Sex estimation from maxillofacial radiographs using a deep learning approach.

The purpose of this study was to construct deep learning models for more efficient and reliable sex estimation. Two deep learning models, VGG16 and DenseNet-121, were used in this retrospective study. In total, 600 lateral cephalograms were analyzed. A saliency map was generated by gradient-weighted class activation mapping for each output. The two deep learning models achieved high values in each performance metric according to accuracy, sensitivity (recall), precision, F1 score, and areas under the receiver operating characteristic curve. Both models showed substantial differences in the positions indicated in saliency maps for male and female images. The positions in saliency maps also differed between VGG16 and DenseNet-121, regardless of sex. This analysis of our proposed system suggested that sex estimation from lateral cephalograms can be achieved with high accuracy using deep learning.

Entire Dental Arch Mesial Movement after Extraction of Maxillary Right Central Incisor due to Root Fracture Treated with Temporary Anchorage Devices.

This case report describes a 19-year-old woman with skeletal Class I crowding and an unsalvageable maxillary right central incisor. She visited our clinic with the chief complaint of mobility of the maxillary right central incisor due to a traffic accident. After extraction of the maxillary right central incisor, the space was closed orthodontically. All the maxillary right teeth were moved mesially with an elastic chain attached to a palatal lever arm which was connected to palatal temporary anchorage devices (TADs). After orthodontic treatment had been completed, the maxillary right lateral incisor and peg-shaped left lateral incisor were restored with a porcelain laminate veneer. The maxillary right canine was morphologically reshaped and built up with composite resin. Consequently, esthetically ideal occlusion and functional lateral guidance with uncontacted molars were obtained. These results show that mesial movement of the entire dental arch with TADs is a useful orthodontic treatment option in patients in whom the maxillary central incisor has been extracted.

The Physiological Function of Deglutition and Swallowing in Patients With Maxillary Protrusion After Mandibular Advancement.

The objective of this study was to determine the tongue-palatal contact changes in patients with skeletal maxillary protrusion after sagittal split ramus osteotomy (SSRO) during swallowing. In this study, 15 patients with maxillary protrusion and 10 normal subjects participated. Before and 3 months after surgery, tongue-palatal contact patterns during swallowing of patients with maxillary protrusion as well as controls were evaluated by electropalatography. The electrode contact number in the alveolar, palatal, and velar parts was examined. The swallowing duration of each phase was also evaluated. In the lateral area of the velar part, incomplete electrode contact was shown at 0.3 seconds in patients with maxillary protrusion. The electrode contact number in the velar part at 0.3 seconds before tongue-palatal complete contact was significantly less in the preoperative patients compared with the controls ( P < 0.05). A small increase in the electrode contact number of the velar part was shown in the postoperative patients at 0.3 and 0.2 seconds before tongue-palatal complete contact ( P < 0.05). The pharyngeal phase duration was significantly larger in the patients with maxillary protrusion before SSRO compared with the controls ( P < 0.05). After SSRO, the pharyngeal phase duration was significantly shortened. It was shown that the tongue-palatal contact pattern during swallowing in patients with maxillary protrusion improved after orthognathic surgery, and the pharyngeal phase duration was also shortened. It is suggested that the changes in the mesiodistal mandibular position by orthognathic surgery can improve tongue posture and movement during swallowing.

Image preprocessing with contrast-limited adaptive histogram equalization improves the segmentation performance of deep learning for the articular disk of the temporomandibular joint on magnetic resonance images.

OBJECTIVES: The objective was to evaluate the robustness of deep learning (DL)-based encoder-decoder convolutional neural networks (ED-CNNs) for segmenting temporomandibular joint (TMJ) articular disks using data sets acquired from 2 different 3.0-T magnetic resonance imaging (MRI) scanners using original images and images subjected to contrast-limited adaptive histogram equalization (CLAHE). STUDY DESIGN: In total, 536 MR images from 49 individuals were examined. An expert orthodontist identified and manually segmented the disks in all images, which were then reviewed by another expert orthodontist and 2 expert oral and maxillofacial radiologists. These images were used to evaluate a DL-based semantic segmentation approach using an ED-CNN. Original and preprocessed CLAHE images were used to train and validate the models whose performances were compared. RESULTS: Original and CLAHE images acquired on 1 scanner had pixel values that were significantly darker and with lower contrast. The values of 3 metrics-the Dice similarity coefficient, sensitivity, and positive predictive value-were low when the original MR images were used for model training and validation. However, these metrics significantly improved when images were preprocessed with CLAHE. CONCLUSIONS: The robustness of the ED-CNN model trained on a dataset obtained from a single device is low but can be improved with CLAHE preprocessing. The proposed system provides promising results for a DL-based, fully automated segmentation method for TMJ articular disks on MRI.

An in vitro study of digital impressions and three-dimensional printed models of orbital defects using mobile devices and monoscopic photogrammetry.

PURPOSE: The purpose of this study was to perform an in vitro evaluation of digital impressions using a mobile device and monoscopic photogrammetry in cases of orbital defects with undercuts. METHODS: Three 10-mm-square cubes were attached to a diagnostic cast of a patient with a right orbital defect. Still images acquired with a mobile device were used to generate facial three-dimensional (3D) data. Two types of still images were used: one was a whole face image, and the other was a defect site-focused image. For comparison, an extraoral scanner was used to obtain facial 3D data. Five dental technicians fabricated 3D printed models using additive manufacturing and measured the distances between the measurement points using a digital caliper. The discrepancy between the distances measured on the diagnostic cast of the patient and the 3D printed model was calculated. Friedman test was used to analyze the discrepancy, and the Bonferroni test was used to verify the differences between the pairs. RESULTS: Statistical significance was found with respect to the type of 3D model fabrication method. CONCLUSION: Within the limitations of this in vitro study, the results suggested that the workflow can be applied to digital impressions of the maxillofacial region.

Analysis of the feasibility of using deep learning for multiclass classification of dental anomalies on panoramic radiographs.

The aim of the feasibility study was to construct deep learning models for the classification of multiple dental anomalies in panoramic radiographs. Panoramic radiographs with single supernumerary teeth and/or odontomas were considered the "case" group; panoramic radiographs with no dental anomalies were considered the "control" group. The dataset comprised 150 panoramic radiographs: 50 each of no dental anomalies, single supernumerary teeth, and odontomas. To classify the panoramic radiographs into case and control categories, we employed AlexNet, which is a convolutional neural network model. AlexNet was able to classify whole panoramic radiographs into two or three classes, according to the presence or absence of supernumerary teeth or odontomas. The performance metrics of the three-class classification were 70%, 70.8%, 70%, and 69.7% for accuracy, precision, sensitivity, and F1 score, respectively, in the macro average. These results support the feasibility of using deep learning to detect multiple dental anomalies in panoramic radiographs.

Lansoprazole and zoledronate delays hard tissue healing of tooth extraction sockets in dexamethasone-treated mice.

Proton pump inhibitors (PPIs) are among the most commonly prescribed medicines for the management of acid-related gastrointestinal diseases. Osteonecrosis of the jaw (ONJ) is a serious adverse event that is associated with the use of antiresorptive and antiangiogenic agents. According to previous clinical reports, the use of PPIs contributes to the pathogenesis of severe ONJ that requires surgery. Here, we investigated the effects of lansoprazole (LP) or LP in combination with zoledronate (ZOL) on ONJ development in mice. C57BL/6J mice were administered ZOL (125 µg/kg intravenously, twice weekly) and/or LP (10 mg/kg intraperitoneally; 3 weeks of 3 consecutive days followed by 1 day off). One week after initiation of the study, the first molar was atraumatically extracted. Concurrently with ZOL administration, dexamethasone (Dex) was administered (5 mg/kg intraperitoneally, twice weekly). Micro-computed tomography and histological evaluation were performed to characterize femoral structures, tooth extraction sockets, and osteonecrosis areas. The results showed that ZOL/Dex significantly increased bone mass compared to saline/Dex, while the simultaneous administration of LP and ZOL/Dex diminished the ZOL-induced enhancement of bone mass. In the alveolar bone around the tooth extraction socket, necrotic bone was significantly increased in the LP/Dex group compared to the saline/Dex group. However, no signs of more severe ONJ-like lesions were observed following combined administration of LP and ZOL/Dex, other than an increase in the number of non-attached TRAP-positive cells. Our findings in a mouse model suggest that LP use can be a risk factor for the development of ONJ.

Automated segmentation of articular disc of the temporomandibular joint on magnetic resonance images using deep learning.

Temporomandibular disorders are typically accompanied by a number of clinical manifestations that involve pain and dysfunction of the masticatory muscles and temporomandibular joint. The most important subgroup of articular abnormalities in patients with temporomandibular disorders includes patients with different forms of articular disc displacement and deformation. Here, we propose a fully automated articular disc detection and segmentation system to support the diagnosis of temporomandibular disorder on magnetic resonance imaging. This system uses deep learning-based semantic segmentation approaches. The study included a total of 217 magnetic resonance images from 10 patients with anterior displacement of the articular disc and 10 healthy control subjects with normal articular discs. These images were used to evaluate three deep learning-based semantic segmentation approaches: our proposed convolutional neural network encoder-decoder named 3DiscNet (Detection for Displaced articular DISC using convolutional neural NETwork), U-Net, and SegNet-Basic. Of the three algorithms, 3DiscNet and SegNet-Basic showed comparably good metrics (Dice coefficient, sensitivity, and positive predictive value). This study provides a proof-of-concept for a fully automated deep learning-based segmentation methodology for articular discs on magnetic resonance images, and obtained promising initial results, indicating that the method could potentially be used in clinical practice for the assessment of temporomandibular disorders.

Detecting the presence of supernumerary teeth during the early mixed dentition stage using deep learning algorithms: A pilot study.

BACKGROUND: Supernumerary teeth are a common anomaly and are frequently observed in paediatric patients. To prevent or minimize complications, early diagnosis and treatment is ideal in children with supernumerary teeth. AIM: This study aimed to apply convolutional neural network (CNN)-based deep learning to detect the presence of supernumerary teeth in children during the early mixed dentition stage. DESIGN: Three CNN models, AlexNet, VGG16-TL, and InceptionV3-TL, were employed in this study. A total of 220 panoramic radiographs (from children aged 6 years 0 months to 9 years 6 months) including supernumerary teeth (cases, n = 120) or no anomalies (controls, n = 100) were retrospectively analyzed. The CNN performances were assessed according to accuracy, sensitivity, specificity, receiver operating characteristic (ROC) curves, and area under the ROC curves for a cross-validation test dataset. RESULTS: The VGG16-TL model had the highest performance according to accuracy, sensitivity, specificity, and area under the ROC curve, but the other models had similar performance. CONCLUSION: CNN-based deep learning is a promising approach for detecting the presence of supernumerary teeth during the early mixed dentition stage.

Occlusal Trauma and Bisphosphonate-Related Osteonecrosis of the Jaw in Mice.

Osteonecrosis of the jaw (ONJ) is a serious adverse event that is associated with antiresorptive agents, and it manifests as bone exposure in the maxillofacial region. Previous clinical reports suggest that mechanical trauma would trigger ONJ in a manner that is similar to tooth extractions. To the best of our knowledge, there have been few detailed pathophysiological investigations of the mechanisms by which occlusal/mechanical trauma influences ONJ. Here, we developed a novel mouse model that exhibits ONJ following experimental hyperocclusion and nitrogen-containing bisphosphonate (N-BP) treatment. This in vivo model exhibited ONJ in alveolar bone, particularly in the mandible. Moreover, the experimental hyperocclusion induced remarkable alveolar bone resorption in both mouse mandible and maxilla, whereas N-BP treatment completely prevented alveolar bone resorption. In this study, we also modeled trauma by exposing clumps of mesenchymal stem cells (MSCs)/extracellular matrix complex to hydrostatic pressure in combination with N-BP. Hydrostatic pressure loading induced lactate dehydrogenase (LDH) release by calcified cell clumps that were differentiated from MSCs; this LDH release was enhanced by N-BP priming. These in vivo and in vitro models may contribute further insights into the effect of excessive mechanical loading on ONJ onset in patients with occlusal trauma.

Landmark annotation and mandibular lateral deviation analysis of posteroanterior cephalograms using a convolutional neural network.

BACKGROUND/PURPOSE: Facial asymmetry is relatively common in the general population. Here, we propose a fully automated annotation system that supports analysis of mandibular deviation and detection of facial asymmetry in posteroanterior (PA) cephalograms by means of a deep learning-based convolutional neural network (CNN) algorithm. MATERIALS AND METHODS: In this retrospective study, 400 PA cephalograms were collected from the medical records of patients aged 4 years 2 months-80 years 3 months (mean age, 17 years 10 months; 255 female patients and 145 male patients). A deep CNN with two optimizers and a random forest algorithm were trained using 320 PA cephalograms; in these images, four PA landmarks were independently identified and manually annotated by two orthodontists. RESULTS: The CNN algorithms had a high coefficient of determination (R 2 ), compared with the random forest algorithm (CNN-stochastic gradient descent, R 2  = 0.715; CNN-Adam, R 2  = 0.700; random forest, R 2  = 0.486). Analysis of the best and worst performances of the algorithms for each landmark demonstrated that the right latero-orbital landmark was most difficult to detect accurately by using the CNN. Based on the annotated landmarks, reference lines were defined using an algorithm coded in Python. The CNN and random forest algorithms exhibited similar accuracy for the distance between the menton and vertical reference line. CONCLUSION: Our findings imply that the proposed deep CNN algorithm for detection of facial asymmetry may enable prompt assessment and reduce the effort involved in orthodontic diagnosis.

In Vivo Efficacy of Lacticaseibacillus rhamnosus L8020 in a Mouse Model of Oral Candidiasis.

Oral candidiasis presents with multiple clinical manifestations. Among known pathogenic Candida species, Candida albicans is the most virulent and acts as the main causative fungus of oral candidiasis. Novel treatment modalities are needed because of emergent drug resistance and frequent candidiasis recurrence. Here, we evaluated the ability of Lacticaseibacillus rhamnosus L8020, isolated from healthy and caries-free volunteers, to prevent against the onset of oral candidiasis in a mouse model. Mice were infected with C. albicans, in the presence or absence of L. rhamnosus L8020. The mice were treated with antibiotics and corticosteroid to disrupt the oral microbiota and induce immunosuppression. We demonstrated that oral consumption of L. rhamnosus L8020 by C. albicans-infected mice abolished the pseudomembranous region of the mouse tongue; it also suppressed changes in the expression levels of pattern recognition receptor and chemokine genes. Our results suggest that L. rhamnosus L8020 has protective or therapeutic potential against oral candidiasis, which supports the potential use of this probiotic strain for oral health management.

Cytotoxicity assay using a human pluripotent stem cell-derived cranial neural crest cell model.

Cleft lip and palate are the most common congenital abnormalities that occur early in pregnancy. The majority of cranial mesenchyme is derived from cranial neural crest cells that differentiate into odontoblasts, cartilage, craniofacial bone, and connective tissue. A subset of these cells differentiates into cranial ganglia. We have previously reported an induction protocol of cranial neural crest cell-like cells from human pluripotent stem cells. This study tested detection of the cytotoxic sensitivities of dental materials, including titanium ions, palladium ions, and hydroxyethyl methacrylate, on the cell viability of induced cranial neural crest cell-like cells (iNC-LCs) derived from Tic human induced pluripotent stem cell (hiPSC) line. Further, the sensitivity was compared with those of human fetal lung fibroblastic cell line MRC-5, which is origin of Tic hiPSC, and osteoblastic cell line MC3T3-E1 which was derived from mouse calvaria. The results suggested that this cell-based assay system using iNC-LCs is a potential method for in vitro screening as an alternative to animal testing to predict toxic effects of dental materials on early craniofacial development.

Comprehensive analysis of transcriptional profiles in oral epithelial-like cells stimulated with oral probiotic Lactobacillus spp.

OBJECTIVE: The mechanisms of action of probiotics can vary among species and among strains of a single species; thus, they can affect host cells in a complex manner. In the present study, Lactobacillus spp. were evaluated for their ability to adhere to gingival epithelial-like cells. Comprehensive analyses of transcriptional profiles of mouse gingival epithelial GE1 cells treated with L. rhamnosus L8020 were performed to assess the putative in vivo probiotic potential of this strain. METHODS: Five Lactobacillus spp., isolated from the oral cavity, traditional Bulgarian yoghurt, and the feces of a healthy human, were each co-cultured with GE1 cells. Adhesion assays with serial dilution plating and DNA microarray analysis were performed to identify differentially expressed genes (DEGs) in GE1 cells grown in co-culture with L. rhamnosus L8020. RESULTS: The oral isolates L. rhamnosus L8020, L. casei YU3, and L. paracasei YU4 demonstrated significantly greater adhesion compared with the non-oral isolates. In total, 536 genes in GE1 cells exhibited more than twofold upregulation or downregulation, compared with the 0 h timepoint, during co-culture with L. rhamnosus L8020. Gene ontology enrichment analysis revealed that DEGs were differentially enriched in a time-dependent manner. Early responses involved widespread changes in gene expression. CONCLUSIONS: This study reveals changes in expression of genes involved in the epithelial physical barrier and immune response in gingival epithelial-like cells co-cultured with L. rhamnosus L8020. Further investigations regarding the molecular mechanisms by which L. rhamnosus L8020 serves as a probiotic may provide evidence to support clinical use.

Accuracy Evaluation of a Three-Dimensional Model Generated from Patient-Specific Monocular Video Data for Maxillofacial Prosthetic Rehabilitation: A Pilot Study.

PURPOSE: To evaluate if the combination of a monoscopic photogrammetry technique and smartphone-recorded monocular video data could be appropriately applied to maxillofacial prosthesis fabrication. MATERIALS AND METHODS: Smartphone video and laser scanning data were recorded for five healthy volunteers (24.1 ± 0.7 years). Three-dimensional (3D) facial models were generated using photogrammetry software and a laser scanner. Smartphone-recorded video data were used to generate a photogrammetric 3D model. The videos were recorded at two resolutions: 1080 × 1920 (high resolution) and 720 × 1280 pixels (low resolution). The lengths of five nasal component parts (nose height, nasal dorsum length, nasal column length, nasal ala length, and nose breadth) were compared in the photogrammetric 3D models (as the test model) and the laser scanned 3D models (as the validation model) using reverse engineering software. RESULTS: There was a significant difference in the nasal dorsum length between the test model and the validation model (high resolution; 95% confidence interval, 2.05-5.07, Low resolution; confidence interval, 2.19-5.69). In contrast to the nasal dorsum length, there were no significant differences in nose height, nose breadth, nasal ala length, and nasal column length. CONCLUSION: Using smartphone-recorded video data and a photogrammetry technique may be a promising technique to apply in the maxillofacial prosthetic rehabilitation workflow.

Applying deep artificial neural network approach to maxillofacial prostheses coloration.

PURPOSE: Maxillofacial prosthetic rehabilitation replaces missing structures to recover the function and aesthetics relating to facial defects or injuries. Deep learning is rapidly expanding with respect to applications in medical fields. In this study, we apply the artificial neural network (ANN)-based deep learning approach to coloration support for fabricating maxillofacial prostheses. METHODS: We compared two machine learning algorithms, ANN-based deep learning and the random forest algorithm, to determine the compounding amount of pigment. We prepared 52 silicone elastomer specimens of varying colors and measured the CIE 1976 L* a* b* color space information using a spectrophotometer on the input dataset. The output of these algorithms indicated the compounding amount of four pigments. According to the algorithms' pigment compounding predictions, we prepared the specimens for validation analysis and measured the CIE 1976 L* a* b* values. We determined the color differences between the real skin color of five research participants (22.3 ± 1.7 years) and that of the silicone elastomer specimens fabricated based on the algorithm predictions using the CIEDE00 ΔE00 color system. RESULTS: The color differences (ΔE00 value) between the real skin color and silicone elastomer validation specimens were 3.45 ± 0.87 (ANN) and 5.54 ± 1.41 (random forest), which indicates that the deep ANN approach produced superior results with respect to the ΔE00 value compared with the random forest algorithm. CONCLUSIONS: These results suggest that applying deep ANN is a promising technique for the coloration of maxillofacial prostheses.

Effect of microslit retention on the bond strength of zirconia to dental materials.

The aim of this study was to investigate the effect of microslits formed by Nd:YVO4 laser beam machining on the bond strength between two types of zirconia, yttria-partially stabilized zirconia (Y-TZP) and ceria-partially stabilized zirconia/alumina nanocomposite (Ce-TZP/A), and porcelain or two types of resin. Zirconia disks were divided into three groups: 1) non-treated (NT); 2) blasted with alumina particles (AB); 3) microslits fabricated on a zirconia surface by laser beam machining (MS). After veneering porcelain or resins on zirconia specimens, halves of the resin specimens were thermocycled up to 20,000 cycles. The shear bond strength between porcelain and both types of zirconia was not improved by the microslits. Before and after thermocycling, the bond strength between an indirect composite resin or acrylic resin and Y-TZP with microslits was the highest. It was concluded that the microslits on Y-TZP enabled micromechanical interlocking and improved the bond strength and durability of the resins.

Fatigue resistance of yttria-stabilized tetragonal zirconia polycrystal clasps for removable partial dentures.

With the rapid development of computer-aided design/computer-aided manufacturing (CAD/CAM) systems, the application of zirconia in removable partial dentures is expected to expand. Clasps composed of zirconia should improve esthetics without inducing the risk of metal allergy. The aim of this study was to examine the fatigue resistance of yttria-stabilized tetragonal zirconia polycrystal (Y-TZP) clasps for removable partial dentures. Yttria-stabilized tetragonal zirconia polycrystal and cobalt-chromium (Co-Cr) alloy were prepared using CAD/CAM systems. Specimens were either of the semicircular type or of the flat type, with cross-sectional areas of taper ratios of 0.50, 0.75, and 1.00. All specimens were tested using the cantilever test and the constant displacement fatigue test, and data were analyzed using ANOVA. During the cantilever test, the maximum displacement prior to fracture was greater than the required undercut, and the semicircular-type specimen exhibited a higher fracture load than the flat type. None of the specimens displayed permanent deformation and showed almost the same degree of deformation after fatigue testing. A lower taper ratio was associated with lower average load values and greater displacement. Within the limitations of this study, it was possible to conclude that Y-TZP provides the required undercut and adequate retentive force for removable partial denture clasps. Additionally, Y-TZP and Co-Cr alloy had almost the same degree of deformation even after the simulated lifespan of removable partial dentures.

Accuracy of a three-dimensional dentition model digitized from an interocclusal record using a non-contact surface scanner.

OBJECTIVES: For orthodontic treatment, it is important to assess the dental morphology, as well as the position and inclination of teeth. The aim of this article was to develop an efficient and accurate method for the three-dimensional (3D) imaging of the maxillary and mandibular dental morphology by measuring interocclusal records using an optical scanner. MATERIALS AND METHODS: The occlusal and incisal morphology of participants was registered in the intercuspal position using a hydrophilic vinyl polysiloxane and digitized into 3D models using an optical scanner. Impressions were made of the maxilla and mandible in alginate materials in order to fabricate plaster models and created into 3D models using the optical scanner based on the principal triangulation method. The occlusal and incisal areas of the interocclusal records were retained. The buccal and lingual areas were added to these regions entirely by the 3D model of the plaster model. The accuracy of this method was evaluated for each tooth, with the dental cast 3D models used as controls. RESULTS: The 3D model created from the interocclusal record and the plaster model of the dental morphology was analysed in 3D software. The difference between the controls and the 3D models digitized from the interocclusal records was 0.068±0.048mm, demonstrating the accuracy of this method. LIMITATIONS: The presence of severe crowding may compromise the ability to separate each tooth and digitize the dental morphology. CONCLUSIONS: The digitization method in this study provides sufficient accuracy to visualize the dental morphology, as well as the position and inclination of these teeth.

Construction of orthodontic setup models on a computer.

INTRODUCTION: Orthodontic setup models are usually limited to the display of teeth, with no information about the roots. The purpose of this article is to present a method for visualizing the tooth roots in setup models by integrating information from cone-beam computed tomography and a laser scanner. The reproducibility of the integration was evaluated. METHODS: The records of 5 patients were used in this study. Three-dimensional digital models were generated from the dental casts. Tooth models were generated from the cone-beam computed tomography slices. The 3-dimensional models were superimposed on the crowns of the teeth in the tooth models and integrated. The integrated 3-dimensional tooth model and 3-dimensional setup model were registered. The reproducibility of the integration was evaluated for each tooth. Unpaired Student t tests were performed on the data between the anterior and posterior teeth, and between the right and left teeth. RESULTS: The discrepancy among the integrated 3-dimensional models at the final positions after we used this technique was 0.025 ± 0.007 mm. There was a significant difference in the distance between the anterior and posterior teeth (P <0.05). However, the average distances between the anterior and posterior teeth were small: 0.023 ± 0.007 and 0.028 ± 0.007 mm, respectively. No significant difference was found between the right and left teeth (P = 0.831). CONCLUSIONS: The methods presented in this study provide a reproducible visualization of tooth roots in virtual setup models by registering accurate crown models to cone-beam computed tomography scans.