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/.

Deep learning for automatic mandible segmentation on dental panoramic x-ray images

Many studies in the last decades have correlated mandible bone structure with systemic diseases like osteoporosis. Mandible segmentation, as well as segmentation of other oral structures, is an essential step in studies that correlate oral structures' conditions with systemic diseases in general. However, manual mandible segmentation is a time-consuming and training-required task that suffers from inter and intra-user variability. Further, the dental panoramic x-ray image (PAN), the most used image in oral studies, contains overlapping of many structures and lacks contrast on structures' interface. Those facts make both manual and automatic mandible segmentation a challenge. In the present study, we propose a precise and robust set of deep learning-based algorithms for automatic mandible segmentation (AMS) on PAN images. Two datasets were considered. An in-house image dataset with 393 image/segmentation pairs was prepared using image data of 321 image patient data and the corresponding manual segmentation performed by an experienced specialist. Additionally, a publicly available third-party image dataset (TPD) composed of 116 image/segmentation pairs was used to train the models. Four deep learning models were trained using U-Net and HRNet architectures with and without data augmentation. An additional morphological refinement routine was proposed to enhance the models' prediction. An ensemble model was proposed combining the four best-trained segmentation models. The ensemble model with morphological refinement achieved the highest scores on the test set (98.27%, 97.60%, 97.18%, ACC, DICE, and IoU respectively), with the other models scoring above 95% in all performance metrics on the test set. The present study achieved the highest ranked performance considering all the previously published results on AMS for PAN images. Additionally, those are the most robust results achieved since it was performed over an image set with considerable gender representativeness, a wide age range, a large variety of oral conditions, and images from different imaging scans.

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%.

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).

Jaw and Teeth Segmentation on the Panoramic X-Ray Images for Dental Human Identification

Due to the damage to biometric properties in the event of natural disasters, like fire or earthquakes, it is very difficult to identify human remains. As teeth are more durable than other biometric properties, identifying information obtained from them is much more reliable. Therefore, in cases where alternative biometric properties cannot be obtained or used, information taken from teeth may be used to identify a person's remains. In recent years, many studies have shown how the identification process, previously performed manually by a forensic dental specialist, can be made faster and more reliable with the assistance of computers and technology. In these studies, the x-ray image is subdivided into meaningful parts, including jaws and teeth, and dental properties are extracted and matched. In order to extract the features accurately and ensure better matching, it is important to segment images properly. In this study, (i) lower and upper jaw and (ii) tooth separation was performed to segment panoramic dental x-ray images to assist in identifying human remains. To separate the jaws, a novel meta-heuristic optimization-based model is proposed. To separate teeth, a user-assisted, semi-automatic approach is presented. The proposed methods have been performed with a computer program. The results of the implementation of these methods of jaw and tooth separation in panoramic tooth images are encouraging.

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.

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

A person may be identified by comparison between ante- and post-mortem dental panoramic radiographs (DPR). However, it is difficult to find reference material if the person is unknown. This is often the case when victims of crime or mass disaster are found. Computer vision can be a helpful solution to automate the finding of reference material in a large database of images. The purpose of the present study was to improve the automated identification of unknown individuals by comparison of ante- and post-mortem DPR using computer vision. The study includes 61,545 DPRs from 33,206 patients, acquired between October 2006 and June 2018. The matching process is based on the Speeded Up Robust Features (SURF) algorithm to find unique corresponding points between two DPRs (unknown person and database entry). The number of matching points found is an indicator for identification. All 43 individuals (100%) were successfully identified by comparison with the content of the feature database. The experimental setup was designed to identify unknown persons based on their DPR using an automatic algorithm system. The proposed tool is able to filter large databases with many entries of potentially matching partners. This identification method is suitable even if dental characteristics were removed or added in the past.