Objective: To evaluate the performance of vision transformer (ViT)-based deep learning models in the classification of open apex on panoramic radiographs (orthopantomograms (OPGs)) and compare their diagnostic accuracy with conventional convolutional neural network (CNN) architectures. Materials and Methods: OPGs were retrospectively collected and labeled by two observers based on apex closure status. Two ViT models (Base Patch16 and Patch32) and three CNN models (ResNet50, VGG19, and EfficientNetB0) were evaluated using eight classifiers (support vector machine (SVM), random forest (RF), XGBoost, logistic regression (LR), K-nearest neighbors (KNN), naïve Bayes (NB), decision tree (DT), and multi-layer perceptron (MLP)). Performance metrics (accuracy, precision, recall, F1 score, and area under the curve (AUC)) were computed. Results: ViT Base Patch16 384 with MLP achieved the highest accuracy (0.8462 ± 0.0330) and AUC (0.914 ± 0.032). Although CNN models like EfficientNetB0 + MLP performed competitively (0.8334 ± 0.0479 accuracy), ViT models demonstrated more balanced and robust performance. Conclusions: ViT models outperformed CNNs in classifying open apex, suggesting their integration into dental radiologic decision support systems. Future studies should focus on multi-center and multimodal data to improve generalizability.

Trigeminal neuralgia (TN) is a psychoneurotic disease involving the sensory branches of the trigeminal nerve (V) and characterized by severe and persistent pain. Pain, which is the most typical feature of the disease, is described by patients as short or long term, electric shock, burning, stinging and throbbing. Pain may involve all sensory branches of the nerve, and it is more common in areas innervated by the maxillary (V2) and mandibular (V3) nerve. Therefore, meticulous and regular follow-ups performed by dentists and oral surgeons and accurate examination of the cranial nerves are very important in order to avoid inappropriate treatments in patients who apply to dentists due to toothaches confused with facial pain. Otherwise, if the correct clinical history and pain type are not questioned, root canal treatments or even tooth extractions can be applied to patients due to misdiagnosis, also the pain does not go away. As can be seen in the three cases discussed in this article, it has been determined that many of the patients' teeth were treated with root canal treatment before the diagnosis of TN, and worse they were removed. It is very important for dentists to be careful about the distinction between TN-induced pain and toothache, in order to prevent incorrect treatments to be applied to patients.

Artificial intelligence architectures represent important technological approaches that contribute to the restructuring of image-based evaluation processes in dentistry. Deep learning–based Convolutional Neural Networks enable the automatic extraction of local features from dental radiological data, while Vision Transformer–based models offer a different perspective for analyzing complex anatomical structures through their capacity to model global contextual relationships within images. Hybrid Convolutional Neural Network –Transformer architectures provide a more comprehensive evaluation framework by integrating both local and global representations. Segmentation-oriented architectures such as U-Net and its variants enable pixel-level delineation of anatomical and pathological structures, thereby supporting diagnostic interpretation and treatment planning processes. Object detection architectures contribute to the automatic identification of clinically relevant regions within dental images characterized by wide fields of view. Due to its reliance on high-resolution imaging data and detailed anatomical assessment, dentistry constitutes a distinctive domain for artificial intelligence applications. In this context, appropriate architectural selection, data quality, and accurate definition of clinical requirements emerge as key determinants for the reliable and sustainable integration of artificial intelligence systems into dental practice. Yapay zekâ mimarileri, diş hekimliğinde görüntü temelli değerlendirme süreçlerinin yeniden yapılandırılmasına katkı sağlayan önemli teknolojik yaklaşımlar arasında yer almaktadır. Derin öğrenme tabanlı Konvolüsyonel Sinir Ağları, dental radyolojik verilerde lokal özelliklerin otomatik olarak çıkarılmasına olanak tanırken; Vision Transformer tabanlı modeller görüntü içerisindeki global bağlamsal ilişkileri modelleyebilme kapasitesiyle karmaşık anatomik yapıların analizine farklı bir perspektif kazandırmaktadır. CNN–Transformer hibrit mimariler ise lokal ve global temsilleri bir araya getirerek daha bütüncül bir değerlendirme imkânı sunmaktadır. Segmentasyon odaklı U-Net ve türevleri, anatomik ve patolojik yapıların piksel düzeyinde sınırlandırılmasını mümkün kılarak tanısal yorumlama ve tedavi planlaması süreçlerini desteklemektedir. Nesne tespiti mimarileri geniş görüş alanına sahip dental görüntülerde klinik açıdan anlamlı bölgelerin otomatik olarak belirlenmesine katkı sağlamaktadır Diş hekimliği pratiği, yüksek çözünürlüklü görüntüleme verileri ve ayrıntılı anatomik değerlendirme gereksinimi nedeniyle yapay zekâ uygulamaları için özgün bir çalışma alanı oluşturmaktadır. Bu bağlamda uygun mimari seçimi, veri kalitesi ve klinik gereksinimlerin doğru tanımlanması, yapay zekâ sistemlerinin güvenilir ve sürdürülebilir entegrasyonu açısından temel belirleyiciler olarak öne çıkmaktadır.