To develop and evaluate a deep learning-based system for automatic detection of the accessory mental foramen (AMF) using cone-beam computed tomography (CBCT) images, and to compare the detection accuracy and clinical reliability performance of two convolutional neural network (CNN) architectures for this model. A total of 3000 CBCT scans were retrospectively screened. After expert evaluation, 700 CBCT scans exhibiting AMFs were identified. For comparative analysis, 700 CBCT scans with normal mental-foramen anatomy were selected as the matched control group. A custom lightweight CNN and a ResNet-50 model were trained for binary classification of AMF presence. Model performance was evaluated by determining accuracy, precision, recall, and the F1-score. Gradient-weighted class activation mapping (Grad-CAM) visualisation was employed to assess the anatomical relevance of the models' attention maps. Statistical analyses were performed to compare the diagnostic performance of the two networks. The ResNet-50 model achieved superior performance (overall accuracy: 85.8% for ResNet-50 vs 71.1% for the custom CNN). With the ResNet-50 model, anomaly recall improved from 0.68 to 0.88, reducing missed detections by 63%. Grad-CAM analysis demonstrated that the models focused primarily on anatomically valid regions around the MF, confirming the interpretability and clinical relevance of the models. Automatic detection of the AMF using CBCT and deep learning represents a reliable, objective, and efficient diagnostic approach that minimises observer bias and enhances clinical decision-making. Deep learning-based detection of AMFs on CBCT can enhance diagnostic accuracy and reduce the risk of surgical complications by providing consistent, observer-independent evaluations.

To evaluate the effectiveness of spatial superimposition of cone-beam computed tomography (CBCT) scans acquired at baseline and 6-month follow-up for monitoring bone healing in mandibular cystic lesions, aiming to reduce variability in conventional volumetric comparisons and enhance accuracy, reproducibility, and spatial fidelity of radiographic assessments. Two CBCT scans of a single patient (baseline, T0; follow-up, T1) were imported into 3D Slicer v5.8.0. The mandible was isolated by manually cropping a region of interest, and the cystic lesion was segmented semi-automatically using the Grow from Seeds tool. A rigid six-degree-of-freedom registration aligned T1 to T0; the resulting transformation matrix was applied to the T1 segmentation to enable direct voxel-wise comparison. Lesion volumes were measured, and spatial subtraction analysis quantified the resorbed area. Lesion volume decreased from 1841.64 mm3 at T0 to 1362.62 mm3 at T1, corresponding to an absolute reduction of 479.02 mm3 (26.0%). The subtraction mask accurately localized regressed voxels, confirming both the magnitude and spatial distribution of bone healing. Potential limitations related to CBCT artifacts and the need for operator-dependent manual steps should be considered when interpreting the results. Overall, CBCT superimposition with semi-automatic segmentation provides an objective, consistent, and anatomically precise approach for monitoring mandibular cyst regression and may represent a useful tool to support conservative management strategies.

Aligning MRI and CBCT for Advanced TMJ Diagnostics: Case Series Using AI-Powered Registration in Dentistry and Orthodontics.

To quantify external apical root resorption (EARR) on the incisors following non-extraction treatment of Class I malocclusion patients with moderate crowding, comparing clear aligners and fixed appliances using a novel 3D analysis of Cone-Beam Computed Tomography (CBCT) derived surface models. In this randomized clinical trial, 32 adult patients, mean age 22.3 years, mean treatment duration 24.2 months, with Class I malocclusion and moderate crowding (mean Little's Index 4.76 mm) were allocated to either clear aligner (n = 15) or fixed appliance (n = 17) treatment. CBCT scans were obtained before and after treatment. EARR was measured using surface-based analysis of 3D models, and associations with patient and treatment-related factors were tested. The overall median EARR was -0.72 mm, with no significant difference between clear aligners (-0.71 mm) and fixed appliances (-0.72 mm). Upper lateral incisors exhibited significantly greater EARR than lower incisors (p = 0.002) and upper central incisors (p < 0.001). No significant predictor for EARR was found considering age, sex, crowding severity or treatment duration. EARR occurred following non-extraction treatment of Class I malocclusion with both clear aligners and fixed appliances, with no significant difference between appliance types. Upper lateral incisors were most susceptible to EARR. The novel 3D analysis enabled comprehensive quantification of total EARR, setting a new methodological standard. Monitoring root health during treatment is important, particularly for upper lateral incisors.

This study aimed to compare and improve the performance of three deep learning models, i.e., U-Net Transformer (UNETR), Swin UNETR, and 3D UX-Net, for the segmentation of the mandibular canal and its branches. A dataset of 173 cone beam computed tomography (CBCT) scans was used for training, validation, and testing. The mandibular canals and branches were segmented manually and by the three AI models. A postprocessing module based on anatomical characteristics was then applied to improve model performance. Evaluations were conducted using Dice similarity coefficient (DSC), intersection over union (IoU), 95th Percentile Hausdorff Distance (HD95), average symmetric surface distance (ASSD), precision, and recall. All models efficiently segmented the mandibular, incisive, and mental canals, operating at least 25 times faster than manual annotation. Both 3D UX-Net and Swin UNETR consistently outperformed the UNETR network across most metrics, with 3D UX-Net demonstrating a slight performance advantage over Swin UNETR in terms of DSC, IoU, and recall. Furthermore, the anatomically-based postprocessing module significantly improved the metrics for all models. Ultimately, the 3D UX-Net with postprocessing achieved the highest accuracy, with mean values of 0.788 (DSC), 0.652 (IoU), 0.23 mm (HD95), 0.083 mm (ASSD), 72.7% (precision), and 87.0% (recall). 3D UX-Net and Swin UNETR are superior to the UNETR for segmenting small dental structures. Between the two, 3D UX-Net demonstrated statistically significant improvements in overlap and recall. Furthermore, the performances of these models can be significantly enhanced by applying postprocessing strategies based on anatomical characteristics.

The objectives of this study were to assess and compare volumetric and morphological changes in the pharyngeal airway (PA) in growing patients with a retrognathic mandible (skeletal Class II patients), assessed via Cone-Beam Computed Tomography (CBCT), after growth modification treatment performed with removable (Twin Block) and fixed (Herbst) functional appliances or Class II elastics. Pre- and post-treatment CBCT scans were collected from an existing practice dataset of 66 patients following completion of multibracket appliance treatment. Three matched groups of 22 patients were treated with either Class II elastics (CII-E), Herbst or Twin Block (TB) appliances. Total and partial PA volumes were calculated, and morphology was assessed via measurements of cross-sectional area (CSA) and hydraulic diameter (DH). Multiple dental and skeletal parameters were also measured three-dimensionally. Predicted mean changes and pairwise comparisons were computed via linear mixed-effects models. The PA volume increased, and the morphology demonstrated reduced constriction (increased minimum CSA and DH) following treatment in all three groups. However, there were no statistically significant differences between any of the groups for all airway parameters measured. Dental and skeletal changes followed a similar pattern to that observed for the airway with most parameters showing significant changes following treatment in all groups, but few significant differences were seen between the groups. Similar changes were observed in the PA after treatment with CII-E or functional appliances (Herbst and TB), but it is unclear whether these were due to comparable treatment effects, normal growth, or a combination of both.

The aim of this study is to investigate the morphological differences in temporomandibular joint (TMJ) structures according to the disc displacement status using cone-beam computed tomography (CBCT)-based three-dimensional (3D) reconstructed models. Overall, 100 TMJs from 50 adult patients were retrospectively analyzed using magnetic resonance imaging (MRI) and categorized into three groups: normal, anterior disc displacement with reduction (ADcR), and without reduction (ADsR). CBCT data were reconstructed into 3D models using the Mimics software (Materialise, Leuven, Belgium), and anatomical parameters, including articular eminence angle, axial condylar inclination, anteroposterior position of the condyle, and vertical position of the condyle, were measured and statistically compared. The articular eminence angle was highest in normal (33.11 ± 9.26°) and lowest in ADsR (26.86 ± 8.90°), with a significant difference (p = 0.011). The axial condylar inclination and vertical position of the condyle significantly differed between the groups (p < 0.05). These parameters decreased in the following order: Normal, ADcR, and ADsR. No significant differences were observed in the anteroposterior condyle positions. TMJ disc displacement is associated with measurable structural changes in joint morphology. CBCT-based 3D analysis of TMJ structures may serve as a useful adjunct method to MRI for the diagnosis and understanding of internal TMJ derangement.

Compromised bone quality and quantity in the posterior maxilla are widely recognised as significant risk factors for implant failures. This study aimed to (1) assess the bone quality of the resorbed posterior maxilla both radiographically and histologically and (2) evaluate the impact of native alveolar bone quality on osteogenesis following maxillary sinus floor augmentation (MSFA). Patients presenting advanced posterior maxillary atrophy (residual bone height ≤ 4 mm) underwent MSFA via a lateral window approach using deproteinised bovine bone matrix (DBBM). Bone core biopsies were collected during second-stage implant placement for histological analyses. Preoperative cone-beam computer tomography (CBCT) images were used to classify bone quality based on cortical bone configuration. Multiple linear regression was employed to identify factors influencing osteogenesis. A total of 190 sinuses from 176 patients (96 males/80 females, mean age: 45.77 ± 4.13 years) underwent MSFA and biopsy. Radiographic analysis revealed the presence of bicortical bone in 42.21% of the maxillary ridges, unicortical bone in 25.97%, no cortical bone in 19.48% and sinus floor-crestal bone fusion in 12.34%. Histologically, native bone comprised 40.01% ± 14.85% of mineralised trabeculae while newly formed bone accounted for 18.77% ± 5.69% in the DBBM grafted areas. A significant positive correlation was observed between native bone trabecula percentage and new bone formation (r = 0.23, P = .04). No significant associations were found with age, sex, healing time or radiographic alveolar type (P > .05). Greater density of native trabecular bone may contribute to enhanced osteogenesis following MSFA. This study enhances our understanding of the bone characteristics of the resorbed posterior maxilla. It emphasises the need to consider the bone quality of the recipient site before conducting a MSFA for determining the surgical procedure and predicting the prognosis.

The concept of a digital or virtual patient represents a paradigm shift in implant dentistry, transforming treatment planning into a data-driven, prosthetically guided, and patient-specific workflow. This article outlines the comprehensive process of constructing a digital patient by integrating multiple datasets: cone-beam computed tomography (CBCT) for volumetric bone representation, intraoral surface scans for dental and soft tissue geometry, facial scans for esthetic harmony, and jaw-motion tracking data for functional analysis. These datasets are merged through precise registration protocols using fiducial markers and AI-assisted alignment to create an accurate three-dimensional model reflecting both anatomical and functional dynamics. The resulting virtual patient enables prosthetically driven implant positioning, real-time occlusal simulations, and seamless communication between the surgical, prosthetic, and laboratory teams. Moreover, advances in artificial intelligence, cloud computing, and blockchain-based data security promise to expand the digital patient’s role from static documentation to a dynamic, continuously updated virtual twin. This integrated workflow enhances surgical precision, esthetic predictability, and clinical efficiency, marking a significant step toward fully digitized, personalized implant rehabilitation.

Background: Zygomatic implants have become an important treatment option for the rehabilitation of patients with severely atrophic maxillae, offering an alternative to extensive bone grafting procedures. Accurate preoperative planning is essential due to the complex anatomical trajectory and proximity to critical structures such as the maxillary sinus and orbit. Cone-beam computed tomography (CBCT) has gained widespread use in implant dentistry because of its ability to provide three-dimensional visualization of craniofacial structures. However, concerns remain regarding the diagnostic accuracy and potential clinical limitations of CBCT imaging in planning zygomatic implant placement. Objective This systematic review aimed to evaluate the accuracy and clinical pitfalls associated with CBCT imaging in the surgical planning of zygomatic implants and to assess its reliability in guiding complex implant procedures. Methods A systematic review was conducted following PRISMA guidelines. Electronic databases including PubMed, Scopus, Web of Science, and the Cochrane Library were searched to identify relevant studies. Eligible studies included clinical and observational research evaluating the use of CBCT in preoperative planning for zygomatic implants. Non-clinical studies, case reports, and non-English publications were excluded. Study selection was performed through a structured screening process, and data were extracted using standardized forms. Methodological quality and risk of bias were assessed using established evaluation tools. A qualitative synthesis of findings was performed due to methodological variability among studies. Results Eight studies met the inclusion criteria and were included in the final analysis. The findings consistently indicated that CBCT imaging provides accurate three-dimensional assessment of anatomical structures relevant to zygomatic implant placement. Measurement deviations between planned and intraoperative findings were generally minimal, often within clinically acceptable ranges. CBCT was shown to improve visualization of implant trajectory and anatomical landmarks; however, certain limitations such as imaging artifacts, variability in voxel resolution, and interpretive challenges were reported in some studies. Conclusion CBCT imaging appears to be a reliable and valuable tool for planning zygomatic implant placement, offering improved visualization and enhanced surgical planning capabilities. Nevertheless, awareness of potential imaging limitations and careful interpretation remain essential. Further large-scale clinical studies are recommended to strengthen the current evidence and refine imaging protocols for complex implant rehabilitation.

Nasopharynx on Cone-beam Computed Tomography.

A unified computed tomography dose index (CTDI) which includes all scattered radiation was used to investigate the difference in absorbed dose to the patient between a computed tomography (CT) and a cone beam CT (CBCT). The unified CTDI, denoted CTDI300,w, was measured using a 100mm pencil ionization chamber at three positions covering 300mm at each of the five phantom slots in three CTDI body phantoms. The dose length product of a lumbar spine protocol for the modalities was comparable (CT: 153 mGycm, CBCT: 126 mGycm), though the normalized CTDI300,w for the CT was 4.6 times higher than for the CBCT. The comparison between CTDI300,w and the two dose indexes provided by the systems showed a substantial underestimation for both modalities. The most accurate, but also the most inconvenient method to compare absorbed dose is to use the CTDI300,w. Most important is to always use the same dose index.

This investigation aimed to compare perceived lesion detectability and visibility between limited-angle cone-beam computed tomography (CBCT) photon-counting (PC) and flat-panel (FP) mammography at equivalent radiation-dose levels. Projection data were collected by rotating a lesion-containing, humanlike breast phantom and acquiring images at every other degree using both a PC and an FP system, resulting in 5-13 projections over an 8-24° angular range. Feldkamp-Davis-Kress (FDK) and simultaneous iterative reconstruction technique (SIRT) algorithms were used to reconstruct three-dimensional (3D) volumes for each system. A reader study evaluated lesion detectability and visibility in the 3D data. PC-CBCT showed better detectability and visibility of masses compared to FP-CBCT, with 90% sensitivity versus 70% and an area under the visual grading curve (AUCVGC) of 0.66-0.71 (P<.05). FDK image volumes demonstrated improved calcification visibility (P<.05) with PC-CBCT over FP-CBCT; however, for SIRT, calcification visibility was lower (AUCVGC=0.46). This study emphasizes the need to reassess photon-counting in 3D mammography to improve lesion detection and visibility.

To evaluate the effect of miniscrew-assisted rapid palatal expansion (MARPE) on root resorption, buccal bone thickness, and tooth inclination. This study included 30 patients with maxillary constriction who were treated with MARPE. The device was left in place for 6 months for retention. Cone-beam computed tomography (CBCT) images were taken before treatment and after 6 months. CBCT images were analyzed for root length, buccal bone thickness, and tooth inclination for premolars and first molar on both sides. A paired t-test was used to compare between pretreatment and posttreatment root lengths, buccal bone thickness, and buccal tooth inclination. No significant differences were found between pretreatment and posttreatment root lengths of all measured teeth (P value > .05). A significant difference was found in the buccal bone thickness at the right second premolar at the levels of 2 mm and 8 mm (P = .049 and P = .039, respectively) and at the level of 2 mm for the left second premolar (P = .009). Buccal bone thickness at the right and left mesiobuccal root of the first molar at 8 mm was significantly reduced after expansion (P = .001). The difference between pre-expansion and postexpansion in tooth inclination ranged between -1.87 and 0.27. Only the inclination of the left first molar was significantly increased after expansion. No root resorption was caused by the bone-borne MARPE. No change in buccal bone thickness and inclination of most of the posterior maxillary teeth was found.

ObjectiveTo evaluate long-term effects of presurgical nasoalveolar molding (PNAM) on occlusal relationships and craniofacial development in children with unilateral cleft lip and palate (UCLP) using the 5-year-olds' index and cone-beam computed tomography (CBCT).DesignRetrospective cohort study.SettingTertiary-level Class A specialized stomatological hospital.ParticipantsSixty-four patients with complete UCLP.InterventionsThirty patients underwent PNAM treatment, and 34 patients did not.Main Outcome MeasureCBCT scans were used to assess dental occlusion via the 5-year-old index and to measure craniofacial variables. Intergroup differences were analyzed using the Mann-Whitney U test and the chi-square test, while the correlation between occlusion and craniofacial variables was examined using the Pearson correlation coefficient. A P value of less than .05 was considered statistically significant.ResultsA comparison of 5-year-olds' index scores in the two groups showed median scores of 3.0 (PNAM group) and 3.5 (non-PNAM group), respectively, with no significant difference found. For prognosis, no significant difference in orthognathic surgery need was found, but the PNAM group had a higher nonsurgery rate (60%) versus 50% in the non-PNAM group. In craniofacial development, the PNAM group had significantly longer maxillary length; favorable prognosis group showed smaller SNB angles, shorter mandibular body, larger ANB angles, and longer maxillary length. Correlation analysis showed 5-year-olds' index positively correlated with SNB/mandibular body length, and negatively with ANB/maxillary length.ConclusionsEarly childhood evaluations showed that treatment with/without PNAM in the neonatal period was not a major factor influencing occlusal relationships or craniofacial development in patients with UCLP. Furthermore, correlation between 5-year-olds' index and skeletal morphology in children with UCLP has been confirmed.

Daily online adaptive radiation therapy (OART) allows the creation of individualized safety margins based on daily optimized treatment planning, taking into account daily anatomy variations. The goal of our work was to compare the protocols and safety margins of a non-adapted, image- guided radiation therapy (IGRT), a published Danish and the National Institute of Oncology of Hungary (NIO) protocol in the first eight bladder cancer patients treated with OART. Three cone beam CTs (CBCT) were taken during the first three fractions: after patient positioning, after adaptation, and at the end of treatment. Bladder contours, extensions and margins were established on all CT images (n=512). Intraadaptational, intrafractional and interfractional changes were examined. Regarding PTV, a mean reduction of 12.2% could be achieved using the Danish, and 34.1% using the NIO protocol compared to IGRT. Healthy tissue irradiated by the prescribed dose could be reduced. The target volume miss is decreased as well. PTV could be reduced and optimized, increasing healthy tissue protection using OART.

Aim: Estimating dental age is essential in forensic and clinical settings. This systematic review and meta-analysis examined the correlation between chronological age and cone-beam computed tomography (CBCT)-derived metrics, pulp volume (PV), and pulp/tooth volume ratio (PTVR) to ascertain their diagnostic use for age assessment. Methods: A thorough literature review was performed to locate studies evaluating the correlation between PV or PTVR and chronological age utilizing CBCT. Meta-analyses were conducted individually for each parameter. The variability of studies, the strength of correlations, and methodological variances were thoroughly examined. Results: Five studies evaluating PV demonstrated a robust inverse connection with chronological age (pooled r = 0.56; 95% CI : 0.41–0.75; p &lt; .0001), albeit considerable heterogeneity ( I ² = 96%). A meta-analysis of nine studies on PTVR revealed no significant link (pooled r = 1.11; 95% CI : 0.74–1.66; p = .61), exhibiting great heterogeneity ( I ² = 99%), which was ascribed to variations in tooth type, imaging techniques, segmentation methods, and demographic factors. Conclusion: The PV evaluated using CBCT is a dependable, non-invasive method for estimating age. Nonetheless, PTVR exhibits inconsistent diagnostic value owing to methodological discrepancies. Standardized methods and population-specific models are essential to improve the precision and relevance of CBCT-based age estimation. The International Prospective Register of Systematic Reviews (PROSPERO) Registration No.: CRD420251052457. Available from: https://www.crd.york.ac.uk/PROSPERO/view/CRD420251052457

To compare the accuracy of the Raypex-6 apex locator and Cone-beam computed tomography (CBCT) in measuring root canal length under different apical foramen diameters and root canal curvature angles. Sixty in vitro root canal specimens were selected. The apical foramen was destroyed retrogradely, and the apical foramen diameter was adjusted artificially to range from 0.2mm to 2.0mm. A stereomicroscope was used to determine the actual root canal length. CBCT and the Raypex-6 apex locator were used to independently measure root canal length. We compared the actual length with that measured by CBCT and the Raypex-6 apex locator. A chi-square test was applied to the data, with measurement differences of 0-0.5mm classified as accurate. Finally, differential analysis and correlation analysis were performed. CBCT accuracy did not differ significantly across apical foramen diameter or root canal curvature angle groups (P > 0.05). Raypex-6 apex locator accuracy also showed no significant differences among groups stratified by root canal curvature angle (P > 0.05). In contrast, Raypex-6 apex locator accuracy varied significantly with apical foramen diameter and area (P < 0.001). The accuracy of CBCT and Raypex-6 apex locator for measuring root canal length does not depend on the root canal curvature angle. However, when the apical foramen diameter exceeds 1.6mm, or the apical foramen area exceeds 1.844mm², Raypex-6 apex locator accuracy decreases significantly. Therefore, CBCT is recommended for clinical root canal length measurement under these conditions.

To describe the clinical and radiographic outcomes associated with the use of a novel posterior occlusal splint (nPOS) in adults with temporomandibular joint osteoarthritis (TMJ-OA) and its impact on condylar morphology, as assessed by cone-beam computed tomography (CBCT). This retrospective study included 156 adult patients (24 males, 132 females; mean age: 27.05 ± 9.18 years) diagnosed with TMJ-OA, with 182 osteoarthritis (OA)-like condyles. All patients received treatment with the nPOS between January 2020 and December 2024. Clinical symptoms were evaluated at six time points: T0 (baseline), T1 (1 month), T2 (3 months), T3 (6 months), T4 (12 months), and T5 (24 months) after treatment initiation. CBCT assessments of condylar morphology were performed at T0, T3, T4, and T5. Wilcoxon signed-rank test, paired t test, McNemar test were employed to evaluate longitudinal changes in clinical symptoms. Significant improvements in maximum mouth opening (MMO) were observed from T2 to T4 (P < 0.01). Significant reductions in spontaneous pain were observed at both T3 and T4 (P < 0.01), while reductions in joint tenderness and joint noises were statistically confirmed at T3 (P < 0.01). No significant changes were noted for intermittent closed lock (ICL), muscle pain, push-back mandibular test, or altered mouth opening patterns (P > 0.01). CBCT showed a significant improvement in condylar cortical score at T3 compared to T0 (P < 0.01). In this observational study, the use of nPOS was associated with improvements in key TMJ-OA symptoms and condylar morphology, especially during the early to mid-treatment phase (3-12 months). These findings warrant validation in randomized controlled trials before clinical adoption. Furthermore, long-term effects require further investigation.

ABSTRACTLarge cyst‐like periapical radiolucencies with cortical perforations are traditionally managed surgically. However, nonsurgical retreatment may be effective when adequate disinfection and seal are achieved. This case report describes the management of a large maxillary periapical lesion involving five endodontically treated teeth (from 21 to 25), associated with buccal and palatal cortical perforations and maxillary sinus involvement. Nonsurgical retreatment using enhanced disinfection and obturation with a hydraulic calcium‐silicate sealer was performed. Successfully establishing patency was reflected in the creation of evident sealer puffs. Radiographic follow‐up at 8 months and 2 years using Cone‐Beam Computed Tomography (CBCT) demonstrated substantial bone fill and near‐complete re‐establishment of both cortical plates. The patient remained asymptomatic throughout the follow‐up period. This case illustrates that, when meticulously executed and supported by modern bioactive materials, nonsurgical retreatment can promote significant healing even in extensive lesions, although surgical intervention remains the recommended secondary option when indicated.