Cone-beam computed tomography (CBCT) is becoming increasingly popular for maxillofacial imaging. It can supply high-resolution three-dimensional (3D) images without distortion and superimposition of bone and other dental structures.
Researchers tested a novel AI system based on deep learning methods to determine its real-time performance of CBCT imaging diagnosis of anatomical landmarks, pathologies, clinical effectiveness, and safety when used by dentists in a clinical setting.
The deep learning-based CBCT diagnosis system consists of five modules: ROI-localization-module (segmentation of teeth and jaws), tooth-localization and numeration-module, periodontitis-module, caries-localization-module, and periapical-lesion-localization-module. These modules use Convolutional Neural Networks (CNNs) based on state-of-the-art architectures that play a significant part in AI-based object detection and segmentation.
The proposed AI system significantly improved the sensitivity and specificity of diagnosing the dental pathologies compared to human observers using CBCT imaging. The results look promising qualitatively and quantitatively. The system can have many uses in the real world, ranging from being a decision support system in clinical settings to a helper system for dental practitioners.