Diagnostic accuracy of artificial intelligence models in detecting osteoporosis using dental images: a systematic review and meta-analysis.

The current study aimed to systematically review the literature on the accuracy of artificial intelligence (AI) models for osteoporosis (OP) diagnosis using dental images. A thorough literature search was executed in October 2022 and updated in November 2023 across multiple databases, including PubMed, Scopus, Web of Science, and Google Scholar. The research targeted studies using AI models for OP diagnosis from dental radiographs. The main outcomes were the sensitivity and specificity of AI models regarding OP diagnosis. The "meta" package from the R Foundation was selected for statistical analysis. A random-effects model, along with 95% confidence intervals, was utilized to estimate pooled values. The Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2) tool was employed for risk of bias and applicability assessment. Among 640 records, 22 studies were included in the qualitative analysis and 12 in the meta-analysis. The overall sensitivity for AI-assisted OP diagnosis was 0.85 (95% CI, 0.70-0.93), while the pooled specificity equaled 0.95 (95% CI, 0.91-0.97). Conventional algorithms led to a pooled sensitivity of 0.82 (95% CI, 0.57-0.94) and a pooled specificity of 0.96 (95% CI, 0.93-0.97). Deep convolutional neural networks exhibited a pooled sensitivity of 0.87 (95% CI, 0.68-0.95) and a pooled specificity of 0.92 (95% CI, 0.83-0.96). This systematic review corroborates the accuracy of AI in OP diagnosis using dental images. Future research should expand sample sizes in test and training datasets and standardize imaging techniques to establish the reliability of AI-assisted methods in OP diagnosis through dental images.

Radiomorphometric indices of the mandible as indicators of decreased bone mineral density and osteoporosis - meta-analysis and systematic review.

This review aims to evaluate the accuracy of various mandibular radiomorphometric indices in comparison with DEXA BMD measurements in the diagnosis of osteopenia and osteoporosis based on a meta-analysis of the sensitivity and specificity of the indices. PRISMA statement was followed. The materials for analysis were collected in August 2023 by searching three databases: PubMed Central, Web of Science, and Scopus. The selection of studies consisted of three selection stages, and 64 articles were finally obtained. Quality assessment was performed with the QUADAS-2 tool, and the general methodological quality of retrieved studies was low. Statistical analysis was performed based on 2 × 2 tables and estimated sensitivity and specificity were obtained using SROC curves. The most used indices were MCI, MCW and PMI. The best results in detecting reduced BMD obtained for MCW ≤ 3 mm, estimated sensitivity and specificity were 0.712 (95% CI, 0.477-0.870) and 0.804 (95% CI, 0.589-0.921), respectively. The most prone to the risk of bias is the MCI due to the examiner's subjectivism. Radiomorphometric indices of the mandible can be useful as a screening tool to identify patients with low BMD, but should not be used as a diagnostic method. Further research needs to focus on analysing the ability of the indices to detect osteoporosis and also in combination the indices with clinical parameters.

Fractal characteristics of the trabecular pattern of the mandible in patients with renal transplantation.

OBJECTIVE: In this study, we examined the mandibular trabecular bone structures by performing fractal dimension (FD) analysis in patients who underwent renal transplantation (RTx). METHODS: Our study is an observational study with 69 RTx patients and 35 control group patients. The mean FD values of the patient and control groups were calculated and compared. In addition, biochemical parathyroid hormone (PTH), serum calcium, phosphorus, alkaline phosphatase (ALP), and vitamin-D parameters and FD values of both groups were analyzed. RESULTS: FD values were significantly lower in the patient group than in the healthy group (p < .05). In the RTx group compared to the control group, ALP (90.71 ± 34.25-66.54 ± 16.8, respectively) (p < .001) and PTH (75.76 ± 38.01-38.17 ± 12.39, respectively) (p < .001) values were higher. There was a positive correlation between the FD values and ALP (rspearman  = .305, p = .011) and a negative correlation between FD values and vitamin-D (rspearman  = .287, p = .017) of patients with RTx. CONCLUSION: FD values were found to be lower in patients who underwent RTx compared to the control group. It should be considered that FD analysis can be a method that can be used to evaluate trabecular bone structure in patients undergoing RTx.

Radiographic diagnosis of periodontal diseases - Current evidence versus innovations.

Accurate diagnosis of periodontal and peri-implant diseases relies significantly on radiographic examination, especially for assessing alveolar bone levels, bone defect morphology, and bone quality. This narrative review aimed to comprehensively outline the current state-of-the-art in radiographic diagnosis of alveolar bone diseases, covering both two-dimensional (2D) and three-dimensional (3D) modalities. Additionally, this review explores recent technological advances in periodontal imaging diagnosis, focusing on their potential integration into clinical practice. Clinical probing and intraoral radiography, while crucial, encounter limitations in effectively assessing complex periodontal bone defects. Recognizing these challenges, 3D imaging modalities, such as cone beam computed tomography (CBCT), have been explored for a more comprehensive understanding of periodontal structures. The significance of the radiographic assessment approach is evidenced by its ability to offer an objective and standardized means of evaluating hard tissues, reducing variability associated with manual clinical measurements and contributing to a more precise diagnosis of periodontal health. However, clinicians should be aware of challenges related to CBCT imaging assessment, including beam-hardening artifacts generated by the high-density materials present in the field of view, which might affect image quality. Integration of digital technologies, such as artificial intelligence-based tools in intraoral radiography software, the enhances the diagnostic process. The overarching recommendation is a judicious combination of CBCT and digital intraoral radiography for enhanced periodontal bone assessment. Therefore, it is crucial for clinicians to weigh the benefits against the risks associated with higher radiation exposure on a case-by-case basis, prioritizing patient safety and treatment outcomes.

Artificial intelligence and radiomics in the diagnosis of intraosseous lesions of the gnathic bones: A systematic review.

BACKGROUND: The purpose of this systematic review (SR) is to gather evidence on the use of machine learning (ML) models in the diagnosis of intraosseous lesions in gnathic bones and to analyze the reliability, impact, and usefulness of such models. This SR was performed in accordance with the PRISMA 2022 guidelines and was registered in the PROSPERO database (CRD42022379298). METHODS: The acronym PICOS was used to structure the inquiry-focused review question "Is Artificial Intelligence reliable for the diagnosis of intraosseous lesions in gnathic bones?" The literature search was conducted in various electronic databases, including PubMed, Embase, Scopus, Cochrane Library, Web of Science, Lilacs, IEEE Xplore, and Gray Literature (Google Scholar and ProQuest). Risk of bias assessment was performed using PROBAST, and the results were synthesized by considering the task and sampling strategy of the dataset. RESULTS: Twenty-six studies were included (21 146 radiographic images). Ameloblastomas, odontogenic keratocysts, dentigerous cysts, and periapical cysts were the most frequently investigated lesions. According to TRIPOD, most studies were classified as type 2 (randomly divided). The F1 score was presented in only 13 studies, which provided the metrics for 20 trials, with a mean of 0.71 (±0.25). CONCLUSION: There is no conclusive evidence to support the usefulness of ML-based models in the detection, segmentation, and classification of intraosseous lesions in gnathic bones for routine clinical application. The lack of detail about data sampling, the lack of a comprehensive set of metrics for training and validation, and the absence of external testing limit experiments and hinder proper evaluation of model performance.

Evaluation of Bone Mineral Changes in Panoramic Radiographs of Hypothyroid and Hyperthyroid Patients Using Fractal Dimension Analysis.

Objective Hyperthyroidism and hypothyroidism are endocrinopathies that cause a decrease in bone mineral density. The aim of this study is to investigate possible bone changes in the mandible caused by hyperthyroidism and hypothyroidism using fractal analysis (FA) on panoramic radiographs. Material and Methods Panoramic radiographs of a total of 180 patients, including 120 patient groups (60 hyperthyroid, 60 hypothyroid) and 60 healthy control groups, were used. Five regions of interests (ROI) were determined from panoramic radiographs and FA was performed. ROI1: geometric midpoint of mandibular notch and mandibular foramen, ROI2: geometric midpoint of mandibular angle, ROI3: anterior of mental foramen, ROI4: basal cortical area from distal mental foramen to distal root of first molar, ROI5: geometric center of mandibular foramen and mandibular ramus. Results While a significant difference was observed between the patient and control groups regarding ROI1 and ROI2 (p < 0.05); there was no significant difference between the groups in relation to ROI3, ROI4, and ROI5. All FA values were lower in the hyperthyroid group than in the hypothyroid group. Conclusion Fractal analysis proves to be an effective method for early detection of bone mass changes. In the present study, it was concluded that while the mandibular cortical bone was intact, trabecular rich regions were affected by osteoporosis caused by thyroid hormones. Necessary precautions should be taken against the risk of osteoporosis in patients with thyroid hormone disorders.

Is the third molar the most frequently extracted tooth? A population-based study utilizing dental panoramic radiographs in adults.

OBJECTIVES: The study aimed to examine the authenticity of the often-mentioned statement that the third molar is the most frequently extracted tooth. This finding has not been shown previously in a large population-based sample. MATERIALS AND METHODS: Data comprised a nationally representative sample of 6082 panoramic radiographs taken from adults in the cross-sectional Health 2000 Survey. From the radiographs, all missing teeth were recorded. Information on congenital agenesis of individual teeth was retrieved from two published meta-analyses. Primary outcome was the frequency of missing teeth by tooth type. Explanatory variables were age, sex, and the jaw (maxilla/mandible). Statistical analyses included χ2 test and binomial logistic regression. RESULTS: Mean age of participants (46% men, 54% women) was 53 years (SD 14.6; range 30‒97 years). Missing teeth occurred more often in women than in men (P < 0.001). The third molar was most frequently missing and the canine least frequently. In the maxilla and mandible, the third molar was missing more often than each of the other tooth types up to the age of 80 years (P < 0.01). CONCLUSIONS: When considering the rates of congenital agenesis of individual teeth, it is concluded that the third molar remained the most common tooth extracted up till the age of 80 years. CLINICAL RELEVANCE: The third molar is the most common target for extraction, but also the most common tooth associated with malpractice claims, and therefore, calls for skills, adequate equipment, and other resources for a successful extraction.

Is cone-beam computed tomography (CBCT) an alternative to plain radiography in assessments of dental disease? A study of method agreement in a medically compromised patient population.

OBJECTIVES: Poor oral health and dental infections can jeopardize medical treatment and be life-threatening. Due to this, patients with head and neck malignancies, generalized tumor spread, organ transplant, or severe infection are referred for a clinical oral and radiographic examination. The aim of this study was to compare the diagnostic agreement of three radiographic modalities: intraoral radiographs (IO), panoramic radiographs (PX), and cone beam computed tomography (CBCT) for diagnosis of dental diseases. MATERIALS AND METHODS: Three hundred patients were examined with IO, PX, and CBCT. Periapical lesions, marginal bone level, and caries lesions were diagnosed separately by four oral radiologists. All observers also assessed six teeth in 30 randomly selected patients at two different occasions. Kappa values and percent agreement were calculated. RESULTS: The highest Kappa value and percent agreement were for diagnosing periapical lesions (0.76, 97.7%), and for the assessment of marginal bone level, it varied between 0.58 and 0.60 (87.8-89.3%). In CBCT, only 44.4% of all teeth were assessable for caries (Kappa 0.68, 93.4%). The intra-observer agreement, for all modalities and diagnoses, showed Kappa values between 0.5 and 0.93 and inter-observer agreement varied from 0.51 to 0.87. CONCLUSIONS: CBCT was an alternative to IO in diagnosing periapical lesions. Both modalities found the same healthy teeth in 93.8%. All modalities were performed equally regarding marginal bone level. In caries diagnosis, artifacts were the major cause of fallout for CBCT. CLINICAL RELEVANCE: Intraoral radiography is the first-hand choice for diagnosing dental disease. For some rare cases where intraoral imaging is not possible, a dedicated panoramic image and/or CBCT examination is an alternative.

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.

Diagnostic accuracy of periapical radiography and panoramic radiography in the detection of apical periodontitis: a systematic review and meta-analysis.

OBJECTIVE: Apical periodontitis (AP) is one of the most common pathologies of the oral cavity. An early and accurate diagnosis of AP lesions is crucial for proper management and planning of endodontic treatments. This study investigated the diagnostic accuracy of periapical radiography (PR) and panoramic radiography (PAN) in the detection of clinically/surgically/histopathologically confirmed AP lesions. METHOD: A systematic literature review was conducted in accordance with the PRISMA guidelines. The search strategy was limited to English language articles via PubMed, Embase and Web of Science databases up to June 30, 2023. Such articles provided diagnostic accuracy values of PR and/or PAN in the detection of AP lesions or alternatively data needed to calculate them. RESULTS: Twelve studies met inclusion criteria and were considered for the analysis. The average value of diagnostic accuracy in assessing AP lesions was 71% for PR and 66% for PAN. According to different accuracy for specific anatomical areas, it is recommended to use PR in the analysis of AP lesions located in the upper arch and lower incisor area, whereas lower premolar and molar areas may be investigated with the same accuracy with PR or PAN. CONCLUSIONS: Two-dimensional imaging must be considered the first-level examination for the diagnosis of AP lesions. PR had an overall slightly higher diagnostic accuracy than PAN. Evidence from this review provided a useful tool to support radiologists and dentists in their decision-making when inflammatory periapical bone lesions are suspected to achieve the best clinical outcome for patients, improving the quality of clinical practice.

Risk factors for impacted lower third molar root perforation through the undercut-shaped lingual plate: preoperative insights from panoramic radiography.

This study aimed to identify risk factors associated with perforation of the undercut (U)-shaped lingual plate (LP) by the lower third molar (LM3) root using panoramic radiography (PAN). We retrospectively examined 468 impacted LM3s from 468 individuals, categorizing LM3-LP associations and LP morphology in the coronal section of cone-beam computed tomography as perforation or nonperforation and U-type or non-U-type, respectively. The outcome was the combination of perforation and U-type, and study variables included patient demographics (age and sex) and PAN-associated features (Winter's classification, Pell-Gregory classification, and two major Rood signs). Multivariate logistic regression methods were used for analysis. Perforated and U-type LPs were observed in 205 (43.8%) and 212 (45.3%) cases, respectively. The double-positive outcome was observed in 126 LM3s (26.9%). In the multivariate model, age ≥ 26 years [odds ratio (OR), 2.66; p = 0.002], men (OR, 2.01; p = 0.002), mesioangular (OR, 2.74; p = 0.038) and horizontal impaction (OR, 3.05; p = 0.019), and root darkening (OR, 1.73; p = 0.039) were independently associated with the risk. Class III impaction (OR, 0.35; p = 0.021) and interruption of the white line (OR, 0.55; p = 0.017) were negatively correlated with the risk. In conclusion, this study highlights the importance of identifying the higher probability of U-type LP perforation by the LM3 root in men aged over their midtwenties with Class I/II impaction and mesioangularly or horizontally impacted LM3s, along with root darkening and no interruption of the white line on PAN.

Mandibular bone mass density in a medieval population and its relationship with stable isotopes δ13C and δ15N.

The aim of this study was to compare the level of bone mass in digital orthopantomograms in two populations (medieval and current) using two radiomorphometric indexes, and to correlate the mandibular bone mass value, in the medieval mandible population, with stable isotope data δ13C and δ15N. An observational, cross-sectional, and analytical study on mandibles from two diachronic groups, 15 mandibles from the medieval settlement of La Torrecilla (Granada, Spain) and 15 mandibles from current patients at the Faculty of Dentistry of the University of Granada (Spain), matched by age and sex was conducted. The bone mass density was determined using the Mandibular Cortical Width Index (MCW) and the Mandibular Panoramic Index (PMI) in digital panoramic radiographs. In the medieval group, the values of bone mass density were correlated with those of two stable isotopes (δ13C and δ15N). The mean value of MCW in mm in the medieval group was 3.96 ± 0.60 (mean ± standard deviation) and in the current group was 4.02 ± 1.01. The PMI was 0.33 ± 0.06 and 0.35 ± 0.08 in the medieval and current groups respectively, with similar results in both groups (p = 0.820 and p = 0.575). A negative correlation was found between both morphometric indices and the δ15N isotope (rs = 0.56, p = 0.030 and rs = 0.61, p = 0.016, respectively). The bone mass density in mandibles belonging to the two compared populations, determined by two quantitative radiomorphometric indices, is similar. Within the medieval population, there is an inverse correlation between the δ15N value and bone mass density.

Exploring the potential of artificial intelligence in paediatric dentistry: A systematic review on deep learning algorithms for dental anomaly detection.

BACKGROUND: Artificial intelligence (AI) based on deep learning (DL) algorithms has shown promise in enhancing the speed and accuracy of dental anomaly detection in paediatric dentistry. AIM: This systematic review aimed to investigate the performance of AI systems in identifying dental anomalies in paediatric dentistry and compare it with human performance. DESIGN: A systematic search of Scopus, PubMed and Google Scholar was conducted from 2012 to 2022. Inclusion criteria were based on problem/patient/population, intervention/indicator, comparison and outcome scheme and specific keywords related to AI, DL, paediatric dentistry, dental anomalies, supernumerary and mesiodens. Six of 3918 initial pool articles were included, assessing nine DL sub-systems that used panoramic radiographs or cone-beam computed tomography. Article quality was assessed using QUADAS-2. RESULTS: Artificial intelligence systems based on DL algorithms showed promising potential in enhancing the speed and accuracy of dental anomaly detection, with an average of 85.38% accuracy and 86.61% sensitivity. Human performance, however, outperformed AI systems, achieving 95% accuracy and 99% sensitivity. Limitations included a limited number of articles and data heterogeneity. CONCLUSION: The potential of AI systems employing DL algorithms is highlighted in detecting dental anomalies in paediatric dentistry. Further research is needed to address limitations, explore additional anomalies and establish the broader applicability of AI in paediatric dentistry.

Detection and classification of mandibular fractures in panoramic radiography using artificial intelligence.

OBJECTIVES: This study evaluated the performance of the YOLOv5 deep learning model in detecting different mandibular fracture types in panoramic images. METHODS: The dataset of panoramic radiographs with mandibular fractures was divided into training, validation, and testing sets, with 60%, 20%, and 20% of the images, respectively. An equal number of control images without fractures were also distributed among the datasets. The YOLOv5 algorithm was trained to detect six mandibular fracture types based on the anatomical location including symphysis, body, angle, ramus, condylar neck, and condylar head. Performance metrics of accuracy, precision, sensitivity (recall), specificity, dice coefficient (F1 score), and area under the curve (AUC) were calculated for each class. RESULTS: A total of 498 panoramic images containing 673 fractures were collected. The accuracy was highest in detecting body (96.21%) and symphysis (95.87%), and was lowest in angle (90.51%) fractures. The highest and lowest precision values were observed in detecting symphysis (95.45%) and condylar head (63.16%) fractures, respectively. The sensitivity was highest in the body (96.67%) fractures and was lowest in the condylar head (80.00%) and condylar neck (81.25%) fractures. The highest specificity was noted in symphysis (98.96%), body (96.08%), and ramus (96.04%) fractures, respectively. The dice coefficient and AUC were highest in detecting body fractures (0.921 and 0.942, respectively), and were lowest in detecting condylar head fractures (0.706 and 0.812, respectively). CONCLUSION: The trained algorithm achieved promising results in detecting most fracture types, particularly in body and symphysis regions indicating machine learning potential as a diagnostic aid for clinicians.

Comparison of quantitative radiomorphometric predictors of healthy and MRONJ-affected bone using panoramic radiography and cone-beam CT.

OBJECTIVES: To determine the most distinctive quantitative radiomorphometric parameter(s) for the detection of MRONJ-affected bone changes in panoramic radiography (PR) and cone-beam CT (CBCT). METHODS: PR and sagittal CBCT slices of 24 MRONJ patients and 22 healthy controls were used for the measurements of mandibular cortical thickness (MCT), fractal dimension (FD), lacunarity, mean gray value (MGV), bone area fraction (BA/TA), trabecular thickness (Tb.Th), trabecular separation (Tb.Sp), trabecular number (Tb.N). MCT was measured in the mental foramen region. While FD and lacunarity were measured on mandibular trabecular and cortical regions-of-interest (ROIs), the remaining parameters were measured on trabecular ROIs. The independent samples t-test was used to compare the measurements between the MRONJ and control groups for both imaging modalities (P = .05). RESULTS: MCT was the only parameter that differentiated MRONJ-affected bone in both PR and CBCT (P < .05). None of the remaining parameters revealed any difference for MRONJ-affected bone in CBCT (P > .05). FD, lacunarity, MGV, BA/TA, and Tb.Sp could distinguish MRONJ-affected trabecular bone in PR (P < .05). The correspondent ROI for both imaging methods that was reliable for detecting MRONJ-affected bone was the trabecular bone distal to the mental foramen above the inferior alveolar canal (ROI-3). CONCLUSIONS: MCT is a reliable parameter for the discrimination of MRONJ-affected bone in both PR and CBCT images. PR may be used to detect MRONJ-affected trabecular bone using FD, lacunarity, MGV, BA/TA, and Tb.Sp measurements as well.

An attempt to generate panoramic radiographs including jaw cysts using StyleGAN3.

OBJECTIVES: The purpose of this study was to generate radiographs including dentigerous cysts by applying the latest generative adversarial network (GAN; StyleGAN3) to panoramic radiography. METHODS: A total of 459 cystic lesions were selected, and 409 images were randomly assigned as training data and 50 images as test data. StyleGAN3 training was performed for 500 000 images. Fifty generated images were objectively evaluated by comparing them with 50 real images according to four metrics: Fréchet inception distance (FID), kernel inception distance (KID), precision and recall, and inception score (IS). A subjective evaluation of the generated images was performed by three specialists who compared them with the real images in a visual Turing test. RESULTS: The results of the metrics were as follows: FID, 199.28; KID, 0.14; precision, 0.0047; recall, 0.00; and IS, 2.48. The overall results of the visual Turing test were 82.3%. No significant difference was found in the human scoring of root resorption. CONCLUSIONS: The images generated by StyleGAN3 were of such high quality that specialists could not distinguish them from the real images.

DMAF-Net: deformable multi-scale adaptive fusion network for dental structure detection with panoramic radiographs.

OBJECTIVES: Panoramic radiography is one of the most commonly used diagnostic modalities in dentistry. Automatic recognition of panoramic radiography helps dentists in decision support. In order to improve the accuracy of the detection of dental structural problems in panoramic radiographs, we have improved the You Only Look Once (YOLO) network and verified the feasibility of this new method in aiding the detection of dental problems. METHODS: We propose a Deformable Multi-scale Adaptive Fusion Net (DMAF-Net) to detect 5 types of dental situations (impacted teeth, missing teeth, implants, crown restorations, and root canal-treated teeth) in panoramic radiography by improving the YOLO network. In DMAF-Net, we propose different modules to enhance the feature extraction capability of the network as well as to acquire high-level features at different scales, while using adaptively spatial feature fusion to solve the problem of scale mismatches of different feature layers, which effectively improves the detection performance. In order to evaluate the detection performance of the models, we compare the experimental results of different models in the test set and select the optimal results of the models by calculating the average of different metrics in each category as the evaluation criteria. RESULTS: About 1474 panoramic radiographs were divided into training, validation, and test sets in the ratio of 7:2:1. In the test set, the average precision and recall of DMAF-Net are 92.7% and 87.6%, respectively; the mean Average Precision (mAP0.5 and mAP[0.5:0.95]) are 91.8% and 63.7%, respectively. CONCLUSIONS: The proposed DMAF-Net model improves existing deep learning models and achieves automatic detection of tooth structure problems in panoramic radiographs. This new method has great potential for new computer-aided diagnostic, teaching, and clinical applications in the future.

Identification and classification of pulp calcifications in posterior teeth according to dental condition using digital panoramic radiography and cone beam CT.

OBJECTIVE: To compare digital panoramic radiography (DPR) and cone beam CT (CBCT) in the detection and classification of pulp calcifications in posterior teeth in relation to tooth condition and its location. METHODS: Two hundred and fifty patients with simultaneous DPR and CBCT scans were selected from a dental image bank. For each posterior tooth, its condition was registered (healthy, restored, or decayed). The presence of calcifications was assessed and classified according to location (coronal or radicular) and type (total obliteration, partial obliteration, narrowing, and no calcification). Chi-square and McNemar tests were used to compare DPR and CBCT results, with a significance level of 5%. DPR diagnostic values were calculated using CBCT as reference. RESULTS: Decayed and restored teeth showed a significantly (P < .001) higher prevalence of pulp calcifications than healthy teeth in both imaging exams. There was a significant discrepancy between the imaging modalities in the identification of calcifications (P < .001), especially for radicular calcifications of maxillary teeth (DPR = 55.2%; CBCT = 30.9%). In the case of coronal calcifications, there was a greater discrepancy between CBCT and DPR in the mandibular teeth (10.7%) than in the maxillary teeth (7.7%). Overall, DPR presents higher sensitivity (0.855) than specificity (0.483) for pulp calcifications detection. CONCLUSION: DPR considerably overestimates pulp calcifications in posterior teeth, as compared to CBCT. A higher prevalence of pulp calcifications is related to decayed and restored teeth.

Detection of periodontal bone loss patterns and furcation defects from panoramic radiographs using deep learning algorithm: a retrospective study.

BACKGROUND: This retrospective study aimed to develop a deep learning algorithm for the interpretation of panoramic radiographs and to examine the performance of this algorithm in the detection of periodontal bone losses and bone loss patterns. METHODS: A total of 1121 panoramic radiographs were used in this study. Bone losses in the maxilla and mandibula (total alveolar bone loss) (n = 2251), interdental bone losses (n = 25303), and furcation defects (n = 2815) were labeled using the segmentation method. In addition, interdental bone losses were divided into horizontal (n = 21839) and vertical (n = 3464) bone losses according to the defect patterns. A Convolutional Neural Network (CNN)-based artificial intelligence (AI) system was developed using U-Net architecture. The performance of the deep learning algorithm was statistically evaluated by the confusion matrix and ROC curve analysis. RESULTS: The system showed the highest diagnostic performance in the detection of total alveolar bone losses (AUC = 0.951) and the lowest in the detection of vertical bone losses (AUC = 0.733). The sensitivity, precision, F1 score, accuracy, and AUC values were found as 1, 0.995, 0.997, 0.994, 0.951 for total alveolar bone loss; found as 0.947, 0.939, 0.943, 0.892, 0.910 for horizontal bone losses; found as 0.558, 0.846, 0.673, 0.506, 0.733 for vertical bone losses and found as 0.892, 0.933, 0.912, 0.837, 0.868 for furcation defects (respectively). CONCLUSIONS: AI systems offer promising results in determining periodontal bone loss patterns and furcation defects from dental radiographs. This suggests that CNN algorithms can also be used to provide more detailed information such as automatic determination of periodontal disease severity and treatment planning in various dental radiographs.

Evaluation of mandibular trabecular bone by fractal analysis in pediatric patients with hypodontia of the mandibular second premolar tooth.

BACKGROUND: It is still unclear whether the trabecular structure of the jaw is different in individuals with hypodontia than in those without hypodontia; this is important for clinicians. The aim was to determine whether the mandibular trabecular bone structure of children and adolescents with hypodontia differs from the control group by using the fractal analysis (FA) method in this study. METHODS: A total of 138 panoramic radiographs of 69 cases and 69 control subjects (mean age 13.2 ± 10.1) were evaluated. The age and gender of subjects in the case and control groups were matched. Three regions of interest (ROIs) were selected from the panoramic radiographs. ROI1 refers to the center of the ramus rising above the mandibular foramen. ROI2 refers to the area between the apical level of the mandibular molar and the upper border of the mandibular canal. ROI3, the missing tooth region, refers to the apical third of the mesial side of the erupting or fully erupted permanent mandibular first molar. Mann-Whitney U and Wilcoxon tests were used. p < 0.05 was accepted for the significance value. RESULTS: The mean fractal dimension (FD) values of ROI1, ROI2, and ROI3 were 1,25, 1,20, and 1,13, respectively. The means FD values obtained from the ramus region were higher than the other regions (p < 0.05). The FD values did not differ significantly according to gender and age (p > 0.05). The FD values of the case group were lower than the control group for ROI3 (p < 0.05). CONCLUSION: The results of this study showed that the mandibular trabecular bone quality of pediatric patients with one missing tooth was different from the healthy group. The difference in the mean FD values from the ROIs indicates that the ramus has a denser structure than the mandibular corpus. Clinicians should factor this into their dental treatment planning process.