Oral Radiology最新文献

Objectives: The goal of this examination was to compare the impact of probable sleep/awake bruxism on the mandibular trabecular bone structure by fractal analysis (FA) with digital panoramic radiograph (DPR) and radiomorphometric measurements in paediatric patients with bruxism.

Methods: The examination included 130 participants with 63 patients with probable sleep/awake bruxism and 67 control groups. Bilateral regions of interest (ROI) in three regions were examined as ROI1: mandibular ramus, ROI2: mandibular angulus, ROI3: anterior to the molar teeth. Radiomorphometric measurements were taken of the mandibular cortical width (MCW), panoramic mandibular index (PMI), and mandibular cortical index (MCI). p < 0.05 was approved for statistical significance.

Results: The ROI-1, ROI-2, and ROI-3 values were defined to be statistically significantly high in the bruxism group (p < 0.05). No significant difference was found between the groups in the other values (p > 0.05). There was no difference in the age and gender for any of the parameters (p > 0.05).

Conclusion: In children and adolescents, the mandibular trabecular bone can be affected by bruxism. FA can be used as an auxiliary method for finding the mandibular trabecular differences of patients with bruxism in paediatric dentistry just as it can for adults.

Objective: To assess quantitatively the effect of metallic materials on MR image uniformity using a standardized method.

Methods: Six types of 1 cm cubic metallic materials (i.e., Au, Ag, Al, Au-Ag-Pd alloy, Ti, and Co-Cr alloy) embedded in a glass phantom filled were examined and compared with no metal condition inserted as a reference. The phantom was scanned five times under each condition using a 1.5-T MR superconducting magnet scanner with an 8-channel phased-array brain coil and head and neck coil. For each examination, the phantom was scanned in three planes: axial, coronal, and sagittal using T1-weighted spin echo (SE) and gradient echo (GRE) sequences in accordance with the American Society for Testing and Materials (ASTM) F2119-07 standard. Image uniformity was assessed using the non-uniformity index (NUI), which was developed by the National Electrical Manufacturers Association (NEMA), as an appropriate standardized measure for investigating magnetic field uniformity.

Results: T1-GRE images with Co-Cr typically elicited the lowest uniformity, followed by T1-GRE images with Ti, while all other metallic materials did not affect image uniformity. In particular, T1-GRE images with Co-Cr showed significantly higher NUI values as far as 6.6 cm at maximum equivalent to 11 slices centering around it in comparison with the measurement uncertainty from images without metallic materials.

Conclusion: We found that MR image uniformity was influenced by the scanning sequence and coil type when Co-Cr and Ti were present. It is assumed that the image non-uniformity in Co-Cr and Ti is caused by their high magnetic susceptibility.

Objectives: This study aimed to train a 3D U-Net convolutional neural network (CNN) for mandible and lower dentition segmentation from cone-beam computed tomography (CBCT) scans.

Methods: In an ambispective cross-sectional design, CBCT scans from two hospitals (2009-2019 and 2021-2022) constituted an internal dataset and external validation set, respectively. Manual segmentation informed CNN training, and evaluations employed Dice similarity coefficient (DSC) for volumetric accuracy. A blinded oral maxillofacial surgeon performed qualitative grading of CBCT scans and object meshes. Statistical analyses included independent t-tests and ANOVA tests to compare DSC across patient subgroups of gender, race, body mass index (BMI), test dataset used, age, and degree of metal artifact. Tests were powered for a minimum detectable difference in DSC of 0.025, with alpha of 0.05 and power level of 0.8.

Results: 648 CBCT scans from 490 patients were included in the study. The CNN achieved high accuracy (average DSC: 0.945 internal, 0.940 external). No DSC differences were observed between test set used, gender, BMI, and race. Significant differences in DSC were identified based on age group and the degree of metal artifact. The majority (80%) of object meshes produced by both manual and automatic segmentation were rated as acceptable or higher quality.

Conclusion: We developed a model for automatic mandible and lower dentition segmentation from CBCT scans in a demographically diverse cohort including a high degree of metal artifacts. The model demonstrated good accuracy on internal and external test sets, with majority acceptable quality from a clinical grader.

Objective: The aim of this study is to determine the contact relationship and position of impacted mandibular third molar teeth (IMM₃) with the mandibular canal (MC) in panoramic radiography (PR) images using deep learning (DL) models trained with the help of cone beam computed tomography (CBCT) and DL to compare the performances of the architectures.

Methods: In this study, a total of 546 IMM₃s from 290 patients with CBCT and PR images were included. The performances of SqueezeNet, GoogLeNet, and Inception-v3 architectures in solving four problems on two different regions of interest (RoI) were evaluated.

Results: The SqueezeNet architecture performed the best on the vertical RoI, showing 93.2% accuracy in the identification of the 2nd problem (contact relationship buccal or lingual). Inception-v3 showed the highest performance with 84.8% accuracy in horizontal RoI for the 1st problem (contact relationship-no contact relationship), GoogLeNet showed 77.4% accuracy in horizontal RoI for the 4th problem (contact relationship buccal, lingual, other category, or no contact relationship), and GoogLeNet showed 70.0% accuracy in horizontal RoI for the 3rd problem (contact relationship buccal, lingual, or other category).

Conclusion: This study found that the Inception-v3 model showed the highest accuracy values in determining the contact relationship, and SqueezeNet architecture showed the highest accuracy values in determining the position of IMM₃ relative to MC in the presence of a contact relationship.

Purpose: This study aimed to evaluate odontogenic keratocysts associated with basal cell nevus syndrome (BCNS) using computed tomography (CT) and magnetic resornance imaging (MRI) including diffusion-weighted imaging (DWI) and compare them with sporadic cases.

Materials and methods: This study investigated 17 outpatients who underwent panoramic radiography, CT, and MRI between August 2012 and January 2021. Five of these patients had BCNS had 16 odontogenic keratocysts, for which the authors recorded detailed findings. DWI analysis compared the apparent diffusion coefficient (ADC) values of odontogenic keratocysts in patients with BCNS and sporadic cases. The Mann-Whitney test was used to analyse bivariate statistics.

Results: Patients with BCNS had an average of 3.2 lesions in the jaw. On DWI, the ADC value ranged from 0.58 to 2.66 × 10^-3 mm²/s. The values for sporadic odontogenic keratocysts ranged from 0.67 to 1.11 × 10^-3 mm²/s. The median values were 0.94 and 0.89 for BCNS-associated and sporadic odontogenic keratocysts cases, respectively (P = .478).

Conclusion: This study presented detailed imaging findings of odontogenic keratocysts in patients with BCNS. Furthermore, the authors revealed a wide range of ADC values for BCNS-associated odontogenic keratocysts.

Objectives: To evaluate previously reported quantitative (tumor thickness 11 mm and depth of invasion [DOI] 7.5 mm) and qualitative (styloglossus/hyoglossus muscle invasion [SHMI]) magnetic resonance imaging (MRI) parameters for predicting occult neck node metastasis in clinical N0 oral tongue squamous cell carcinoma.

Methods: This single-center retrospective study included 76 patients. MRI images were independently reviewed by two radiologists for tumor thickness, DOI, and SHMI. Statistical analysis assessed the predictive capability of these parameters for 2-year potential lymph node metastasis.

Results: Among the 76 cases, 30.2% developed 2-year potential lymph node metastasis. For tumor thickness ≥ 11 mm, the sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy were 0.46, 0.68, 0.37, 0.75, and 0.61, respectively. DOI ≥ 7.5 mm exhibited a sensitivity, specificity, PPV, NPV, and accuracy of 0.73, 0.59, 0.42, 0.84, and 0.63, respectively. SHMI demonstrated a sensitivity, specificity, PPV, NPV, and accuracy of 0.87, 0.51, 0.46, 0.89, and 0.63, respectively.

Conclusion: DOI ≥ 7.5 mm and SHMI demonstrated comparable diagnostic accuracy in predicting neck metastasis, surpassing tumor thickness of > 11 mm. These findings underscore their potential utility in guiding decisions concerning elective neck dissection.

Objective: The horizontal tube-shifting technique can be adopted to separate overlapping buccal roots of the maxillary molar from the palatal root. A simulation study was performed to determine an appropriate tube-shift angulation when adopting three-dimensional computed tomography imaging.

Methods: Cone-beam computed tomography images of 21 volunteers were used for simulation. Adopting image analysis software, maximum intensity projection (MIP) three-dimensional images of maxillary molars observed from directions similar to the direction of an intraoral radiography X-ray beam were reconstructed. The orthoradial projection angle was taken as the baseline and the tube-shifting technique was then simulated to separate superimposed buccal roots from the palatal root.

Results and conclusion: The tube-shift technique was considered applicable to maxillary molars for 29/42 teeth (69%) in the case of a mesial tube shift and 40/42 teeth (95.2%) in the case of a distal tube shift. The specific shifting angle suitable for observing the buccal root apex of the maxillary molar without overlapping was obtained.

Objective: The mandibular canal (MC) houses the inferior alveolar nerve. Extraction of the mandibular third molar (MM3) is a common dental surgery, often complicated by nerve damage. CBCT is the most effective imaging method to assess the relationship between MM3 and MC. With advancements in artificial intelligence, deep learning has shown promising results in dentistry. The aim of this study is to evaluate the MC-MM3 relationship using CBCT and a deep learning technique, as well as to automatically segment the mandibular impacted third molar, mandibular canal, mental and mandibular foramen.

Methods: This retrospective study analyzed CBCT data from 300 patients. Segmentation was used for labeling, dividing the data into training (n = 270) and test (n = 30) sets. The nnU-NetV2 architecture was employed to develop an optimal deep learning model. The model's success was validated using the test set, with metrics including accuracy, sensitivity, precision, Dice score, Jaccard index, and AUC.

Results: For the MM3 annotated on CBCT, the accuracy was 0.99, sensitivity 0.90, precision 0.85, Dice score 0.85, Jaccard index 0.78, AUC value 0.95. In MC evaluation, accuracy was 0.99, sensitivity 0.75, precision 0.78, Dice score 0.76, Jaccard index 0.62, AUC value 0.88. For the evaluation of mental foramen; accuracy 0.99, sensitivity 0.64, precision 0.66, Dice score 0.64, Jaccard index 0.57, AUC value 0.82. In the evaluation of mandibular foramen, accuracy was found to be 0.99, sensitivity 0.79, precision 0.68, Dice score 0.71, and AUC value 0.90. Evaluating the MM3-MC relationship, the model showed an 80% correlation with observer assessments.

Conclusion: The nnU-NetV2 deep learning architecture reliably identifies the MC-MM3 relationship in CBCT images, aiding in diagnosis, surgical planning, and complication prediction.

Objectives: To explore the effectiveness of radiographic biomarkers on transition area (TA)-the grayscale gradient zone from carious lesion to normal dentine on radiographs-for identifying deep caries/reversible pulpitis and chronic pulpitis via diagnostic model analysis.

Methods: This retrospective study included 392 caries cases. Canny edge detection was used to define the TA region. Texture parameters were extracted from the carious lesions (S1) and TA region (S2) by MaZda software on radiographs. Least absolute shrinkage and selection operator (LASSO) regression analysis was used to select biomarkers. Diagnostic models were fitted and model performance was furtherly evaluated by internal and external validation, decision curve analysis was applied to evaluate clinical benefits.

Results: TA-based biomarkers (e.g., TA thickness, TA ratio, S2-S(5,-5) contrast and S2-WavEnLL-s-4) were significantly associated with the diagnosis of deep caries/reversible pulpitis versus chronic pulpitis, model performance significantly improved when adding the above biomarkers (likelihood-ratio test; p < 0.05, with an increase of AUC from 0.67 (reference model) to 0.89), and these results were maintained in a small external validation cohort. Clinical benefit was greater with the application of TA-based biomarkers.

Conclusion: TA-based biomarkers are proven to be an effective tool in differentiating deep caries/reversible pulpitis and chronic pulpitis, preoperative diagnosis was improved with the above biomarkers compared to the reference model.

Objectives: Deep learning has revolutionized image analysis for dentistry. Automated segmentation of dental radiographs is of great importance towards digital dentistry. The performance of deep learning models heavily relies on the quality and diversity of the training data. Data augmentation is a widely used technique implemented in machine learning and deep learning to artificially increase the size and diversity of a training dataset by applying various transformations to the original data.

Methods: This work aims to automatically segment implants, prostheses, and fillings in panoramic images using 9 different deep learning segmentation models. Later, it explores the effect of data augmentation methods on segmentation performance of the models. Eight different data augmentation techniques are examined. Performance is evaluated by well-accepted metrics such as intersection over union (IoU) and Dice coefficient.

Results: While averaging the segmentation results for the three classes, IoU varies between 0.62 and 0.82 while Dice score is between 0.75 and 0.9 among deep learning models used. Augmentation techniques provided performance improvements of up to 3.37%, 5.75% and 8.75% for implant, prosthesis and filling classes, respectively.

Conclusions: Findings reveal that choosing optimal augmentation strategies depends on both model architecture and dental structure type.