Deep focus approach for accurate bone age estimation from lateral cephalogram.

Background/purpose: Bone age is a useful indicator of children's growth and development. Recently, the rapid development of deep-learning technique has shown promising results in estimating bone age. This study aimed to devise a deep-learning approach for accurate bone-age estimation by focusing on the cervical vertebrae on lateral cephalograms of growing children using image segmentation. Materials and methods: We included 900 participants, aged 4-18 years, who underwent lateral cephalogram and hand-wrist radiograph on the same day. First, cervical vertebrae segmentation was performed from the lateral cephalogram using DeepLabv3+ architecture. Second, after extracting the region of interest from the segmented image for preprocessing, bone age was estimated through transfer learning using a regression model based on Inception-ResNet-v2 architecture. The dataset was divided into train:test sets in a ratio of 4:1; five-fold cross-validation was performed at each step. Results: The segmentation model possessed average accuracy, intersection over union, and mean boundary F1 scores of 0.956, 0.913, and 0.895, respectively, for the segmentation of cervical vertebrae from lateral cephalogram. The regression model for estimating bone age from segmented cervical vertebrae images yielded average mean absolute error and root mean squared error values of 0.300 and 0.390 years, respectively. The coefficient of determination of the proposed method for the actual and estimated bone age was 0.983. Our method visualized important regions on cervical vertebral images to make a prediction using the gradient-weighted regression activation map technique. Conclusion: Results showed that our proposed method can estimate bone age by lateral cephalogram with sufficiently high accuracy.

Aggressive central odontogenic fibroma in the maxilla: A case report.

A central odontogenic fibroma is a rare benign tumor composed of mature fibrous connective tissue with variable amounts of odontogenic epithelium. It appears at similar rates in the maxilla and mandible. In the maxilla, it usually occurs anterior to the molars. Radiographically, central odontogenic fibroma commonly presents as a multilocular or unilocular radiolucency with a distinct border. This paper reports a case of an aggressive central odontogenic fibroma involving the right posterior maxilla of a 53-year-old man. Radiographs showed an extensive soft tissue mass involving the entire right maxilla with frank bone resorption. The patient had a history of 2 operations in the region, both more than 2 decades ago. Although it was impossible to confirm the previous diagnoses, it was presumed that this case was a recurrent lesion.

Analysis of the root position and angulation of maxillary premolars in alveolar bone using cone-beam computed tomography.

Purpose: This study investigated whether the relationship between the maxillary sinus and the root of the maxillary premolar is correlated with the root position and whether there is a difference in the long axis angle of premolars and the buccal bone thickness according to the sinus-root relationship and root position. Materials and Methods: Cone-beam computed tomographic images of 587 maxillary first premolars and 580 second premolars from 303 patients were retrospectively reviewed. The maxillary sinus floor-root relationship was classified into 4 types, and the root position in the alveolar bone was evaluated as buccal, middle, or palatal. The long axis angle of the maxillary premolars in the alveolar bone and the buccal bone thickness were measured. The correlation between these parameters was analyzed. Results: The maxillary sinus floor-root relationship showed a statistically significant correlation with the root position in the alveolar bone. Most maxillary first premolars were buccally located, and more than half of the second premolars had their roots in the middle. The long axis angle of the premolars was significantly larger in buccal-positioned teeth than in middle-positioned teeth, and the buccal bone was thinner. Conclusion: When the root of the maxillary premolar was separated from the sinus floor, the premolar was often located on the buccal side. Most of the maxillary first premolars had a thinner buccal bone and larger inclination than the second premolars. It is recommended to evaluate the root position, sagittal angle and buccal bone thickness using CBCT for implant treatment planning.

Automated Mesiodens Classification System Using Deep Learning on Panoramic Radiographs of Children.

In this study, we aimed to develop and evaluate the performance of deep-learning models that automatically classify mesiodens in primary or mixed dentition panoramic radiographs. Panoramic radiographs of 550 patients with mesiodens and 550 patients without mesiodens were used. Primary or mixed dentition patients were included. SqueezeNet, ResNet-18, ResNet-101, and Inception-ResNet-V2 were each used to create deep-learning models. The accuracy, precision, recall, and F1 score of ResNet-101 and Inception-ResNet-V2 were higher than 90%. SqueezeNet exhibited relatively inferior results. In addition, we attempted to visualize the models using a class activation map. In images with mesiodens, the deep-learning models focused on the actual locations of the mesiodens in many cases. Deep-learning technologies may help clinicians with insufficient clinical experience in more accurate and faster diagnosis.

Tomographic similarity scan with a computed modified absolute mandibular midsagittal plane for precise and objective localization of mandibular asymmetry.

The application of 3D imaging is at its cusp in craniofacial diagnosis and treatment planning. However, most applications are limited to simple subjective superimposition-based analysis. As the diagnostic accuracy dictates the precision in operability, we propose a novel method that enables objective clinical decision making for patients with mandibular asymmetry. We analyzed cone-beam computed tomography (CBCT) scans of 34 patients who underwent surgical correction for mandibular asymmetry using a high-throughput computing algorithm. Radiomic segmentation of quantitative features of surface and volume followed by exploration resulted in identification of a computed modified absolute mandibular midsagittal plane (cmAMP). Tomographic similarity scan (ToSS) curves were generated via bilateral equidistant scanning in an antero-posterior direction with cmAMP as the reference. ToSS comprised of a comprehensive similarity index (SI) score curve and a segment-wise volume curve. The SI score was computed using the Sørensen-Dice similarity coefficient ranging from 0 to 1. The volumetric analysis was represented as the non-overlapping volume (NOV) and overlapping volume (OV) for each segment, with two segmentation lines, at the mental foramen anteriorly and the intraoral vertical ramus osteotomy region posteriorly. Statistical analysis showed strong negative correlation between the NOV and SI scores for the anterior, middle, and total mandible (P < 0.001). Additionally, a significant correlation was observed between the change in the SI scores for anterior (P = 0.044) and middle segments (P < 0.001) to the total mandible when comparing the data before and after the surgery. This work demonstrated the potential of incorporating ToSS curves in surgical simulation software to improve precision in the clinical decision-making process.

Transfer Learning via Deep Neural Networks for Implant Fixture System Classification Using Periapical Radiographs.

In the absence of accurate medical records, it is critical to correctly classify implant fixture systems using periapical radiographs to provide accurate diagnoses and treatments to patients or to respond to complications. The purpose of this study was to evaluate whether deep neural networks can identify four different types of implants on intraoral radiographs. In this study, images of 801 patients who underwent periapical radiographs between 2005 and 2019 at Yonsei University Dental Hospital were used. Images containing the following four types of implants were selected: Brånemark Mk TiUnite, Dentium Implantium, Straumann Bone Level, and Straumann Tissue Level. SqueezeNet, GoogLeNet, ResNet-18, MobileNet-v2, and ResNet-50 were tested to determine the optimal pre-trained network architecture. The accuracy, precision, recall, and F1 score were calculated for each network using a confusion matrix. All five models showed a test accuracy exceeding 90%. SqueezeNet and MobileNet-v2, which are small networks with less than four million parameters, showed an accuracy of approximately 96% and 97%, respectively. The results of this study confirmed that convolutional neural networks can classify the four implant fixtures with high accuracy even with a relatively small network and a small number of images. This may solve the inconveniences associated with unnecessary treatments and medical expenses caused by lack of knowledge about the exact type of implant.

Automatic mandibular canal detection using a deep convolutional neural network.

The practicability of deep learning techniques has been demonstrated by their successful implementation in varied fields, including diagnostic imaging for clinicians. In accordance with the increasing demands in the healthcare industry, techniques for automatic prediction and detection are being widely researched. Particularly in dentistry, for various reasons, automated mandibular canal detection has become highly desirable. The positioning of the inferior alveolar nerve (IAN), which is one of the major structures in the mandible, is crucial to prevent nerve injury during surgical procedures. However, automatic segmentation using Cone beam computed tomography (CBCT) poses certain difficulties, such as the complex appearance of the human skull, limited number of datasets, unclear edges, and noisy images. Using work-in-progress automation software, experiments were conducted with models based on 2D SegNet, 2D and 3D U-Nets as preliminary research for a dental segmentation automation tool. The 2D U-Net with adjacent images demonstrates higher global accuracy of 0.82 than naïve U-Net variants. The 2D SegNet showed the second highest global accuracy of 0.96, and the 3D U-Net showed the best global accuracy of 0.99. The automated canal detection system through deep learning will contribute significantly to efficient treatment planning and to reducing patients' discomfort by a dentist. This study will be a preliminary report and an opportunity to explore the application of deep learning to other dental fields.

Comparison of panoramic radiography and cone-beam computed tomography for assessing radiographic signs indicating root protrusion into the maxillary sinus.

PURPOSE: This study investigated correlations between findings on panoramic radiographs and cone-beam computed tomography (CBCT) to assess the relationship between the maxillary sinus floor and the roots of maxillary posterior teeth. In addition, radiographic signs indicating actual root protrusion into the maxillary sinus were evaluated on panoramic radiographs. MATERIALS AND METHODS: Paired panoramic radiographs and CBCT images from 305 subjects were analyzed. This analysis classified 2,440 maxillary premolars and molars according to their relationship with the maxillary sinus floor on panoramic radiographs and CBCT images. In addition, interruption of the sinus floor was examined on panoramic radiographs. RESULTS: Root protrusion into the maxillary sinus occurred most frequently in the mesiobuccal roots of the second molars. The classification according to panoramic radiographs and CBCT images was the same in more than 90% of cases when there was no contact between the root apex and the sinus floor. When the panoramic radiograph showed root protrusion into the sinus, the CBCT images showed the same classification in 67.5% of second molars, 48.8% of first molars, and 53.3% of second premolars. There was a statistically significant relationship between interruption of the sinus floor on panoramic radiographs and root protrusion into the sinus on CBCT images. CONCLUSION: The presence of root protrusion into the sinus on panoramic radiographs demonstrated a moderate ability to predict root protrusion into the maxillary sinus. Interruption of the maxillary sinus floor could be considered an indicator of actual root protrusion into the maxillary sinus.

Very deep super-resolution for efficient cone-beam computed tomographic image restoration.

PURPOSE: As cone-beam computed tomography (CBCT) has become the most widely used 3-dimensional (3D) imaging modality in the dental field, storage space and costs for large-capacity data have become an important issue. Therefore, if 3D data can be stored at a clinically acceptable compression rate, the burden in terms of storage space and cost can be reduced and data can be managed more efficiently. In this study, a deep learning network for super-resolution was tested to restore compressed virtual CBCT images. MATERIALS AND METHODS: Virtual CBCT image data were created with a publicly available online dataset (CQ500) of multidetector computed tomography images using CBCT reconstruction software (TIGRE). A very deep super-resolution (VDSR) network was trained to restore high-resolution virtual CBCT images from the low-resolution virtual CBCT images. RESULTS: The images reconstructed by VDSR showed better image quality than bicubic interpolation in restored images at various scale ratios. The highest scale ratio with clinically acceptable reconstruction accuracy using VDSR was 2.1. CONCLUSION: VDSR showed promising restoration accuracy in this study. In the future, it will be necessary to experiment with new deep learning algorithms and large-scale data for clinical application of this technology.

Ameloblastic carcinoma of the mandible: A case report.

Ameloblastic carcinoma is a rare odontogenic malignant tumor with the histologic features of both ameloblastoma and carcinoma. It occurs more frequently in the mandible than in the maxilla and it may appear de novo or develop from a preexisting ameloblastoma or odontogenic cyst. Rapidly progressing, painful swelling is the most common symptom, and radiographically, it shows significant bone resorption and cortical perforation. This report described a case of ameloblastic carcinoma in a 45-year-old man who presented with swelling in the left mandible. The lesion showed combined features of benign findings, such as an expansile cortex with a distinct border, and malignant findings, such as a large exophytic mass with frank bone resorption. Excisional biopsy was performed and a final diagnosis of ameloblastic carcinoma was made.

Langerhans cell histiocytosis of the mandible: two case reports and literature review.

Langerhans cell histiocytosis (LCH) is a rare disorder characterized by the proliferation of dendritic cells resulting in local or systemic symptoms. The clinical symptoms of patients with Langerhans cell histiocytosis depend on the site and the degree of involvement. This article describes two case histories of unifocal bony Langerhans cell histiocytosis with mandibular involvement and further discusses the appropriate management of such via a review of the literature.

Similarity index for intuitive assessment of three-dimensional facial asymmetry.

Evaluation of facial asymmetry generally involves landmark-based analyses that cannot intuitively assess differences in three-dimensional (3D) stereoscopic structures between deviation and non-deviation sides. This study tested a newly developed similarity index that uses a mirroring technique to intuitively evaluate 3D mandibular asymmetry, and characterised the resulting lower facial soft tissue asymmetry. The similarity index was used to evaluate asymmetry before and after surgery in 46 adult patients (27 men, 19 women; age, 22 ± 4.8 years) with skeletal Class III malocclusion and facial asymmetry who underwent conventional bimaxillary orthognathic surgery. Relative to the midsagittal plane used as the reference plane, the non-overlapping volume of the mandible significantly decreased, and the similarity index significantly increased after surgery. Similarity indexes of the mandible and lower facial soft tissue were strongly negatively correlated with non-overlapping volumes of each measurement. Differences in bilateral hemi-mandibular and hemi-lower facial soft tissue surface and volume measurements before surgery were significantly negatively correlated with similarity indexes of the mandible before and after surgery. This newly developed similarity index and non-overlapping volume using a mirroring technique can easily and intuitively evaluate overall 3D morphological discrepancies, especially 3D mandibular asymmetry, before and after surgery in skeletal Class III patients with facial asymmetry.

An overview of deep learning in the field of dentistry.

PURPOSE: Artificial intelligence (AI), represented by deep learning, can be used for real-life problems and is applied across all sectors of society including medical and dental field. The purpose of this study is to review articles about deep learning that were applied to the field of oral and maxillofacial radiology. MATERIALS AND METHODS: A systematic review was performed using Pubmed, Scopus, and IEEE explore databases to identify articles using deep learning in English literature. The variables from 25 articles included network architecture, number of training data, evaluation result, pros and cons, study object and imaging modality. RESULTS: Convolutional Neural network (CNN) was used as a main network component. The number of published paper and training datasets tended to increase, dealing with various field of dentistry. CONCLUSION: Dental public datasets need to be constructed and data standardization is necessary for clinical application of deep learning in dental field.

Maxillary antroliths detected by cone-beam computed tomography in an adult dental population.

PURPOSE: This study evaluated the prevalence and characteristics of maxillary antroliths using cone-beam computed tomography (CBCT) scans performed for maxillofacial diagnostic purposes. MATERIALS AND METHODS: CBCT scans of 13,946 patients over the age of 20 were reviewed for maxillary antroliths, and prevalence according to sex, age, and the side of the jaw was calculated. The relationships of single or multiple antroliths with sex, side, and the degree of sinus inflammation were evaluated. The shape and dimension of antroliths were also assessed. The data were analyzed using descriptive statistics, the chi-square or Fisher exact tests, and Kendall's tau-b. RESULTS: A total of 138 (0.99%) of the 13,946 patients showed an antrolith in at least 1 sinus. Only 18 patients presented a bilateral manifestation, which brought the total number of sinuses containing an antrolith to 156 (0.56%). Multiple antroliths were observed in 36 sinuses, and the total number of antroliths was 207: 110 punctate, 65 linear, and 32 amorphous. The antrolith dimensions varied from 1 mm2 to 91 mm2 (average, 10.2±15.5 mm2). No statistically significant differences were found according to sex, side, and age group (P>0.05). However, there was a statistically significant difference between the multiplicity of antrolith and the degree of sinus inflammation (P<0.05). CONCLUSION: Cone-beam computed tomography is an effective modality for the detection of incidental antroliths. Maxillary antroliths were found to be very rare and were usually asymptomatic. Dentists should have a comprehensive understanding of their diagnosis and treatment in light of possible associated dental problems.

Location and shape of the mandibular lingula: Comparison of skeletal class I and class III patients using panoramic radiography and cone-beam computed tomography.

PURPOSE: The aim of this study was to compare the location and the shape of the mandibular lingula in skeletal class I and III patients using panoramic radiography and cone-beam computed tomography. MATERIALS AND METHODS: The sample group included 190 skeletal class I patients and 157 class III patients. The location of the lingula in relation to the deepest point of the coronoid notch was classified into 3 types using panoramic radiographs. The shapes of the lingulae were classified into nodular, triangular, truncated, or assimilated types using cone-beam computed tomographic images. The data were analyzed using the chi-square test. RESULTS: The tips of the lingulae were at the same level as the coronoid notch in 75.3% of skeletal class I patients and above the coronoid notch in 66.6% of class III patients. The positions of the lingulae in relation to the deepest point of the coronoid notch showed statistically significant differences between class I and class III patients. The most common shape was nodular, and the least common was the assimilated shape. Although this trend was not statistically significant, the triangular shape was more frequently observed in class III patients than in class I patients. CONCLUSION: The locations and the shapes of the mandibular lingulae were variable. Most of the lingulae were at the same level as the coronoid notch in skeletal class I patients and above the coronoid notch in skeletal class III patients. The nodular and assimilated-shaped lingulae were the most and the least prevalent, respectively.

Automatic analysis algorithm for acquiring standard dental and mandibular shape data using cone-beam computed tomography.

This study aims to introduce a new algorithm developed using retrospective cone-beam computed tomography (CBCT) data to obtain a standard dental and mandibular arch shape automatically for an optimal panoramic focal trough. A custom-made program was developed to analyze each arch shape of randomly collected 30 CBCT images. First, volumetric data of the mandible were binarized and projected in the axial direction to obtain 2-dimensional arch images. Second, 30 patients' mandibular arches were superimposed on the center of the bilateral distal contact points of the mandibular canines to generate an average arch shape. Third, the center and boundary of a panoramic focal trough were obtained using smoothing splines. As a result, the minimum thickness and transition of the focal trough could be obtained. If this new algorithm is applied to big data of retrospective CBCT images, standard focal troughs could be established by race, sex, and age group, which would improve the image quality of dental panoramic radiography.

Widely disseminated sporadic Burkitt lymphoma initially presented as oral manifestations in a 6-year-old boy.

Burkitt lymphoma, a subtype of non-Hodgkin's lymphoma, is an aggressive neoplasm with three variants that are endemic, sporadic, and immunodeficiency associated. We present an unusual case of sporadic Burkitt lymphoma in a 6-year-old boy who initially presented with hypermobile teeth and no other specific signs or symptoms. On dental radiography, the patient was found to have alveolar bone resorption adjacent to the maxillary first molars, with the appearance of floating teeth. In addition, magnetic resonance imaging (MRI) showed extensive soft tissue masses involving four quadrants of the jaws. A definitive diagnosis of Burkitt lymphoma was made based on tissue and bone marrow biopsy. Subsequent images, including abdominal computed tomography (CT) and bone scan, revealed wide dissemination of the lymphoma into the abdominal cavity, pancreas, and numerous bones. This case suggests the possibility of dental complaints as an initial clinical manifestation of sporadic Burkitt lymphoma and emphasizes the role of dentists in early detection of the disease to improve prognosis.

Acetylshikonin suppresses invasion of Porphyromonas gingivalis­infected YD10B oral cancer cells by modulating the interleukin-8/matrix metalloproteinase axis.

The development of pharmaceutical agents possessing anti­invasive and anti­metastatic abilities, as well as apoptotic activity, is important in decreasing the incidence and recurrence of oral cancer. Cancer cells are known to acquire invasiveness not only through epigenetic changes, but also from inflammatory stimuli within the tumor microenvironment. Accordingly, the identification of agents that can suppress the inflammation­promoted invasiveness of cancer cells may be important in treating cancer and improving the prognosis of patients with cancer. Acetylshikonin, a flavonoid with anti­inflammatory activity, inhibits proliferation and induces apoptosis of oral cancer cells. In the present study, the anti­invasive effect of acetylshikonin on YD10B oral cancer cells infected with Porphyromonas gingivalis, a major pathogen of chronic periodontitis, and the mechanisms involved were investigated. Firstly, we examined whether P. gingivalis infection increased the invasiveness of YD10B cells. Results suggested that YD10B oral cancer cells become more aggressive when they are infected with P. gingivalis. Secondly, acetylshikonin significantly inhibited the invasion of P. gingivalis­infected YD10B cells by suppressing IL­8 release and IL­8­dependent MMP release. These data suggest that acetylshikonin may be a useful preventive and therapeutic candidate for oral cancer that is chronically infected with periodontal pathogens.

Analysis of the root position of the maxillary incisors in the alveolar bone using cone-beam computed tomography.

PURPOSE: The purpose of this study was to measure the buccal bone thickness and angulation of the maxillary incisors and to analyze the correlation between these parameters and the root position in the alveolar bone using cone-beam computed tomography (CBCT). MATERIALS AND METHODS: CBCT images of 398 maxillary central and lateral incisors from 199 patients were retrospectively reviewed. The root position in the alveolar bone was classified as buccal, middle, or palatal, and the buccal type was further classified into subtypes I, II, and III. In addition, the buccolingual inclination of the tooth and buccal bone thickness were evaluated. RESULTS: A majority of the maxillary incisors were positioned more buccally within the alveolar bone, and only 2 lateral incisors (0.5%) were positioned more palatally. The angulation of buccal subtype III was the greatest and that of the middle type was the lowest. Most of the maxillary incisors exhibited a thin facial bone wall, and the lateral incisors had a significantly thinner buccal bone than the central incisors. The buccal bone of buccal subtypes II and III was significantly thinner than that of buccal subtype I. CONCLUSION: A majority of the maxillary incisor roots were positioned close to the buccal cortical plate and had a thin buccal bone wall. Significant relationships were observed between the root position in the alveolar bone, the angulation of the tooth in the alveolar bone, and buccal bone thickness. CBCT analyses of the buccal bone and sagittal root position are recommended for the selection of the appropriate treatment approach.