Advancements in diagnosing oral potentially malignant disorders: leveraging Vision transformers for multi-class detection.

OBJECTIVES: Diagnosing oral potentially malignant disorders (OPMD) is critical to prevent oral cancer. This study aims to automatically detect and classify the most common pre-malignant oral lesions, such as leukoplakia and oral lichen planus (OLP), and distinguish them from oral squamous cell carcinomas (OSCC) and healthy oral mucosa on clinical photographs using vision transformers. METHODS: 4,161 photographs of healthy mucosa, leukoplakia, OLP, and OSCC were included. Findings were annotated pixel-wise and reviewed by three clinicians. The photographs were divided into 3,337 for training and validation and 824 for testing. The training and validation images were further divided into five folds with stratification. A Mask R-CNN with a Swin Transformer was trained five times with cross-validation, and the held-out test split was used to evaluate the model performance. The precision, F1-score, sensitivity, specificity, and accuracy were calculated. The area under the receiver operating characteristics curve (AUC) and the confusion matrix of the most effective model were presented. RESULTS: The detection of OSCC with the employed model yielded an F1 of 0.852 and AUC of 0.974. The detection of OLP had an F1 of 0.825 and AUC of 0.948. For leukoplakia the F1 was 0.796 and the AUC was 0.938. CONCLUSIONS: OSCC were effectively detected with the employed model, whereas the detection of OLP and leukoplakia was moderately effective. CLINICAL RELEVANCE: Oral cancer is often detected in advanced stages. The demonstrated technology may support the detection and observation of OPMD to lower the disease burden and identify malignant oral cavity lesions earlier.

Combining public datasets for automated tooth assessment in panoramic radiographs.

OBJECTIVE: Panoramic radiographs (PRs) provide a comprehensive view of the oral and maxillofacial region and are used routinely to assess dental and osseous pathologies. Artificial intelligence (AI) can be used to improve the diagnostic accuracy of PRs compared to bitewings and periapical radiographs. This study aimed to evaluate the advantages and challenges of using publicly available datasets in dental AI research, focusing on solving the novel task of predicting tooth segmentations, FDI numbers, and tooth diagnoses, simultaneously. MATERIALS AND METHODS: Datasets from the OdontoAI platform (tooth instance segmentations) and the DENTEX challenge (tooth bounding boxes with associated diagnoses) were combined to develop a two-stage AI model. The first stage implemented tooth instance segmentation with FDI numbering and extracted regions of interest around each tooth segmentation, whereafter the second stage implemented multi-label classification to detect dental caries, impacted teeth, and periapical lesions in PRs. The performance of the automated tooth segmentation algorithm was evaluated using a free-response receiver-operating-characteristics (FROC) curve and mean average precision (mAP) metrics. The diagnostic accuracy of detection and classification of dental pathology was evaluated with ROC curves and F1 and AUC metrics. RESULTS: The two-stage AI model achieved high accuracy in tooth segmentations with a FROC score of 0.988 and a mAP of 0.848. High accuracy was also achieved in the diagnostic classification of impacted teeth (F1 = 0.901, AUC = 0.996), whereas moderate accuracy was achieved in the diagnostic classification of deep caries (F1 = 0.683, AUC = 0.960), early caries (F1 = 0.662, AUC = 0.881), and periapical lesions (F1 = 0.603, AUC = 0.974). The model's performance correlated positively with the quality of annotations in the used public datasets. Selected samples from the DENTEX dataset revealed cases of missing (false-negative) and incorrect (false-positive) diagnoses, which negatively influenced the performance of the AI model. CONCLUSIONS: The use and pooling of public datasets in dental AI research can significantly accelerate the development of new AI models and enable fast exploration of novel tasks. However, standardized quality assurance is essential before using the datasets to ensure reliable outcomes and limit potential biases.

Enhancing intraoral scanner accuracy using scan aid for multiple implants in the edentulous arch: An in vivo study.

OBJECTIVES: Intraoral scans of multiple implants in the edentulous arch are challenged by the absence of a distinct surface morphology between scan bodies. A scan aid was applied in such situation and evaluated for intraoral scanning accuracy in vivo. MATERIALS AND METHODS: 87 implants in 22 patients were scanned with scan aid (SA) and without scan aid (NO) using two different intraoral scanners (CS3600 [CS] and TRIOS3 [TR]). Master casts were digitized by a laboratory scanner. Virtual models were superimposed using an inspection software and Linear deviation and precision were measured. Statistical analysis was performed using linear mixed models (α = .05). RESULTS: Total mean linear deviation within the CS group was 189 µm without scan aid and 135 µm when using the scan aid. The TR group's total mean deviation was 165 µm with and without a scan aid. Significant improvement with scan aid was observed for the CS group (p = .001), and no difference was found in the TR group. 96% of scan bodies were successfully scanned in the TR-SA group compared to 86% for the TR-NO group, 83% for the CS-SA, and 70% for the CS-NO group, respectively. CONCLUSIONS: The evaluated scan aid improved linear deviation compared to unsplinted scans for the CS group but not for the TR group. These differences could originate from different scanning technologies used, active triangulation (CS) and confocal microscopy (TR). The scan aid improved the ability to recognize scan bodies successfully with both systems, which could have a favorable clinical impact overall.

Intra-oral scan segmentation using deep learning.

OBJECTIVE: Intra-oral scans and gypsum cast scans (OS) are widely used in orthodontics, prosthetics, implantology, and orthognathic surgery to plan patient-specific treatments, which require teeth segmentations with high accuracy and resolution. Manual teeth segmentation, the gold standard up until now, is time-consuming, tedious, and observer-dependent. This study aims to develop an automated teeth segmentation and labeling system using deep learning. MATERIAL AND METHODS: As a reference, 1750 OS were manually segmented and labeled. A deep-learning approach based on PointCNN and 3D U-net in combination with a rule-based heuristic algorithm and a combinatorial search algorithm was trained and validated on 1400 OS. Subsequently, the trained algorithm was applied to a test set consisting of 350 OS. The intersection over union (IoU), as a measure of accuracy, was calculated to quantify the degree of similarity between the annotated ground truth and the model predictions. RESULTS: The model achieved accurate teeth segmentations with a mean IoU score of 0.915. The FDI labels of the teeth were predicted with a mean accuracy of 0.894. The optical inspection showed excellent position agreements between the automatically and manually segmented teeth components. Minor flaws were mostly seen at the edges. CONCLUSION: The proposed method forms a promising foundation for time-effective and observer-independent teeth segmentation and labeling on intra-oral scans. CLINICAL SIGNIFICANCE: Deep learning may assist clinicians in virtual treatment planning in orthodontics, prosthetics, implantology, and orthognathic surgery. The impact of using such models in clinical practice should be explored.

In vitro scanning accuracy using different aids for multiple implants in the edentulous arch.

OBJECTIVES: Optical impressions of implants in the edentulous arch are challenged by the absence of distinct surface morphology between multiple implant scan bodies. The purpose of this in vitro study was to evaluate a newly developed scan aid in various designs and colors to improve the accuracy of multi-implant scans. MATERIALS AND METHODS: A universal scan aid in three different designs (circular, square, irregular) and three different materials (colors: beige, gray, white) was applied to the implant scan bodies of a master model of six implants in an edentulous maxilla. Ten scans using an intra-oral scanner of each scan aid were acquired. Reference scanning was performed using a desktop optical scanner. Alignment of scans was performed at the first scan body in a three-dimensional modeling and inspection software and deviations for trueness and precision were calculated using the signed nearest-neighbor method and then statistically analyzed (α = .05). RESULTS: Overall, the beige irregular scan aid had the highest trueness and showed significant differences compared to unsplinted scans. The precision showed more heterogenous results and decreased when using the scan aid. Ease of use was observed with the irregular gray scan aid due to its increased elongation at break. CONCLUSIONS: The evaluated scan aid led to improved trueness when compared to unsplinted scans. Even though the irregular design in beige color showed highest trueness, the poor fracture strength of the tested material requires further improvement.

Accuracy of intraoral scans: An in vivo study of different scanning devices.

STATEMENT OF PROBLEM: The accuracy of intraoral scanners is a prerequisite for the fabrication of dental restorations in computer-aided design and computer-aided manufacturing (CAD-CAM) dentistry. While the precision of intraoral scanners has been investigated in vitro, clinical data on the accuracy of intraoral scanning (IOS) are limited. PURPOSE: The purpose of this clinical study was to determine the accuracy of intraoral scanning with different devices compared with extraoral scanning. MATERIAL AND METHODS: An experimental appliance was fabricated for 11 participants and then scanned intraorally and extraorally with 3 different intraoral scanners and a reference scanner. Intraoral and extraoral scans were subdivided into complete-arch and short-span scans and compared with the reference scan to assess trueness. Repeated scans in each group were assessed for precision. RESULTS: Precision and trueness were higher for extraoral scans compared with intraoral scans, except for complete-arch scans with 1 intraoral scanner. The median precision of short-span scans was higher (extraoral: 22 to 29 µm, intraoral: 23 to 43 µm) compared with complete-arch scans (extraoral: 81 to 165 µm, intraoral: 80 to 198 µm). The median trueness of short-span scans (extraoral: 28 to 40 µm, intraoral: 38 to 47 µm) was higher than that of complete-arch scans (extraoral: 118 to 581 µm, intraoral: 147 to 433 µm) for intraoral and extraoral scanning. CONCLUSIONS: Intraoral conditions negatively influenced the accuracy of the scanning devices, which was also reduced for the complete-arch scans.

Digital implantology-a review of virtual planning software for guided implant surgery. Part II: Prosthetic set-up and virtual implant planning.

BACKGROUND: Patient- and technology-related parameters influence the successful implementation of virtual implant planning and guided implant surgery. Besides data processing and computer aided design of drill guides as described in Part I, the possibilities and limitations for prosthetic set-up and virtual implant planning are essential (Part II). METHODS: The following software systems were examined using two different clinical situations for implant therapy: coDiagnostiX™, DentalWings, Canada (CDX); Simplant Pro™, Dentsply, Sweden (SIM); Smop™, Swissmeda, Switzerland (SMP); NobelClinician™, Nobel Biocare, Switzerland (NC); Implant Studio, 3Shape, Denmark (IST). Assessment criteria geared towards interfaces and integrated tools for prosthetic set-up and virtual implant planning. RESULTS: A software interface for an individual virtual prosthetic set-up was provided by two systems (CDX, IST), whereas the set-up of standardized teeth was provided by four systems (CDX, SIM, SMP, IST). Alternatively, a conventional set-up could be scanned and imported. One system could solely work with the digitization of a conventional set-up for virtual implant planning (NC). Stock abutments could be displayed for implant planning, but none of the tested software systems provided tools for the design of an individual abutment. All systems displayed three-dimensional reconstructions or two-dimensional cross-sections with varying orientation for virtual implant placement. The inferior alveolar nerve could be marked to respect a minimum distance between the nerve and the planned implant. Three implant planning systems provided a library to display more than 50 implant systems (CDX, SIM, IST), one system provided 33 implant systems (SMP) and one implant system provided 4 implant systems (NC). CONCLUSION: Depending on the used software system, there are limited options for a virtual set-up, virtual articulators and the display of a virtual prosthetic set-up. The implant systems used by the clinician is important for the decision which software system to choose, as there is a discrepancy between available implant systems and the number of supported systems in each software.

Gender- and age-related differences in the width of attached gingiva and clinical crown length in anterior teeth.

BACKGROUND: The anatomical features of the gingiva and the clinical crowns and their interrelation, especially in aesthetically and functionally demanding areas, are important in complex dental or implant-retained prosthetic rehabilitations. This observational cross-sectional study was designed to evaluate gender- and age-related differences in the width of attached gingiva (WAG), the clinical crown length (CCL), and their interrelation in the anterior teeth to determine the relationship between the pink and white aesthetics. METHODS: Eighty (54 females, 26 males) fully dentate Caucasian participants between the ages of 20 and 25 years and 36 probands (23 females, 13 males) between the ages of 45 and 55 years were included in the present study. The CCL of the maxillary and mandibular canines, as well as the central incisors of the maxilla and mandible, were determined with a dental sliding caliper measuring from the middle margin of the gingiva at its deepest point to the incisal edge. The clinical investigation of the WAG was performed by inserting a periodontal probe into the gingival sulcus in the middle of the buccal surface to firstly measure the probing pocket depth. The distance between the gingival margin and mucogingival junction (MGJ) was then measured with a Beerendonk sliding caliper in the middle of the labial curvature, and the clinical WAG was determined by subtraction of the measured probing depth. For the statistical analysis, the Mann-Whitney U test, the Wilcoxon-Test, the Spearman's rank correlation, and a two-factorial non-parametric analysis were used. RESULTS: There was no correlation between the CCL and the WAG in a healthy periodontium. Gender influenced the CCL, with men having significantly longer teeth than women in both maxilla (P ≤ 0.01) and mandible (P ≤ 0.05). Age did not influence the CCL significantly neither in the upper (P = 0.06) nor in the lower jaw (P = 0.33). Gender did not show to have a significant influence on the WAG of maxilla (P = 0.69) and mandible (P = 0.26). But differences in the WAG between young and old participants were observed in both upper (P ≤ 0.01) and lower jaw (P ≤ 0.05). CONCLUSION: The present observational study demonstrated that the mean values of cohorts with mixed age groups and genders should not be considered when attempting to determine the ideal relationships between the pink and white aesthetics since the statistical analyses showed significant differences between different age groups and genders.

Fully guided implant surgery using Magnetic Resonance Imaging - An in vitro study on accuracy in human mandibles.

OBJECTIVES: The objective of this in vitro study was to assess the accuracy of fully guided implant placement following virtual implant planning based on MRI. MATERIAL AND METHODS: Sixteen human cadaver hemimandibles with single missing teeth (n = 3), partially edentulous (n = 6) and edentulous situations (n = 7) were imaged using MRI. MRI and optical scans obtained with an intraoral scanner, were imported into an implant planning software. Virtual prosthetic and implant planning were performed regarding hard- and soft-tissue anatomy. Drill guides were manufactured, and fully guided implant placement was performed. Buccal and lingual bone and implant nerve distance were measured by three examiners in preoperative MRI and postoperative CBCT. The implant position was assessed using a software for deviation of implant positions displayed in CBCT and optical scans, respectively. RESULTS: MRI displayed relevant structures for implant planning such as cortical and cancellous bone, inferior alveolar nerve and neighboring teeth. Implant planning, CAD/CAM of drill guides and guided implant placement were performed. Deviations between planned and actual implant positions in postoperative CBCT and optical scans were 1.34 mm (SD 0.84 mm) and 1.03 mm (SD 0.46 mm) at implant shoulder; 1.41 mm (SD 0.88 mm) and 1.28 mm (SD 0.52 mm) at implant apex, and 4.84° (SD 3.18°) and 4.21° (SD 2.01°). Measurements in preoperative MRI and postoperative CBCT confirmed the compliance with minimum distances of implants to anatomical structures. CONCLUSIONS: Relevant anatomical structures for imaging diagnostics in implant dentistry are displayed with MRI. The accuracy of MRI-based fully guided implant placement in vitro is comparable to the workflow using CBCT.

Horizontal bone grafting using equine-derived cancellous bone blocks is associated with severe complications: A prospective clinical and histological pilot study.

AIMS: The aim of this prospective, clinical study was to evaluate the clinical performance and histological outcome of a new equine hydroxyapatite collagenated bone block (eHAC) for horizontal bone grafting prior to implant placement. MATERIALS AND METHODS: Five patients (two male/three female) with a mean age of 51.6 years (range 22-66 years) and a reduced horizontal bone width of the alveolar ridge (mean 3.5 mm) underwent horizontal bone grafting using eHAC at 10 grafting sites. Reentry was performed 6.9 months after the horizontal grafting procedure. Clinical follow-up (mean 28.9 month) considered width gain of the alveolar ridge, soft tissue healing, and complications. To evaluate graft incorporation, four additional patients underwent histological assessment of equine blocks adjacent to autologous blocks 3 and 6 months after grafting. RESULTS: The study was terminated after graft failure was observed in four of five patients. Mean horizontal bone width had increased by 3.6 ± 1.22 mm. Three out of nine implants placed had to be removed due to graft failure. Histological evaluation revealed large amounts of soft connective tissue within the grafts (mean 67.3 ± 9.5%). The proportion of new bone formation 3 months after the lateral grafting procedure revealed an average of 8.6%, compared to 11.4% after 6 to 7 months. CONCLUSION: Lateral ridge grafting using eHAC achieved measurable horizontal width gain but revealed high rates of severe complications. CLINICAL IMPLICATIONS: Within the limitations of this study, eHAC bone blocks cannot be recommended for horizontal bone grafting.

Virtual implant planning and fully guided implant surgery using magnetic resonance imaging-Proof of principle.

OBJECTIVES: To present a workflow of virtual implant planning and guided implant surgery with magnetic resonance imaging (MRI) and virtual dental models without the use of ionizing radiation. METHODS: Five patients scheduled for implant placement underwent an MR examination at three Tesla using individualized 2D and 3D turbo spin-echo (TSE) sequences and dedicated head coils. The MRI data and virtual dental models derived from either optical model scans or intraoral scans were imported to a virtual implant planning software (coDiagnostiX, Dental Wings, Montreal, Canada). Virtual prosthetic planning and implant planning were performed regarding the hard and soft tissue anatomy. A drill guide was designed on the virtual dental model using computer-aided design (CAD) and manufactured in-house, using a 3D printer (Eden 260V, Stratasys, Eden Prairie, MN, USA). RESULTS: The MRI displayed all relevant anatomical structures for dental implant planning such as cortical and cancellous bone, floor of the nasal and maxillary sinus, inferior alveolar nerve and neighboring teeth. The manual alignment of virtual dental models with the MRI was possible using anatomical landmarks. Dental implant planning, CAD/CAM of a drill guide and fully guided implant placement were successfully performed. CONCLUSIONS: Guided implant surgery is feasible with MRI without ionizing radiation. Further studies will have to be conducted to study the accuracy of the presented protocol and compare it to the current workflow of guided surgery using CBCT.

A review of virtual planning software for guided implant surgery - data import and visualization, drill guide design and manufacturing.

BACKGROUND: Virtual implant planning systems integrate (cone beam-) computed tomography data to assess bone quantity and virtual models for the design of the implant-retained prosthesis and drill guides. Five commercially available systems for virtual implant planning were examined regarding the modalities of integration of radiographic data, virtual dental models and the design of drill guides for guided implant surgery. The purpose of this review was to describe the limitations of these available systems regarding the import of imaging data and the design and fabrication of a drill guide. METHODS: The following software systems were examined regarding the import of imaging data and the export of the virtual implant planning for the design and fabrication of a drill guide with the help of two clinical situations requiring dental implant therapy: coDiagnostiX™, DentalWings, Canada (CDX); Simplant Pro™, Dentsply, Sweden (SIM); Smop™, Swissmeda, Switzerland (SMP); NobelClinician™, Nobel Biocare, Switzerland (NC); Implant Studio, 3Shape, Denmark (IST). Assessment criteria included data formats and management as well as the workflow for the design and production of drill guides. RESULTS: All systems have a DICOM-interface ("Digital Imaging and Communication in Medicine") for the import of radiographic data. Imaging artefacts could be reduced but not eliminated by manual data processing. The import of virtual dental models in a universal format (STL: Standard Tesselation Language) was possible with three systems; one system could only be used with a proprietary data format. All systems display three-dimensional surface models or two-dimensional cross-sections with varying orientation for virtual implant planning. Computer aided design and manufacturing (CAD/CAM) of drill guides may be performed by the user with the help of default parameters or solely by the provider of the software and thus without the influence of the clinician. CONCLUSION: Data bases of commonly used implant systems are available in all tested software, however not all systems allow to plan and execute fully guided implant placement. An individual design and in-house manufacturing of the drill guide is only available in some software systems. However, at the time of publication most recent software versions showed flexibility in individual design and in-house manufacturing of drill guides.

The accuracy of computer-guided implant surgery with tooth-supported, digitally designed drill guides based on CBCT and intraoral scanning. A prospective cohort study.

OBJECTIVES: The purpose of this prospective cohort study was to evaluate computer-guided implant surgery with tooth-supported drill guides based on CBCT scans and intraoral scanning. MATERIALS AND METHODS: For partially edentulous patients, a prosthetic and surgical planning was completed in the guided surgery software (coDiagnostiX) and drill guides were 3D-printed accordingly. Three months after implant placement, an intraoral scan of the implant's position was used to evaluate the accuracy of placement using the coDiagnostiX treatment evaluation tool. Deviations were reported in degrees and in distance at implant's entry point and apex. Several risk factors, which might influence the accuracy, were evaluated separately: treated jaw, flap design, prior augmentations, amount of unrestored teeth, crowding, location of implants, cortical interference, and implant's length and diameter. RESULTS: A total of 66 patients received 145 Straumann tissue level implants that were eligible for accuracy analysis. The mean angular deviation was 2.72° ± 1.42. The mean three-dimensional deviation at the implant's entry point was 0.75 mm ± 0.34. At implant's apex, the mean was 1.06 mm ± 0.44. The amount of unrestored teeth (p = .002 & p = .003), the implant's location (p < .001), the implant's length (p = .004), and cortical interference (p = .033) had a significant influence on the accuracy of placement. Implant survival was 99.3% (n = 1 failed implant) at 12 and 24 months. CONCLUSIONS: Guided surgery with tooth-supported drill guides made in a digital workflow is a feasible treatment option. However, deviations do occur and the implant's length, location, cortical interference and the amount of unrestored teeth have a significant influence on the accuracy.

The accuracy of different dental impression techniques for implant-supported dental prostheses: A systematic review and meta-analysis.

AIM: This systematic review and meta-analysis were conducted to assess and compare the accuracy of conventional and digital implant impressions. The review was registered on the PROSPERO register (registration number: CRD42016050730). MATERIAL AND METHODS: A systematic literature search was conducted adhering to PRISMA guidelines to identify studies on implant impressions published between 2012 and 2017. Experimental and clinical studies at all levels of evidence published in peer-reviewed journals were included, excluding expert opinions. Data extraction was performed along defined parameters for studied specimens, digital and conventional impression specifications and outcome assessment. RESULTS: Seventy-nine studies were included for the systematic review, thereof 77 experimental studies, one RCT and one retrospective study. The study setting was in vitro for most of the included studies (75 studies) and in vivo for four studies. Accuracy of conventional impressions was examined in 59 studies, whereas digital impressions were examined in 11 studies. Nine studies compared the accuracy of conventional and digital implant impressions. Reported measurements for the accuracy include the following: (a) linear and angular deviations between reference models and test models fabricated with each impression technique; (b) three-dimensional deviations between impression posts and scan bodies respectively; and (c) fit of implant-supported frameworks, assessed by measuring marginal discrepancy along implant abutments.) Meta-analysis was performed of 62 studies. The results of conventional and digital implant impressions exhibited high values for heterogeneity. CONCLUSIONS: The available data for accuracy of digital and conventional implant impressions have a low evidence level and do not include sufficient data on in vivo application to derive clinical recommendations.

Group 5 ITI Consensus Report: Digital technologies.

OBJECTIVES: Working Group 5 was assigned the task to review the current knowledge in the area of digital technologies. Focused questions on accuracy of linear measurements when using CBCT, digital vs. conventional implant planning, using digital vs. conventional impressions and assessing the accuracy of static computer-aided implant surgery (s-CAIS) and patient-related outcome measurements when using s-CAIS were addressed. MATERIALS AND METHODS: The literature was systematically searched, and in total, 232 articles were selected and critically reviewed following PRISMA guidelines. Four systematic reviews were produced in the four subject areas and amply discussed in the group. After emendation, they were presented to the plenary where after further modification, they were accepted. RESULTS: Static computer-aided surgery (s-CAIS), in terms of pain & discomfort, economics and intraoperative complications, is beneficial compared with conventional implant surgery. When using s-CAIS in partially edentulous cases, a higher level of accuracy can be achieved when compared to fully edentulous cases. When using an intraoral scanner in edentulous cases, the results are dependent on the protocol that has been followed. The accuracy of measurements on CBCT scans is software dependent. CONCLUSIONS: Because the precision intraoral scans and of measurements on CBCT scans and is not high enough to allow for the required accuracy, s-CAIS should be considered as an additional tool for comprehensive diagnosis, treatment planning, and surgical procedures. Flapless s-CAIS can lead to implant placement outside of the zone of keratinized mucosa and thus must be executed with utmost care.

Registration of cone beam computed tomography data and intraoral surface scans - A prerequisite for guided implant surgery with CAD/CAM drilling guides.

OBJECTIVES: Guided implant surgery (GIS) is performed with drilling guides that are produced on the virtual tooth model using CAD/CAM technology. The prerequisite for this workflow is the alignment of patients cone beam computed tomography CBCT and surface scan (registration). Dental restorations may cause deteriorating imaging artifacts in CBCT data, which in turn can have an impact on the registration process. The influence of the user and the preprocessing of data and of image artifacts on the registration accuracy were examined. MATERIAL AND METHODS: CBCT data and intraoral surface scans of 36 patients were used for virtual implant planning in coDiagnostiX (Dentalwings, Montreal, Canada). CBCT data were reconstructed to a three-dimensional anatomical model with the default settings provided by the software and also manually by four different examiners. Subsequently, the CBCT and intraoral surface models were registered by each examiner with the help of anatomical landmarks. Patients' data were subdivided into four groups (A-D) according to the number of metallic restorations: A = 0-2 restorations, B = 3-5 restorations, C = 6-8 restorations and D > 8 restorations. After registration, the distances between CBCT and dental surface models were measured. Linear regression models were used to assess the influence of the segmentation, the examiner and to the number of restorations (P < 0.05). RESULTS: The deviations between surface scan and CBCT models accounted to 0.54 mm (mean). The mean deviations were 0.69 mm (max. 24.8 mm) and 0.4 mm (max. 9.1 mm) for default and manual segmentation, respectively. Mean deviations of 0.36 mm (Group A), 0.43 mm (Group B), 0.67 mm (Group C) and 1.01 mm (Group D) were recorded. The segmentation (P = 0.000), the user (P = 0.0052) and the number of restorations (P = 0.0337) had a significant influence on the registration accuracy. CONCLUSIONS: The deviation between CBCT and surface scan model resulting from inaccurate registration is transferred to the surgical field and results in a deviation between the planned and actual implant position. The registration accuracy in commercial virtual implant planning software is significantly influenced by the preprocessing of imported data, by the user and by the number of restorations resulting in clinically non-acceptable deviations encoded in drilling guides.

Magnetic resonance imaging of intraoral hard and soft tissues using an intraoral coil and FLASH sequences.

OBJECTIVES: To ascertain the feasibility of MRI as a non-ionizing protocol for routine dentomaxillofacial diagnostic imaging. Wireless coils were used for MRI of intraoral hard and soft tissues. METHODS: FLASH MRI was applied in vivo with a mandible voxel size of 250 × 250 × 500 µm3, FOV of 64 × 64 × 28 mm3 and acquisition time of 3:57 min and with a maxilla voxel size of 350 µm3 and FOV of 34 cm3 in 6:40 min. Ex vivo imaging was performed in 4:38 min, with a resolution of 200 µm3 and FOV of 36.5 cm3. Cone beam (CB) CT of the mandible and subjects were acquired. MRI was compared to CBCT and histological sections. Deviations were calculated with intraclass correlation coefficient (ICC) and coefficient of variation (cv). RESULTS: A high congruence between CBCT, MRI and specimens was demonstrated. Hard and soft tissues including dental pulp, periodontium, gingiva, cancellous bone and mandibular canal contents were adequately displayed with MRI. CONCLUSIONS: Imaging of select intraoral tissues was achieved using custom MRI protocols with an easily applicable intraoral coil in a clinically acceptable acquisition time. Comparison with CBCT and histological sections helped demonstrate dimensional accuracy of the MR images. The course of the mandibular canal was accurately displayed with CBCT and MRI. KEY POINTS: • MRI is a clinically available diagnostic tool in dentistry • Intraoral hard and soft tissues can be imaged with a high resolution with MRI • The dimensional accuracy of MRI is comparable to cone beam CT.

2-Dimensional changes of the soft tissue profile of augmented and non-augmented human extraction sockets: a randomized pilot study.

AIM: This study identified the soft tissue changes of the alveolar ridge at different time points within 12 weeks after tooth extraction with and without socket augmentation. MATERIALS AND METHODS: In 38 patients with single tooth extractions, 40 sockets were augmented and 39 extraction sockets were not augmented. At 2, 4, 6, 8 and 12 weeks impressions were taken and casts digitized with a laser scanner. The horizontal and vertical changes were compared between augmented and non-augmented sites. A p-value <0.05 was considered statistically significant. RESULTS: The mean changes of augmented sockets were between 0.4 mm (2 weeks) and 0.8 mm (12 weeks). In non-augmented sockets changes of 0.7 mm (2 weeks) and of 1.0 mm (12 weeks) were demonstrated. The mean values differed significantly between the buccal and oral region (p < 0.01). Overall, there were significant differences of the mean dimensional changes regarding time (p < 0.01) and augmentation (p < 0.01). CONCLUSIONS: Augmented sockets showed less resorption within 4 weeks after extraction compared to non-augmented sockets. Non-augmented sockets showed a continuous dimensional loss with a great variation over 12 weeks whereas augmented sockets had the highest degree of resorption between 4 and 6 weeks. At 12 weeks a comparable resorption in augmented and non-augmented sockets was observed.

A Comparison of Different Methods to Generate Tooth Surface Models Without Applying Ionizing Radiation for Digital 3-Dimensional Image Fusion With Magnetic Resonance Imaging-Based Data of the Head and Neck Region.

OBJECTIVE: The 3-dimensional display of the vocal tract and teeth is necessary in numerous clinical and scientific contexts. Due to the different tissue properties this can only be achieved by combining different imaging techniques. A comparison of methods to record the tooth surface to create combined models of the vocal tract and teeth without applying ionizing radiation is displayed in this study. METHODS: Four methods to record the tooth surface were compared concerning their clinical accuracy after combination with magnetic resonance imaging (MRI)-based vocal tract models. With 2 of the presented methods the information of the tooth surface was obtained from MRI data. With the other 2 methods, the tooth surface was captured using a digital and a conventional impression technique. RESULTS: Impression techniques received the best rating results. CONCLUSIONS: The digital 3-dimensional image fusion of dental impression and MRI resulted in a virtual model of the vocal tract and teeth with a high clinical accuracy without applying ionizing radiation.

Detection of caries around restorations on bitewings using deep learning.

OBJECTIVE: Secondary caries lesions adjacent to restorations, a leading cause of restoration failure, require accurate diagnostic methods to ensure an optimal treatment outcome. Traditional diagnostic strategies rely on visual inspection complemented by radiographs. Recent advancements in artificial intelligence (AI), particularly deep learning, provide potential improvements in caries detection. This study aimed to develop a convolutional neural network (CNN)-based algorithm for detecting primary caries and secondary caries around restorations using bitewings. METHODS: Clinical data from 7 general dental practices in the Netherlands, comprising 425 bitewings of 383 patients, were utilized. The study used the Mask-RCNN architecture, for instance, segmentation, supported by the Swin Transformer backbone. After data augmentation, model training was performed through a ten-fold cross-validation. The diagnostic accuracy of the algorithm was evaluated by calculating the area under the Free-Response Receiver Operating Characteristics curve, sensitivity, precision, and F1 scores. RESULTS: The model achieved areas under FROC curves of 0.806 and 0.804, and F1-scores of 0.689 and 0.719 for primary and secondary caries detection, respectively. CONCLUSION: An accurate CNN-based automated system was developed to detect primary and secondary caries lesions on bitewings, highlighting a significant advancement in automated caries diagnostics. CLINICAL SIGNIFICANCE: An accurate algorithm that integrates the detection of both primary and secondary caries will permit the development of automated systems to aid clinicians in their daily clinical practice.