Imaging Science in Dentistry最新文献

Purpose: This study was conducted to develop and evaluate a deep learning-based super-resolution approach for enhancing the quality of cone-beam computed tomography (CBCT) images in dentomaxillofacial imaging.

Materials and methods: A deep learning-based super-resolution method using the MIRNet-v2 model was developed to enhance CBCT image quality. The study used a dataset comprising 6,961 anonymized axial slices from 15 CBCT scans. High-resolution images served as ground truth, while low-resolution versions were created through artificial degradation, including downscaling, blurring, and noise addition. The model was evaluated using a 5-fold cross-validation strategy, employing peak signal-to-noise ratio (PSNR) and structural similarity index measure (SSIM) as metrics. Qualitative assessments conducted by 2 experienced radiologists involved criteria such as noise, sharpness, spatial resolution, and diagnostic quality, scored using a CBCT evaluation chart.

Results: The model significantly improved degraded CBCT images across all evaluation metrics. Enhanced images demonstrated mean PSNR values exceeding 35 dB and SSIM values over 0.85, with the highest performance achieved for blurred images (PSNR: 43.86±1.61, SSIM: 0.98±0.01). Subjective assessments indicated improvements in diagnostic quality, noise reduction, and spatial resolution, with outputs comparable to the original images in several degradation scenarios. Interobserver reliability was fair (Cohen kappa: 0.335). Notable improvements were observed for noise and artifact reduction in specific degradation groups, suggesting improved diagnostic utility.

Conclusion: Deep learning-based super-resolution demonstrates considerable potential for enhancing CBCT image quality, especially in scenarios involving blur and downscaling. These results suggest possible applications in low-dose imaging protocols and improved clinical decision-making.

Purpose: Temporomandibular joint osteoarthritis (TMJOA) is a significant subtype of temporomandibular joint disorders (TMDs). The purpose of this study was to comprehensively summarize the current literature on the use of artificial intelligence (AI) technologies in the diagnosis and management of TMJOA using cone-beam computed tomography (CBCT).

Materials and methods: This systematic review was pre-registered in the PROSPERO database (PROSPERO CRD42024509772). Up to December 2023, research was conducted using Google Scholar, Embase, MEDLINE, and Web of Science databases to identify studies evaluating the use of AI technologies in the management and diagnosis of TMJOA via CBCT. The search strategy included MeSH terms, keywords, and their combinations. Risk of bias was assessed using the ROBINS-I tool.

Results: Out of 2,543 articles retrieved, a total of 9 studies were included in this systematic review. All included studies were observational and employed AI models based on convolutional neural networks, including SVA, SSD, LightGBM, XGBoost, and YOLO. The performance of these models varied, with accuracy ranging from 73.5% to 99% and F1-scores between 0.80 and 0.86. Among these, YOLO demonstrated the highest accuracy for the assessment and diagnosis of TMJOA using CBCT scans.

Conclusion: AI algorithms developed for the automated diagnosis of TMJOA can be utilized by clinicians as decision-support tools. Incorporating diverse input data types, such as electronic medical records, radiomics features, and biomarkers, alongside diagnostic imaging may further increase the diagnostic accuracy for TMDs.

Purpose: This retrospective study aimed to evaluate the accuracy of facial scan (FS) to cone-beam computed tomography (CBCT) registration by comparing superimpositions on full-cranium and reduced field-of-view (FOV) CBCT, with the goal of assessing its potential to reduce radiation exposure without compromising diagnostic quality.

Materials and methods: CBCT scans from 50 patients were analyzed, integrating FS data obtained via 3D laser scanning. FSs were registered to both full-cranium and reduced FOV CBCT using landmark-based matching and a best-fit algorithm. Accuracy was evaluated by calculating the point-to-point surface distance between FS and CBCT soft-tissue renderings. The metrics used were root mean square distance (RMSD), Hausdorff distance (HD), and median distance (MD). Registration of FS onto full FOV CBCT served as the ground truth. Statistical analysis employed the Mann-Whitney U test to compare registration performance on both the overall surface and the facial midline.

Results: There was no significant difference in HD (P=0.288) between the 2 methods. However, median RMSD and MD were significantly lower for full-cranium CBCT (P=0.019). Midline alignment between FS and reduced FOV CBCT showed no visual discrepancies, with an MD of 0.35 mm along the midsagittal plane.

Conclusion: FS registration to reduced FOV CBCT provides clinically acceptable accuracy, particularly in the midline region, while substantially reducing radiation exposure. This approach is promising for a range of dental applications, especially in pediatric cases and situations prioritizing facial aesthetics. Further research is warranted to optimize this technique for diverse clinical contexts.

Purpose: Knowledge of the angular artery (AA) anatomy in the nasolabial fold is essential to prevent necrosis during esthetic procedures. This study aimed to assess the depth and hemodynamics of the AA in the pyriform space using 2-dimensional ultrasonography (USG), as well as color and spectral Doppler imaging.

Materials and methods: This study evaluated AA diameter, epidermis-AA distance (EP-AA), AA-bone distance (AA-BO), systolic velocity, diastolic velocity, pulsatility index, and resistance index. These parameters were analyzed in relation to independent variables, including age, sex, and body mass index (BMI). The study sample consisted of 58 Brazilian participants, with both hemifaces independently assessed by 2 trained examiners. Statistical analyses of the variables were performed with a significance level of 5%.

Results: The EP-AA distance was greater in men (5.8 mm) than in women (4.8 mm, P<0.05). BMI was also found to significantly influence the EP-AA distance (P<0.05). No independent variable affected the AA-BO distance. With respect to AA diameter, only age demonstrated a significant influence (P<0.05), with the diameter decreasing as age increased. Similarly, only age was associated with diastolic and systolic velocities (P<0.05), with each additional year corresponding to a reduction of 0.13 cm/s in diastolic velocity and 0.28 cm/s in systolic velocity.

Conclusion: USG demonstrated that AA depth in relation to the epidermis is greater in men and in individuals with higher BMI. Aging is associated with a reduction in both the diameter of the AA and its diastolic and systolic velocities.

Purpose: This study was performed to assess the relationship of the third molars with the mandibular canal as a predictor of inferior alveolar nerve (IAN) sensory disturbances using panoramic radiography (PR) and cone-beam computed tomography (CBCT).

Materials and methods: A systematic search was conducted of 4 databases-PubMed, Scopus, Web of Science, and Google Scholar-for the period from 1985 to 2024. In the retrieved articles, the outcome of interest was the relationship of the mandibular canal with the third molars on PR and CBCT scans. The risk of bias was assessed using the Newcastle-Ottawa Scale, and quantitative meta-analysis was performed using STATA. A random-effects restricted maximum likelihood model was employed for the meta-analysis, and the I² statistic was used to assess heterogeneity.

Results: A total of 1,635 articles were initially retrieved. After a rigorous selection process, 20 studies were included in the qualitative synthesis, and 8 were selected for the meta-analysis. The findings indicated that CBCT yielded higher prevalence rates for root darkening, root deflection, interruption of the white line, diversion of the mandibular canal, and narrowing of the mandibular canal (theta values: 49.962, 4.76, 8.09, 2.229, and 4.708, respectively) compared with PR (theta values: 1.363, 1.605, 6.322, 0.655, and 1.449, respectively).

Conclusion: CBCT was more accurate than PR in investigating predictors of IAN paresthesia in mandibular third molar surgery. Considering the higher prevalence of paresthesia in the presence of root darkening, CBCT may be highly efficient in detecting this parameter and thus aiding in the prevention of paresthesia.

Purpose: This study aimed to review, evaluate, and synthesize existing evidence on the effectiveness of fractal analysis (FA) in assessing bone formation in periapical lesions following endodontic treatment.

Materials and methods: Two reviewers systematically searched 6 electronic databases and gray literature. Studies were deemed eligible if they implemented the desired intervention and included a follow-up period of at least 12 months. Methodological quality was assessed using tools from the Joanna Briggs Institute. The meta-analysis calculated the mean difference (MD) in FA measurements of periapical lesion regions before and after endodontic treatment, with subgroup analyses based on bone and treatment type. The GRADE tool was employed to evaluate the certainty of the evidence.

Results: Ten studies were included for qualitative synthesis and 8 in the meta-analysis. Overall, the mean fractal dimension (FD) increased following 12 months of endodontic treatment, with an MD of 0.223 (95% CI: 0.100-0.346; P<0.001; I²=99%). Subgroup analyses revealed significantly increases in mean FD values for lesions in the maxilla (P<0.01) and for the treatment subgroup (P<0.01). However, the certainty of evidence was classified as very low.

Conclusion: The observed increase in mean FD 12 months post-endodontic treatment across all included studies indicates bone formation in the periapical lesion regions.

Purpose: This study aimed to evaluate the correlation between incidental vascular calcification-like imaging findings and self-reported medical data, as well as to assess the relationship between reported predisposing factors and imaging findings using cone-beam computed tomography (CBCT) data.

Materials and methods: A total of 391 CBCT scans from 188 males and 203 females were anonymously analyzed for the presence of extra- and intra-cranial carotid artery calcifications (ECAC and ICAC, respectively) and signs of Mönckeberg medial sclerosis (MMS). The patients were categorized into 4 groups based on their self-reported medical histories. Descriptive statistics were used to evaluate the data, which were subsequently validated through simple univariate logistic regression analysis.

Results: Among the 391 CBCT scans reviewed, 23.27% exhibited ECAC, 42.71% demonstrated ICAC, and 1.8% showed MMS. Statistical analysis revealed a significant correlation (P<0.05) between both ECAC and ICAC and self-reported predisposing factors-including hypertension, cardiovascular disease, dyslipidemia, diabetes mellitus, and sleep apnea/chronic obstructive pulmonary disease-with notable differences among the study categories (P<0.05). In addition, a strong correlation (P<0.001) was found between the presence of ECAC, ICAC, and MMS and increasing age. Men were significantly more susceptible to ECAC than women (P<0.05).

Conclusion: These findings underscore the importance of a thorough pre-treatment medical history assessment in dental patients, particularly when vascular calcification-like signs are observed on CBCT imaging.

Purpose: Panoramic radiographs are instrumental in dental diagnosis but face quality issues related to contrast, artifacts, positioning, and coverage, which can impact diagnostic accuracy. Although expert assessment is the accepted standard, it is time-consuming and prone to inconsistency. Artificial intelligence offers an automated, objective solution for evaluating radiograph quality, increasing efficiency and reducing inter-rater variability.

Materials and methods: This study aimed to develop a deep learning (DL)-based model for evaluating the quality of dental panoramic radiographs. A dataset of 1,000 panoramic images, collected from 2018 to 2023, was assessed by 2 trained dentists using predefined grading criteria for contrast/density, artifact presence, coverage area, patient positioning, and overall quality. These expert-annotated scores were used as the ground truth to train and validate 5 YOLOv8 classification models, each targeting a specific quality criterion. The models' performance was evaluated on a separate test set using performance metrics.

Results: The YOLOv8 models achieved classification accuracies of 87.2%, 74.1%, 77.3%, 97.9%, and 79.3% for artifact detection, coverage area, patient positioning, contrast/density, and overall image quality, respectively. The model used to classify images as clinically acceptable or unacceptable exhibited an average accuracy of 81.4%, demonstrating its potential for real-world application.

Conclusion: These findings highlight the feasibility of DL-based automated image quality assessment for panoramic radiographs. The high accuracy of the proposed model suggests its potential integration into clinical workflows to assist practitioners in efficiently evaluating radiograph quality. Additionally, such a model could represent an educational tool for dental students, improving radiographic techniques and reducing unnecessary retakes.

Purpose: Panoramic radiographs have recently become a platform for deep learning models, which show potential in enhancing diagnostic accuracy for detecting contact between mandibular third molars and the mandibular canal. However, detailed information regarding the accuracy of these models in identifying such contact remains limited.

Materials and methods: In accordance with the PRISMA-2020 and PRISMA-DTA guidelines, the PubMed, ScienceDirect, Web of Science, Embase, and EBSCO databases were systematically searched up to September 2024. Eligible studies employed deep learning models based on convolutional neural networks to classify the contact between mandibular third molars and the mandibular canal. Extracted metrics included accuracy, sensitivity, specificity, precision, and F1-score. A meta-analysis using random effects models pooled these performance metrics, while univariate and multivariate meta-regressions were conducted to explore sources of heterogeneity. Study quality was assessed using the QUADAS-2 tool.

Results: Seven studies incorporating 4,955 panoramic radiographs reported pooled performance metrics of 83.4% accuracy, 80.2% sensitivity, 85.8% specificity, 83.3% precision, and an F1-score of 80.9%. High heterogeneity (I² > 90%) was primarily attributable to variations in sample size, image resolution, model architecture, and model complexity. Meta-regression analyses identified image resolution and architecture (e.g., VGG-16, AlexNet) as key factors. Although the overall risk of bias was low, the patient selection domain was often unclear.

Conclusion: Deep learning models exhibit significant promise in evaluating mandibular third molar and mandibular canal contact on panoramic radiographs, potentially complementing traditional methods. The adoption of standardized protocols, diverse datasets, and explainable artificial intelligence will be crucial for broader clinical application.

Purpose: This study employs bibliometric analysis to evaluate research trends, key contributors, and applications of artificial intelligence (AI) models in orthodontic imaging. It highlights the impact and evolution of AI in this field from 1991 to 2024.

Material and methods: A total of 130 documents were extracted from the Scopus database, spanning 33 years of research. The analysis examined annual growth rates, citation metrics, AI model adoption, and international collaborations. Network visualization was performed using VOSviewer to map research trends and co-authorship networks.

Results: The study analyzed 96 publications from 47 sources, revealing exponential growth in AI research-particularly after 2010, with a peak in 2023. The findings show a steady annual growth rate of 9.66% and a maximum citation count of 138 for an AI-based cephalometric analysis study. Convolutional neural networks (CNNs) and artificial neural networks (ANNs) dominate AI applications in orthodontic image analysis. An h-index of 23 and a g-index of 38 reflect the field's significant research impact. Strong international collaborations were observed, with 28.12% of studies involving cross-border research.

Conclusion: This analysis highlights the growing influence of AI in orthodontic imaging and emphasizes the need for larger datasets, improved model interpretability, and seamless clinical integration. Addressing these challenges will further enhance AI-driven diagnostics and treatment planning, guiding future research and broader clinical applications.