Orhan K, et al. Imaging Sci Dent. 2023 Aug;53:e26. https://doi.org/10.5624/isd.20230109

Cone-beam computed tomography (CBCT) is used to provide multiplanar views of the temporomandibular joint (TMJ) bone components as well as TMJ pathologies without superposition, magnification or distortion. The study aimed to analyze degenerative changes in the condylar surface, and their relationship with patient age and gender, and the TMJ space measurements by using CBCT images. A total of 258 individuals were retrospectively analyzed. The degenerative bone changes of the condylar head were evaluated and classified on the right and left sides. The shortest distances from the anterior, superior and posterior parts of the condylar head to the glenoid fossa were measured to represent the TMJ space. Univariate and multivariate logistic regression analyses then evaluated the effect of age and gender on the presence of degenerative changes. Condylar flattening was most frequently observed (413 TMJs, 53.5%). However, the presence or absence of the change types did not differ according to the sides. The mean values of the TMJ space measurements on the right and left sides were narrower in the group with changes than in the group without changes. Nonetheless, no statistically significant difference in the TMJ space was found between the groups (p > 0.05). An increased risk of radiographically detectable degenerative alterations in left TMJs was detected for males and for increasing age. Degenerative changes in the condylar surface may affect the dimensions of the TMJ space.

Background: The aim of our study is to provide successful automatic segmentation of various objects on orthopantomographs (OPGs). Methods: 8138 OPGs obtained from the archives of the Department of Dentomaxillofacial Radiology were included. OPGs were converted into PNGs and transferred to the segmentation tool’s database. All teeth, crown–bridge restorations, dental implants, composite–amalgam fillings, dental caries, residual roots, and root canal fillings were manually segmented by two experts with the manual drawing semantic segmentation technique. Results: The intra-class correlation coefficient (ICC) for both inter- and intra-observers for manual segmentation was excellent (ICC > 0.75). The intra-observer ICC was found to be 0.994, while the inter-observer reliability was 0.989. No significant difference was detected amongst observers (p = 0.947). The calculated DSC and accuracy values across all OPGs were 0.85 and 0.95 for the tooth segmentation, 0.88 and 0.99 for dental caries, 0.87 and 0.99 for dental restorations, 0.93 and 0.99 for crown–bridge restorations, 0.94 and 0.99 for dental implants, 0.78 and 0.99 for root canal fillings, and 0.78 and 0.99 for residual roots, respectively. Conclusions: Thanks to faster and automated diagnoses on 2D as well as 3D dental images, dentists will have higher diagnosis rates in a shorter time even without excluding cases.

The study aimed to generate a fused deep learning algorithm that detects and classifies the relationship between the mandibular third molar and mandibular canal on orthopantomographs. Radiographs (n = 1880) were randomly selected from the hospital archive. Two dentomaxillofacial radiologists annotated the data via MATLAB and classified them into four groups according to the overlap of the root of the mandibular third molar and mandibular canal. Each radiograph was segmented using a U-Net-like architecture. The segmented images were classified by AlexNet. Accuracy, the weighted intersection over union score, the dice coefficient, specificity, sensitivity, and area under curve metrics were used to quantify the performance of the models. Also, three dental practitioners were asked to classify the same test data, their success rate was assessed using the Intraclass Correlation Coefficient. The segmentation network achieved a global accuracy of 0.99 and a weighted intersection over union score of 0.98, average dice score overall images was 0.91. The classification network achieved an accuracy of 0.80, per class sensitivity of 0.74, 0.83, 0.86, 0.67, per class specificity of 0.92, 0.95, 0.88, 0.96 and AUC score of 0.85. The most successful dental practitioner achieved a success rate of 0.79. The fused segmentation and classification networks produced encouraging results. The final model achieved almost the same classification performance as dental practitioners. Better diagnostic accuracy of the combined artificial intelligence tools may help to improve the prediction of the risk factors, especially for recognizing such anatomical variations.

Background: Genetic and environmental factors especially climatic conditions are thought to influence the shape and size of the paranasal sinuses and anatomic variations may create both a diagnostic and therapeutic challenge. However, no study has been published about the climatic adaptation of the paranasal sinus region in different populations. This study aimed to compare the prevalence of anatomical variants in the paranasal sinus and nasal cavity using Cone-Beam Computed Tomography (CBCT) between Polish and Turkish Cypriot populations. Methods: The material consisted of volumes acquired utilizing Galileos (Sirona, Germany) as well as Newtom 3G (QR Verona, Newtom, Italy) CBCT units. There were examined 356 Polish and 359 Turkish Cypriot patients in whom paranasal sinuses were included in the field of view. Paranasal sinus anatomic variations were assessed in both populations. Results: In the Polish population, the most common anatomic variation was septum deviation followed by the Agger nasi cell and concha bullosa with a prevalence of 87.7%, 83.2%, and 54.8% respectively. For the Turkish Cypriot population, the most common anatomic variation was Agger nasi cell followed by concha bullosa and supraorbital ethmoid cells with a prevalence of 81.6%, 68%, and 57.8% respectively. Many anatomic variations were found to show substantial differences among both populations. Incidence rates of hyperpneumatization of the frontal sinus, septum pneumatization, supraorbital ethmoid cells, concha bullosa, uncinate bulla, and internal carotid artery protrusion into the sphenoid sinus were significantly higher in the Turkish Cypriot group, while the incidence of Haller cell, frontal sinus hypoplasia, maxillary sinus hypoplasia, ethmomaxillary sinus, sphenomaxillary plate, and septum deviation were significantly higher in Polish population. Conclusion: According to the Köppen-Geiger world climatic map, the climate is warmer and drier in Turkish Cypriote populations than in the Polish population. These climatic differences influence the paranasal sinus variations between the Turkish Cypriot and Polish populations that must be taken into account by rhinologic surgeons especially when performing frontal and sphenoid sinus surgery.

The aim of this study was to compare the reliability of both linear and angular measurements conducted on two-dimensional (2D) lateral cephalometric images and three-dimensional (3D) cone-beam computed tomography-generated cephalograms derived from various rendering software. Pre-treatment cephalometric digital radiographs of 15 patients and their corresponding cone beam computed tomographic images were randomly selected. Vista Dent OC as 2D, In vivo 5.1.2, Maxilim and Romexis software were used to generate cephalograms from the CBCT scans (NewTom 3G, QR Verona, Italy). In total, 19 cephalometric landmarks were identified and 18 widely used (11 linear, 7 angular) measurements were performed by an independent observer. Mann–Whitney and Kruskall–Wallis H tests were also used to compare the four methods (p < 0.05). Intra-class correlation coefficients (ICCs) were used to examine the intra-observer reliability. Intra-observer examinations showed various unreliable ICCs, which was an indication of poor reproducibility for condylion-gnathion (Co-Gn), gonion-mentum (Go-Me), anterior nasal spina-mentum (ANS-Me), Wits for all 2D and 3D rendering software (p > 0.05). The ICCs for Vista Dent OC (2D) measurements indicated high reproducibility (p < 0.05). The 2D and 3D generated cephalograms from various rendering software were found to be similar; however, measurements on curved surfaces are not easily reproducible for 3D software.