Diagnostic accuracy of artificial intelligence versus manual detection in marginal bone loss around fixed semicolon. a systematic review.

Objectives: The aim of the review is to evaluate the existing precision of artificial intelligence (AI) in detecting Marginal Bone Loss (MBL) around prosthetic crowns using 2-Dimentional radiographs. It also summarises the recent advances and future challenges associated to their clinical application. Methodology: A literature survey of electronic databases was conducted in November 2023 to recognize the relevant articles. MeSH terms/keywords were used to search ("panoramic" OR "pantomogram" OR "orthopantomogram" OR "opg" OR "periapical") AND ("artificial intelligence" OR "deep" OR "machine" OR "automated" OR "learning") AND ("periodontal bone loss") AND ("prosthetic crown") in PubMed database, SCOPUS, COCHRANE library, EMBASE, CINAHL and Science Direct. RESULTS: The searches identified 49 relevant articles, of them 5 articles met the inclusion criteria were included. The outcomes measured were sensitivity, specificity and accuracy of AI models versus manual detection in panoramic and intraoral radiographs. Few studies reported no significant difference between AI and manual detection, whereas majority demonstrated the superior ability of AI in detecting MBL. CONCLUSIONS: AI models show promising accuracy in analysing complex datasets and generate accurate predictions in the MBL around fixed prosthesis. However, these models are still in the developmental phase. Therefore, it is crucial to assess the effectiveness and reliability of these models before recommending their use in clinical practice.

Third molar eruption in dental panoramic radiographs as a feature for forensic age assessment - new reference data from a German population.

Forensic age assessment in the living can provide legal certainty when an individual's chronological age is unknown or when age-related information is questionable. An established method involves assessing the eruption of mandibular third molars through dental panoramic radiographs (PAN). In age assessment procedures, the respective findings are compared to reference data. The objective of this study was to generate new reference data in line with the required standards for mandibular third molar eruption within a German population. For this purpose, 605 PANs from 302 females and 303 males aged 15.04 to 25.99 years were examined. The PANs were acquired between 2013 and 2020, and the development of the mandibular third molars was rated independently by two experienced examiners using the Olze et al. staging scale from 2012. In case of disagreement in the assigned ratings, a consensus was reached through arbitration. While the mean, median and minimum ages were observed to increase with each stage of mandibular third molar eruption according to the Olze method, there was considerable overlap in the distribution of age between the stages. The minimum age for stage D, which corresponds to complete tooth eruption, was 16.1 years for females and 17.1 years for males. Thus, the completion of mandibular third molar eruption was found in both sexes before reaching the age of 18. In all individuals who had at least one tooth with completed eruption and who were younger than 17.4 years of age (n = 10), mineralization of the teeth in question was not complete. Based on our findings, the feature of assessing mandibular third molar eruption in PAN cannot be relied upon for determining age of majority.

The comparison of visibility of the maxillary sinus septa between cone-beam computed tomography scans and panoramic radiograph images as dependent on the cortical bone thickness: a retrospective comparative study.

PURPOSE: To analyze the visibility of the maxillary sinus septa (MSS) in panoramic radiography (PR) versus cone beam computed tomography (CBCT) and to investigate whether the buccal cortical bone thickness (BT) or the septa dimensions influence their visibility. METHODS: Corresponding PR and CBCT images of 355 patients were selected and examined for MSS visibility. The septa dimensions (width, height, depth) and the BT were measured. Results were analysed statistically. RESULTS: Comparing the corresponding regions on CBCT and PR, 170 MSS were identified; however, only 106 of these were also visible using PR. The MSS visibility was significantly higher on CBCT versus PR images (P1: p = 0.039, P2: p = 0.015, M1: p = 0.041, M2: p = 0.017, M3: p = 0.000), except region C (p = 0.625). Regarding the measurements of MSS dimensions, only the height in region M1 (p = 0.013) and the width in region P2 (p = 0.034) were significantly more visible on CBCT. The BT in the area of the MSS was found to have a marginal influence on its visibility on the PR images only in regions M3 and M1 (M3: p = 0.043, M1: p = 0.047). In terms of MSS visibility based on the dimensions, significance was found for all three influencing variables only in region P2 (width; p = 0.041, height; p = 0.001, depth; p = 0.007). There were only isolated cases of further significance: M3 for width (p = 0.043), M2 for height (p = 0.024), and P1 for depth (p = 0.034), no further significance was noted. CONCLUSION: MSS visibility appears significantly higher on CBCT versus PR images. It is concluded that the septa dimensions and BT can influence MSS visibility on PR images just in certain regions.

Comparative efficacy of digital panoramic radiograph and cone beam computed tomography in locating mandibular foramen.

OBJECTIVE: The aim of this study was to compare the accuracy of panoramic radiographs (PAN) with cone-beam computed tomography (CBCT) in locating the position of the mandibular foramen (MF). PATIENTS AND METHODS: A total of 100 patients who underwent CBCT and panoramic imaging were included in the study. The location of the MF was evaluated anterior-posteriorly and superior-inferiorly on both CBCT and panoramic radiographs. Measurements were taken by two examiners, reviewed by a radiologist, and intra-examiner variability was assessed. A comparison of statistical analysis was performed using the Mann-Whitney U test, independent and paired t-test. A p-value ≤0.05 was deemed significant. RESULTS: The mean age of the patients was 35.03 years, with 51% females and 49% males. CBCT and panoramic radiographs showed comparable accuracy in locating the left (p=0.937) and right (p=0.371) MF anterior-posteriorly. In the superior-inferior dimension, the accuracy of CBCT and panoramic radiographs were comparable in locating the right (p=0.292) and left (p=0.640) MF. The gender-based accuracy of PAN and CBCT radiographs in locating the right (p=0.353) and left (p=0.985) MF was comparable. CONCLUSIONS: The study concludes that panoramic radiographs showed comparable accuracy in effectively locating MF in comparison to CBCT. The influence of gender and anatomic location (right and left sides of MF) on MF identification with panoramic radiographs was insignificant. Digital panoramic radiographs are an effective tool in accurately identifying MF location in clinical practice.

[Imaging in temporomandibular joint disorders]. / Beeldvorming bij kaakgewrichtsstoornissen.

The temporomandibular joint is a unique and complex joint. Various imaging techniques have been developed to properly visualize this complex joint, such as conventional radiology, orthopantomography, CBCT and MRI. Imaging can contribute to the differential diagnosis of temporomandibular joint disorders. Common joint disorders are arthritis and internal derangement. Osseous changes of the temporomandibular joint can be clearly visualized with CBCT. MRI is superior for imaging the internal anatomy of the temporomandibular joint and is preferred in the context of internal derangement.

Isolated dentinogenesis imperfecta: Novel DSPP variants and insights on genetic counselling.

OBJECTIVE: Dentinogenesis imperfecta (DI) is an inherited dentin defect and may be isolated or associated with disorders such as osteogenesis imperfecta, odontochondrodysplasia Ehler-Danlos and others. Isolated DI is caused mainly by pathogenic variants in DSPP gene and around 50 different variants have been described in this gene. Herein, we report on 19 patients from two unrelated Egyptian families with isolated DI. Additionally, we focused on genetic counselling of the two families. MATERIALS AND METHODS: The patients were examined clinically and dentally. Panoramic X-rays were done to some patients. Whole exome sequencing (WES) and Sanger sequencing were used. RESULTS: WES revealed two new nonsense variants in DSPP gene, c.288T > A (p.Tyr96Ter) and c.255G > A (p.Trp85Ter). Segregation analysis by Sanger sequencing confirmed the presence of the first variant in all affected members of Family 1 while the second variant was confirmed to be de novo in the patient of Family 2. CONCLUSIONS AND CLINICAL RELEVANCE: Our study extends the number of DSPP pathogenic variants and strengthens the fact that DSPP is the most common DI causative gene irrespective of patients' ethnicity. In addition, we provide insights on genetic counseling issues in patients with inherited DSPP variants taking into consideration the variable religion, culture and laws in our society.

Assessing current clinical eruption stage of mandibular third molars by dental panoramic radiography.

OBJECTIVE: We examined whether dental panoramic radiography (PAN) can be used to identify the clinical stage of eruption of mandibular third molars at the time of radiological examination. MATERIALS AND METHODS: Cross-sectional data included records from clinical oral examination and PANs of university students. In the retrospective analysis of 345 mandibular third molars in 189 participants (20% men, 80% women; mean age 20.7 years; standard deviation [SD] ± 0.6), clinical stages of eruption were compared with their radiographic depth in bone, inclination, and root development. Statistics included χ2, Mann-Whitney U tests, and logistic regression. RESULTS: Significant (p < 0.001) predictor variables for assessing the clinical stage of eruption were radiographic depth in bone and inclination. All teeth radiologically at a depth of the cementoenamel (CE) junction of the neighbouring second molar or deeper were clinically unerupted. Above the CE junction, 80% of vertical and 97% of distoangular teeth were connected to the oral cavity, and 82% of mesioangular and 69% of horizontal teeth were clinically unerupted. CONCLUSION: All teeth below or at the CE junction are clinically unerupted. Above the CE junction, stage of eruption should be assessed together with the inclination, but horizontally inclined teeth are recommended to be verified clinically.

Fully automated deep learning approach to dental development assessment in panoramic radiographs.

BACKGROUND: Dental development assessment is an important factor in dental age estimation and dental maturity evaluation. This study aimed to develop and evaluate the performance of an automated dental development staging system based on Demirjian's method using deep learning. METHODS: The study included 5133 anonymous panoramic radiographs obtained from the Department of Pediatric Dentistry database at Seoul National University Dental Hospital between 2020 and 2021. The proposed methodology involves a three-step procedure for dental staging: detection, segmentation, and classification. The panoramic data were randomly divided into training and validating sets (8:2), and YOLOv5, U-Net, and EfficientNet were trained and employed for each stage. The models' performance, along with the Grad-CAM analysis of EfficientNet, was evaluated. RESULTS: The mean average precision (mAP) was 0.995 for detection, and the segmentation achieved an accuracy of 0.978. The classification performance showed F1 scores of 69.23, 80.67, 84.97, and 90.81 for the Incisor, Canine, Premolar, and Molar models, respectively. In the Grad-CAM analysis, the classification model focused on the apical portion of the developing tooth, a crucial feature for staging according to Demirjian's method. CONCLUSIONS: These results indicate that the proposed deep learning approach for automated dental staging can serve as a supportive tool for dentists, facilitating rapid and objective dental age estimation and dental maturity evaluation.

Endoscopically-assisted extraction of broken roots or fragments within the mandibular canal: a retrospective case series study.

PURPOSE: To assess the impact of endoscope-assisted fractured roots or fragments extraction within the mandibular canal, along with quantitative sensory testing (QST) alterations in the inferior alveolar nerve (IAN). METHODS: Six patients with lower lip numbness following mandibular third molar extraction were selected. All patients had broken roots or fragments within the mandibular canal that were extracted under real-time endoscopic assistance. Follow-up assessments were conducted on postoperative days 1, 7, and 35, including a standardized QST of the lower lip skin. RESULTS: The average surgical duration was 32.5 min, with the IAN exposed in all cases. Two of the patient exhibited complete recovery of lower lip numbness, three experienced symptom improvement, and one patient remained unaffected 35 days after the surgery. Preoperative QST results showed that the mechanical detection and pain thresholds on the affected side were significantly higher than those on the healthy side, but improved significantly by postoperative day 7 in five patients, and returned to baseline in two patients on day 35. There were no significant differences in the remaining QST parameters. CONCLUSIONS: All endoscopic surgical procedures were successfully completed without any additional postoperative complications. There were no cases of deterioration of IAN injury, and lower lip numbness recovered in the majority of cases. Endoscopy allowed direct visualization and examination of the affected nerve, facilitating a comprehensive analysis of the IAN.

Radiological features and treatment of erupting maxillary canines in relation to the occurrence of dental developmental abnormalities.

OBJECTIVE: This study aimed to investigate the radiological features of the permanent canines and the treatment carried out to ensure their eruption relative to certain features involved in Dental Anomaly Patterns (DAP). MATERIAL AND METHODS: The cross-sectional part of this retrospective register--based study focused on 1,315 dental panoramic tomographs (DPTs) of children aged 8.5-10.5 years, while the longitudinal part involved information on the treatment provided for 1,269 canines after the DPTs and on their eruption into the oral cavity. RESULTS: The canines of the DAP children more often showed evidence of early treatment (p = 0.014), with girls having more frequently interceptive treatment (p = 0.004) and boys early headgear (p = 0.022). Delayed dental age was associated with early treatment (OR 3.29, 95% CI 1.08-9.99). Either no or clear overlapping of a canine with the lateral incisor occurred more often in the DAP children, whereas canine inclination did not differ between the groups. The root development stage of the canine was more often either beginning or well advanced in the DAP children. CONCLUSIONS: The children with dental developmental abnormalities more often showed evidence of early treatment for the canines. Monitoring of the erupting canines after the first mixed stage is important to enable timely early treatment.

Predicting chronological age of 14 or 18 in adolescents: integrating dental assessments with machine learning.

AIM: Age estimation plays a critical role in personal identification, especially when determining compliance with the age of consent for adolescents. The age of consent refers to the minimum age at which an individual is legally considered capable of providing informed consent for sexual activities. The purpose of this study is to determine whether adolescents meet the age of 14 or 18 by using dental development combined with machine learning. METHODS: This study combines dental assessment and machine learning techniques to predict whether adolescents have reached the consent age of 14 or 18. Factors such as the staging of the third molar, the third molar index, and the visibility of the periodontal ligament of the second molar are evaluated. RESULTS: Differences in performance metrics indicate that the posterior probabilities achieved by machine learning exceed 93% for the age of 14 and slightly lower for the age of 18. CONCLUSION: This study provides valuable insights for forensic identification for adolescents in personal identification, emphasizing the potential to improve the accuracy of age determination within this population by combining traditional methods with machine learning. It underscores the importance of protecting and respecting the dignity of all individuals involved.

Accuracy of different dental age assessment methods to determine chronological age among children with special needs.

AIM: This study aimed to validate the accuracy of dental age (DA) based on the dental development of permanent teeth in children with special needs using Demirjian, Willems, and London Atlas methods and to correlate the dental and chronological age (CA) of children with special needs in Malaysia. MATERIALS AND METHODS: The panoramic radiographic images belonging to children with special needs from the two teaching dental hospitals in Malaysia aged between 5 and 16 years were included in the study. The evaluation was performed by two observers using three methods (London Atlas, Demirjian, and Willems methods) to estimate the accurate DA. The outcome was determined by comparing the mean of the DA and CA. RESULTS: A total of 52 panoramic radiographs were available for the analysis. The London Atlas and Demirjian methods overestimated the DA with a mean of 0.05 and 0.20 years, respectively, while the Willems method underestimated by 0.19 years. The London Atlas method was highly precise and accurate, while Demirjian and Willems methods were the least precise and accurate. CONCLUSION: The London Atlas method of DA estimation is highly accurate and valid for children with special needs in the Malaysian population, followed by the Willems and Demirjian methods.

Graphical user interface-based convolutional neural network models for detecting nasopalatine duct cysts using panoramic radiography.

Nasopalatine duct cysts are difficult to detect on panoramic radiographs due to obstructive shadows and are often overlooked. Therefore, sensitive detection using panoramic radiography is clinically important. This study aimed to create a trained model to detect nasopalatine duct cysts from panoramic radiographs in a graphical user interface-based environment. This study was conducted on panoramic radiographs and CT images of 115 patients with nasopalatine duct cysts. As controls, 230 age- and sex-matched patients without cysts were selected from the same database. The 345 pre-processed panoramic radiographs were divided into 216 training data sets, 54 validation data sets, and 75 test data sets. Deep learning was performed for 400 epochs using pretrained-LeNet and pretrained-VGG16 as the convolutional neural networks to classify the cysts. The deep learning system's accuracy, sensitivity, and specificity using LeNet and VGG16 were calculated. LeNet and VGG16 showed an accuracy rate of 85.3% and 88.0%, respectively. A simple deep learning method using a graphical user interface-based Windows machine was able to create a trained model to detect nasopalatine duct cysts from panoramic radiographs, and may be used to prevent such cysts being overlooked during imaging.

Accuracy in the Diagnostics of Styloid Process - Panoramic Radiograph vs. Computed Tomography: A Comparative Study.

BACKGROUND/AIM: The styloid process (SP) becomes clinically relevant when it shows enlargement (>30 mm) in the sense of an elongated SP (ESP) and/or increasing calcification leading to Eagle Syndrome (ES). Panoramic radiograph (PR) or computed tomography (CT) are part of the routine diagnostics in ES. Currently, CT is considered the gold standard. The aim of this study was to investigate the accuracy in the diagnostics/measurements of SP/ESP throughout a comparative study between PR and CT. Furthermore, in addition to measuring established parameters, this study aimed to determine the currently unexamined width in the base and tip of the SP. PATIENTS AND METHODS: The present study examined the radiological findings of bilateral SP in 100 patients who received both PR and CT on the same day. Measurements of the length of the SP and width at the basis and tip were performed. Furthermore, calcification patterns, Langlais classification and the prevalence of ESP were analyzed. RESULTS: There was a highly significant correlation between PR and CT measuring SP for every parameter. Males showed significantly longer SP than females among the age group between 18-75 years. The results of the length measurements of the SP (male: right SP=32.98 mm; left SP=35.21 mm; female: right SP=30.31 mm; left SP=30.92 mm) significantly exceeded the values of comparable studies. CONCLUSION: Consequently, it can be concluded that PR provides accurate measurements when compared to CT for measuring and diagnosing SP/ESP/Eagle syndrome. This study was one of the first to examine the width of the SP in the base and tip, thus these measurements can serve as a baseline for further studies. Since the mean lengths of SP exceeded 30.0 mm in the present study, these findings raise the question of whether the cut-off of 30.0 mm is adequate for the diagnosis of ESP.

Applicability of Cameriere's third molar maturity index on orthopantomograms and computed tomography scans from a French population.

Cameriere developed a method on orthopantomograms (OPG) to assess adult age of 18 years based on the relationship between age and the third molar maturity index I3M. The aim of this study was to evaluate whether Cameriere's method could be applied to computed-tomography scans (CT-scans) from a population of French juveniles and young adults and compare the results obtained from OPG of the same individuals. Our sample comprised 200 examinations that had been performed at the radiological department of a French University hospital between 2007 and 2020. Each patient had received an OPG and a cranial CT scan for medical purposes, and we used a similar adaptation of I3M based on OPG to determine the I3M based on CT scans. Due to exclusion criteria, our final sample comprised 71 OPGs and 63 CT scans. Based on the 71 OPGs, there was concordance between chronological age and estimated age, with a sensitivity of 78.57%, a specificity of 89.47%, and a misclassified rate of 18.03% based on tooth 38, and a sensitivity of 78.79%, a specificity of 91.67%, and a misclassified rate of 17.78% based on tooth 48. Our results based on CT scans presented concordance between chronological age and estimated age for tooth 38 described by a sensitivity of 77.78%, a specificity of 94.12%, and a misclassified rate of 16.98%. The concordance between chronological age and estimated age based on 48 had a sensitivity of 75.00%, a specificity of 93.75%, and a misclassified rate of 19.23%. The > 90% ICC indicate an excellent similarity between measurements of teeth 38 and 48 based on OPGs and CT scans. This study has revealed the applicability of the Cameriere's method to calculate the I3M based on CT scans from a French population. The results based on CT scans are similar to results based on OPGs from the same individuals.

Investigation of the relationship between the mandibular third molar teeth and the inferior alveolar nerve using posteroanterior radiographs: a pilot study.

BACKGROUND: The most severe complication that can occur after mandibular third molar (MM3) surgery is inferior alveolar nerve (IAN) damage. It is crucial to have a comprehensive radiographic evaluation to reduce the possibility of nerve damage. The objective of this study is to assess the diagnostic accuracy of panoramic radiographs (PR) and posteroanterior (PA) radiographs in identifying the association between impacted MM3 roots and IAN. METHODS: This study included individuals who had PR, PA radiographs, and cone beam computed tomography (CBCT) and who had at least one impacted MM3. A total of 141 impacted MM3s were evaluated on CBCT images, and the findings were considered gold standard. The relationship between impacted MM3 roots and IAN was also evaluated on PR and PA radiographies. The data was analyzed using the McNemar and Chi-squared tests. The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and diagnostic accuracy of PR and PA radiographies were determined. RESULTS: Considering CBCT the gold standard, the relationship between MM3 roots and IAN was found to be statistically significant between PR and CBCT (p = 0.00). However, there was no statistically significant relationship between PA radiography and CBCT (0.227). The study revealed that the most prevalent limitation of the PR in assessing the relationship between MM3 roots and IAN was the identification of false-positive relationship. CONCLUSIONS: PA radiography may be a good alternative in developing countries to find out if there is a contact between MM3 roots and IAN because it is easier to get to, cheaper, and uses less radiation.

Correlation between dental and skeletal maturity in Korean children based on dental maturity percentile: a retrospective study.

BACKGROUND: The correlation between dental maturity and skeletal maturity has been proposed, but its clinical application remains challenging. Moreover, the varying correlations observed in different studies indicate the necessity for research tailored to specific populations. AIM: To compare skeletal maturity in Korean children with advanced and delayed dental maturity using dental maturity percentile. DESIGN: Dental panoramic radiographs and cephalometric radiographs were obtained from 5133 and 395 healthy Korean children aged between 4 and 16 years old. Dental maturity was assessed with Demirjian's method, while skeletal maturity was assessed with the cervical vertebral maturation method. Standard percentile curves were developed through quantile regression. Advanced (93 boys and 110 girls) and delayed (92 boys and 100 girls) dental maturity groups were defined by the 50th percentile. RESULTS: The advanced group showed earlier skeletal maturity in multiple cervical stages (CS) in both boys (CS 1, 2, 3, 4, and 6) and girls (CS 1, 3, 4, 5, and 6). Significant differences, as determined by Mann-Whitney U tests, were observed in CS 1 for boys (p = 0.004) and in CS 4 for girls (p = 0.037). High Spearman correlation coefficients between dental maturity and cervical vertebral maturity exceeded 0.826 (p = 0.000) in all groups. CONCLUSION: A correlation between dental and skeletal maturity, as well as advanced skeletal maturity in the advanced dental maturity group, was observed. Using percentile curves to determine dental maturity may aid in assessing skeletal maturity, with potential applications in orthodontic diagnosis and treatment planning.

Performance of a commercially available Generative Pre-trained Transformer (GPT) in describing radiolucent lesions in panoramic radiographs and establishing differential diagnoses.

OBJECTIVES: To evaluate the performance of a commercially available Generative Pre-trained Transformer (GPT) in describing and establishing differential diagnoses for radiolucent lesions in panoramic radiographs. MATERIALS AND METHODS: Twenty-eight panoramic radiographs, each containing a single radiolucent lesion, were evaluated in consensus by three examiners and a commercially available ChatGPT-3.5 model. They provided descriptions regarding internal structure (radiodensity, loculation), periphery (margin type, cortication), shape, location (bone, side, region, teeth/structures), and effects on adjacent structures (effect, adjacent structure). Diagnostic impressions related to origin, behavior, and nature were also provided. The GPT program was additionally prompted to provide differential diagnoses. Keywords used by the GPT program were compared to those used by the examiners and scored as 0 (incorrect), 0.5 (partially correct), or 1 (correct). Mean score values and standard deviation were calculated for each description. Performance in establishing differential diagnoses was assessed using Rank-1, -2, and - 3. RESULTS: Descriptions of margination, affected bone, and origin received the highest scores: 0.93, 0.93, and 0.87, respectively. Shape, region, teeth/structures, effect, affected region, and nature received considerably lower scores ranging from 0.22 to 0.50. Rank-1, -2, and - 3 demonstrated accuracy in 25%, 57.14%, and 67.85% of cases, respectively. CONCLUSION: The performance of the GPT program in describing and providing differential diagnoses for radiolucent lesions in panoramic radiographs is variable and at this stage limited in its use for clinical application. CLINICAL RELEVANCE: Understanding the potential role of GPT systems as an auxiliary tool in image interpretation is imperative to validate their clinical applicability.

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.

A comparison of estimated age based on pulp volume from cone beam computed tomography (CT) images and panoramic radiography data with chronological age.

This retrospective study was conducted to evaluate different methods for dental age estimation in children and to examine the feasibility of using cone beam computed tomography (CBCT) data for age estimation. A total of 200 radiographic records (both digital panoramic radiographs and CBCTs) were acquired from 100 children aged 9 to 16 years, all taken on the same dates. Radiographic data was acquired from archived records and included both panoramic radiography and CBCT data belonging to the same individual. CBCT was used when panoramic radiographic data was insufficient. The pulp volume and pulp/tooth volume ratio of the left first molar teeth in the mandible were calculated from the CBCT data using MIMICS software. In addition, age was estimated by the Demirjian and Willems methods from data obtained from panoramic radiography images. Statistical analyses and linear regression analysis were performed as necessary. There was a statistically significant difference between the mean difference between the Demirjian method and chronological age, and between the Willems method and chronological age (p < 0.001). Statistically significance was achieved in a linear regression model created from pulp volume (R2 = 0.098) and pulp/tooth volume ratio (R2 = 0.395) data for the estimated dental age analysis (p < 0.001) and a negative correlation was observed with chronological age. When compared estimated dental age from CBCT data with chronological age, the pulp/tooth volume ratio method yielded results closer to chronological age than using only pulp volume data. When considering both panoramic radiographic age estimation methods and age estimation methods using CBCT data, we found that the results obtained with the Willems method, a panoramic radiographic age estimation technique, provided the closest results to the chronological age. More contributions should be made to the literature regarding the feasibility of age estimation using pulp and tooth volume as an alternative method.