Objective To evaluate the performance of a new artificial intelligence (AI)-driven tool for tooth detection and segmentation on panoramic radiographs. Materials and methods In total, 153 radiographs were collected. A dentomaxillofacial radiologist labeled and segmented each tooth, serving as the ground truth. Class-agnostic crops with one tooth resulted in 3576 training teeth. The AI-driven tool combined two deep convolutional neural networks with expert refinement. Accuracy of the system to detect and segment teeth was the primary outcome, time analysis secondary. The Kruskal-Wallis test was used to evaluate differences of performance metrics among teeth groups and different devices and chi-square test to verify associations among the amount of corrections, presence of false positive and false negative, and crown and root parts of teeth with potential AI misinterpretations. Results The system achieved a sensitivity of 98.9% and a precision of 99.6% for tooth detection. For segmenting teeth, lower canines presented best results with the following values for intersection over union, precision, recall, F1-score, and Hausdorff distances: 95.3%, 96.9%, 98.3%, 97.5%, and 7.9, respectively. Although still above 90%, segmentation results for both upper and lower molars were somewhat lower. The method showed a clinically significant reduction of 67% of the time consumed for the manual. Conclusions The AI tool yielded a highly accurate and fast performance for detecting and segmenting teeth, faster than the ground truth alone. Clinical significance An innovative clinical AI-driven tool showed a faster and more accurate performance to detect and segment teeth on panoramic radiographs compared with manual segmentation.

With the growing demand for dental work, trigeminal nerve injuries are increasingly common. This retrospective cohort study examined 53 cases of iatrogenic trigeminal nerve injury seen at the Department of Oral and Maxillofacial Surgery, University Hospitals of Leuven between 2013 and 2014 (0.6% among 8845 new patient visits). Patient records were screened for post-traumatic trigeminal nerve neuropathy caused by nerve injury incurred during implant surgery, endodontic treatment, local anaesthesia, tooth extraction, or specifically third molar removal. The patients ranged in age from 15 to 80years (mean age 42.1years) and 68% were female. The referral delay ranged from 1day to 6.5years (average 10months). The inferior alveolar nerve (IAN) was most frequently injured (28 cases), followed by the lingual nerve (LN) (21 cases). Most nerve injuries were caused during third molar removal (24 cases), followed by implant placement (nine cases) and local anaesthesia injuries (nine cases). Pain symptoms were experienced by 54% of patients suffering IAN injury, compared to 10% of patients with LN injury. Persistent neurosensory disturbances were identified in 60% of patients. While prevention remains the key issue, timely referral seems to be a critical factor for the successful treatment of post-traumatic neuropathy.

BackgroundIntraoral scanners are devices for capturing digital impressions in dentistry. Until now, several in vitro studies have assessed the trueness of digital impressions, but in vivo studies are missing. Therefore, the purpose of this study was to introduce a new method to assess trueness of intraoral scanners and digital impressions in an in vivo clinical set-up.MethodsA digital impression using an intraoral scanner (Trios® 3 Cart wired, 3Shape, Copenhagen, Denmark) and a conventional alginate impression (Cavex Impressional®, Cavex, Haarlem, the Netherlands) as clinical reference were made for two patients assigned for full mouth extraction. A total of 30 teeth were collected upon surgery after impressions making. The gypsum model created from conventional impression and extracted teeth were then scanned in a lab scanner (Activity 885®, SmartOptics, Bochum, Germany). Digital model of the intraoral scanner (DM), digital model of the conventional gypsum cast (CM) and those of the extracted natural teeth (NT) were imported to a reverse engineering software (3-matic®, Materialise, Leuven, Belgium) in which the three models were registered then DM and CM were compared to their corresponding teeth in NT by distance map calculations.ResultsDM had statistically insignificant better trueness when compared to CM for total dataset (p = 0.15), statistically insignificant better trueness for CM when mandibular arches analyzed alone (p = 0.56), while a significantly better DM trueness (p = 0.013) was found when only maxillary arches were compared.ConclusionsOur results show that digital impression technique is clinically as good as or better than the current reference standard for study models of orthognathic surgery patients.

The purpose of the presented Artificial Intelligence (AI)-tool was to automatically segment the mandibular molars on panoramic radiographs and extract the molar orientations in order to predict the third molars’ eruption potential. In total, 838 panoramic radiographs were used for training (n = 588) and validation (n = 250) of the network. A fully convolutional neural network with ResNet-101 backbone jointly predicted the molar segmentation maps and an estimate of the orientation lines, which was then iteratively refined by regression on the mesial and distal sides of the segmentation contours. Accuracy was quantified as the fraction of correct angulations (with predefined error intervals) compared to human reference measurements. Performance differences between the network and reference measurements were visually assessed using Bland−Altman plots. The quantitative analysis for automatic molar segmentation resulted in mean IoUs approximating 90%. Mean Hausdorff distances were lowest for first and second molars. The network angulation measurements reached accuracies of 79.7% [−2.5°; 2.5°] and 98.1% [−5°; 5°], combined with a clinically significant reduction in user-time of >53%. In conclusion, this study validated a new and unique AI-driven tool for fast, accurate, and consistent automated measurement of molar angulations on panoramic radiographs. Complementing the dental practitioner with accurate AI-tools will facilitate and optimize dental care and synergistically lead to ever-increasing diagnostic accuracies.

Structured Abstract Objectives The aim of our study was to identify and predict patients at risk of impeded mandibular third molar eruption and potential relation between the third molar roots and the mandibular canal, based on molar angulations in an early development stage. Setting and Sample Population A total of 1011 adolescent orthodontic patients were included in this longitudinal study. Materials and Methods We analysed pre‐eruptive rotational changes and root development of mandibular third molars on 2022 panoramic radiographs (two time‐points). Five variables were evaluated: third molar eruption level, development stage, risk of relation between the third molar and the mandibular canal, the molar angulations and orthodontic treatment. The relation between early third molar angulation and mean annual angulation change was assessed using a linear mixed model. Logistic regression was applied to investigate a potential correlation of the radiographic variables with the eruption potential and risk of developing a relation between the third molar and the mandibular canal. Results Mandibular third molar follicles with an initial angulation exceeding 27.0° relative to the second molar tend to progressively increase their angulation during further development. A significant correlation was found between the hemimandibular molar angulations and the probability of eruption (P < 0.0001). The second to first molar angulation was predictive for potential development of a relation with the mandibular canal (P = 0.005). Conclusion From the present data, it appears that severely angulated mandibular third molars (>27.0°) have a minimal chance of future eruption and a maximal risk of developing a relation with the mandibular canal.