Dental calculus is a significant contributor to periodontal disease, making its accurate identification crucial for effective prevention and treatment. Traditional methods for identifying dental calculus have the problems of strong subjectivity and low accuracy, especially when it comes to detecting calculus in the concealed subgingival areas, which cannot meet the demands of modern clinical practice. Recently, optical detection technology has attracted widespread attention due to its non-invasive and high-sensitivity characteristics. A variety of optical techniques have provided new ideas for the identification of dental calculus, such as polarization detection, optical coherence tomography (OCT), differential reflectometry, hyperspectral imaging (HSI), and fluorescence spectroscopy systems. Furthermore, artificial intelligence (AI) technology, particularly the combination of machine learning and deep learning with optical techniques, has significantly enhanced the level of automation and intelligence in the identification of dental calculus. This review provides the current mainstream methods for identifying dental calculus, compares and analyzes the advantages and disadvantages of each technology, and looks forward to the future development direction. This work aims to guide research and clinical application of dental calculus detection technologies.

Cephalometric analysis is indispensable for orthodontic and orthognathic treatment. With the development of three-dimensional (3D) imaging technology, 3D imaging is increasingly used to assess dentomaxillofacial deformity and formulate treatment plans. 3D cephalometric analysis based on multi-modal data contains more anatomical information than traditional 2D cephalometric analysis, which can be used to conduct a more comprehensive diagnosis of patients with dentomaxillofacial deformities, and has become a research hotspot. However, its application is accompanied by the problem of time-consuming and laborious. In recent years, the emergence of artificial intelligence (AI) can assist in the automation of landmark positioning, data collection and analysis in 3D cephalometric measurement. In this article, research status and the auxiliary application of AI in 3D cephalometric analysis were reviewed and summarized.

The application of deep learning (DL) has become widespread with the development of digital medicine. At present, DL has been gradually applied to the fields of stomatology. Multiple studies have applied DL, combined with preoperative examination images such as X ray and cone beam CT (CBCT) images, to assist clinical diagnosis and decision-making in dealing with impacted mandibular third molar (IMTM). Besides, inferior alveolar nerve (IAN) injury is one of the most serious sequelae after extraction of IMTM. Combined with imageological examination, DL can provide objective and accurate estimation of the risk of IAN injury to improve the outcome of treatment. This paper reviews the current application of DL in preoperative image recognition, preoperative auxiliary diagnosis and evaluation, and IAN injury prognosis prediction in the extraction of IMTM, and looked into the role of DL in the extraction of IMTM in the future.

Objective: To establish a design program for orthognathic surgical splints based on artificial intelligence (AI), and to compare the precision between the AI splints and manual digital splints. Methods: This research established an AI algorithm for orthognathic surgical splints and obtained the available automatic design program. The time required for designing surgical splints in the same case was compared. Besides, 40 dentition models of patients with skeletal Class Ⅱ malocclusion (20 for mandibular surgery, and 20 for bimaxillary surgery) were included for comparison. The splints were designed by AI program and manual digital software respectively and scanned for the digital data to compare the difference. Results: The overall deviation between the AI and manual digital splints in guiding the positioning of the models was less than 0.1 mm, indicating that there was no significant difference between them. Comparison of cusp position showed that the mean deviation distance between the AI and manual digital splints was about 0.10-0.14 mm. The application of the AI program can greatly reduce the design time in both the mandibular surgery [(10.7±2.4) s] and the bimaxillary surgery [(21.5±3.9) s]. Conclusion: There is no significant difference in accuracy between AI and manual digital splints, and AI program can improve the design efficiency of orthognathic surgical splints.

The research and application of artificial intelligence (AI) in oral clinical medicine has become widespread with numerous clinical practice in digital medicine and precision treatment. X ray and computed tomography (CT) are routine examination methods for oral diseases. By combining X ray and CT images with AI and analyzing a large amount of data, the data-driven analysis algorithm based on machine learning can finally be used for supporting clinical diagnosis and decision-making and assisting in establishing appropriate treatment plan. Lots of studies indicate that performance of the program system based on artificial intelligence technology is exceptionally excellent, and its accuracy is approaching or even better than trained professionals. This article aims to review the application of AI in oral and maxillofacial radiology and CT image processing, in order to provide a new idea for the prediction, diagnosis, treatment and prognosis of oral and maxillofacial diseases.