteristics are combined in the identification process, most frequently DNA with dental status (37.1%). DNA analysis is used most frequently. Of the respondents, 62.9% agreed that dental identification is used "often." The percentage of identifications using dental status is estimated to be 1.6-8.1%. For forensic dentistry, 19.4% have a fixed point of contact. A digital platform to contact dentists was estimated to be helpful by 56.5%.

DISCUSSION

Forensic dentistry is currently still lagging behind DNA analysis, which could change through increasing digitalization if, for example, ante-mortem data are more reliably available and platforms for interdisciplinary exchange are created.

Implementation of a personal identification system using alveolar bone images

OBJECTIVE

In recent years, personal identification has been performed using antemortem panoramic X-ray images and postmortem-CT images. Using these, we have developed a personal identification method that focuses on the alveolar bone. This study examined the effectiveness of this method and aimed to implement a reproducible system.

MATERIALS AND METHODS

For personal identification, a total of 633 CT images and panoramic X-ray images belonging to three groups with different conditions were used.　These images were 160 sets in the same person group and 96,820 in the other groups. The similarity of alveolar bone images was calculated using the landmark method of Procrustes analysis. The processes were system implemented and the methodology was validated.

RESULTS

The ability to identify between the same person group and other person groups showed 0.9769 as the area under the curve (AUC: ROC curve). At the cutoff value of 4.978, there was no false rejection rate, but false acceptance rate was slightly higher.

CONCLUSION

This method was useful as a screening method for personal identification. In addition, system implementation was efficient and reduced human error. In the future, we aim to realize a more efficient personal identification method using distortion-corrected images and including auto-detective landmarks using deep learning.

Automatic identification of individuals using deep learning method on panoramic radiographs

Abstract Background/purpose

The dentition shows individual characteristics and dental structures are stable with respect to postmortem decomposition, allowing the dentition to be used as an effective tool in forensic dentistry. We developed an automatic identification system using panoramic radiographs (PRs) with a deep learning method.

Materials and methods

In total, 4966 PRs from 1663 individuals with various changes in image characteristics due to various dental treatments were collected. In total, 3303 images were included in the data set used for model training. Vgg16, Vgg19, ResNet50, ResNet101, and EfficientNet models were applied for identification. The precision curves were evaluated.

Results

The matching precision rates of all models (Vgg16, Vgg19, ResNet50, ResNet101, and EfficientNet) were examined. Vgg16 was the best model, with a precision of around 80-90% on 200 epochs, using the Top-N metrics concept with 5-15 candidate labels. The model can successfully identify the individual even with low quantities of dental features in 5-10 s.

Conclusion

This identification system with PRs using a deep learning method appears useful. This identification system could prove useful not only for unidentified bodies, but also for unidentified wandering elderly people. This project will be beneficial for police departments and government offices and support disaster responses.

Age group prediction with panoramic radiomorphometric parameters using machine learning algorithms

The aim of this study is to investigate the relationship of 18 radiomorphometric parameters of panoramic radiographs based on age, and to estimate the age group of people with permanent dentition in a non-invasive, comprehensive, and accurate manner using five machine learning algorithms. For the study population (209 men and 262 women; mean age, 32.12 ± 18.71 years), 471 digital panoramic radiographs of Korean individuals were applied. The participants were divided into three groups (with a 20-year age gap) and six groups (with a 10-year age gap), and each age group was estimated using the following five machine learning models: a linear discriminant analysis, logistic regression, kernelized support vector machines, multilayer perceptron, and extreme gradient boosting. Finally, a Fisher discriminant analysis was used to visualize the data configuration. In the prediction of the three age-group classification, the areas under the curve (AUCs) obtained for classifying young ages (10-19 years) ranged from 0.85 to 0.88 for five different machine learning models. The AUC values of the older age group (50-69 years) ranged from 0.82 to 0.88, and those of adults (20-49 years) were approximately 0.73. In the six age-group classification, the best scores were also found in age groups 1 (10-19 years) and 6 (60-69 years), with mean AUCs ranging from 0.85 to 0.87 and 80 to 0.90, respectively. A feature analysis based on LDA weights showed that the L-Pulp Area was important for discriminating young ages (10-49 years), and L-Crown, U-Crown, L-Implant, U-Implant, and Periodontitis were used as predictors for discriminating older ages (50-69 years). We established acceptable linear and nonlinear machine learning models for a dental age group estimation using multiple maxillary and mandibular radiomorphometric parameters. Since certain radiomorphological characteristics of young and the elderly were linearly related to age, young and old groups could be easily distinguished from other age groups with automated machine learning models.

Dental diagnosis for inlay restoration using an intraoral optical coherence tomography system: A case report

PATIENTS

The patient was a 32-year-old man who underwent amalgam restoration of the mandibular right second molar. An amalgam restoration fracture was diagnosed by intraoral optical coherence tomography (OCT), and pulp exposure was examined during cavity preparation. Subsequently, a definitive ceramic restoration was fabricated, and the marginal fit in the oral cavity was evaluated using the OCT system.

DISCUSSION

The existing OCT system cannot acquire images inside the oral cavity because of the large probe size. However, the proposed intraoral OCT system can access the prostheses in the mandibular right second molar. Therefore, dental diagnosis for restoration treatment with dental prosthesis fracture, marginal gap, and pulp exposure after tooth preparation is possible using the proposed intraoral OCT system.

CONCLUSIONS

The use of the intraoral OCT system improved dental diagnosis by allowing the dentist to confirm quantitative values through cross-sectional images, rather than that by determining a treatment plan after visual dental diagnosis.

Mandibular Premolar Identification System based on a Deep Learning Model

OBJECTIVES

For constructing an isolated tooth identification system using deep learning, Igarashi et al. (2021) began constructing a learning model as basic research to identify the left and right mandibular first and second premolars. These teeth were chosen for analysis because they are difficult to identify from one another. The learning method itself was proven appropriate but presented low accuracy. Therefore, further improvement in the learning data should increase the accuracy of the model. The study objectives were to modify the learning data and increase the learning model accuracy for enabling the identification of isolated lower premolars.

METHODS

Static images of the occlusal surface of the premolars made from the dental plaster casts of dental students were used as the training, validation, and test data. A convolutional neural network with 32 hidden layers, AlexNet, convolutional architecture for fast feature embedding, and stochastic gradient descent was used to construct four learning models.

RESULTS

The accuracy of the identification model increased using static images of the occlusal surface of the teeth with the adjacent teeth deleted as the training and validation data; however, a learning model that could perfectly identify the teeth could not be realized.

CONCLUSIONS

Static images of the occlusal surface of the teeth with the adjacent teeth deleted should be used as both training and validation data. The ratio of the numbers of training, validation, and test data should be optimized.

Dental Radiographic/Digital Radiography Technology along with Biological Agents in Human Identification

The heavy casualties associated with mass disasters necessitate substantial resources to be managed. The unexpectedly violent nature of such occurrences usually remains a problematic amount of victims that urgently require to be identified by a reliable and economical method. Conventional identification methods are inefficient in many cases such as plane crashes and fire accidents that have damaged the macrobiometric features such as fingerprints or faces. An appropriate recognition method for such cases should use features more resistant to destruction. Forensic dentistry provides the most appropriate available method for the successful identification of victims using careful techniques and precise data interpretation. Since bones and teeth are the most persistent parts of the demolished bodies in sudden mass disasters, scanning and radiographs are unrepeatable parts of forensic dentistry. Forensic dentistry as a scientific method of human remain identification has been considerably referred to be efficient in disasters. Forensic dentistry can be used for either "sex and age estimation," "Medical biotechnology techniques," or "identification with dental records," etc. The present review is aimed at discussing the development and implementation of forensic dentistry methods for human identification. For this object, the literature from the last decade has been searched for the innovations in forensic dentistry for human identification based on the PubMed database.

A fully automated method of human identification based on dental panoramic radiographs using a convolutional neural network

OBJECTIVES

This study aimed to develop a fully automated human identification method based on a convolutional neural network (CNN) with a large-scale dental panoramic radiograph (DPR) data set.

METHODS

In total, 2760 DPRs from 746 subjects who had 2-17 DPRs with various changes in image characteristics due to various dental treatments (tooth extraction, oral surgery, prosthetics, orthodontics, or tooth development) were collected. The test data set included the latest DPR of each subject (746 images) and the other DPRs (2014 images) were used for model training. A modified VGG16 model with two fully connected layers was applied for human identification. The proposed model was evaluated with rank-1, -3, and -5 accuracies, running time, and gradient-weighted class activation mapping (Grad-CAM)-applied images.

RESULTS

This model had rank-1, -3, and -5 accuracies of 82.84%, 89.14%, and 92.23%, respectively. All rank-1 accuracy values of the proposed model were above 80% regardless of changes in image characteristics. The average running time to train the proposed model was 60.9 s per epoch, and the prediction time for 746 test DPRs was short (3.2 s/image). The Grad-CAM technique verified that the model automatically identified humans by focusing on identifiable dental information.

CONCLUSION

The proposed model showed good performance in fully automatic human identification despite differing image characteristics of DPRs acquired from the same patients. Our model is expected to assist in the fast and accurate identification by experts by comparing large amounts of images and proposing identification candidates at high speed.