Dentistry is a technically oriented profession, and the ubiquitous trend of digitalization significantly influences the healthcare sector. Some of these digital developments have the potential to result in disruptive changes for dental practice, while others may turn out to be just a pipedream. 

AI influences the field of dentistry by 

• Dental eHealth data management
• Clinical and technical healthcare applications
• Services and operations

AI systems enable personalized dental medicine workflows by analyzing all eHealth data gathered from an individual patient. Besides dental-specific data, this also includes genomic proteomic, facilitating optimized and personalized treatment strategies and risk management. Based on the power of AI, the triangular frame of “data”/“health care”/“service” is supplemented by technological advancements in the field of social media, Internet of things, augmented and virtual reality, rapid prototyping, and intraoral optical scanning as well as teledentistry.

Innovation remains critical to tackling dental problems until its routine implementation is based on sound scientific evidence. Novel technologies must be viewed critically in relation to the cost-benefit ratio and the ethical implications of a misleading diagnosis or treatment produced by AI algorithms. Highly sensitive eHealth data must be handled responsibly to enable the immense benefits of these technologies to be realized for society. The focus on patient-centred research and the development of personalized dental medicine can potentially improve individual and public health and clarify the interconnectivity of disease in a more cost-effective way.

AI in dentistry

Bone loss is among the most challenging oral and maxillofacial conditions to treat, resulting from progressive dentoalveolar diseases in which invasive microorganisms from the oral environment or the periodontium, orthodontic treatments, trauma and pathological systemic/local conditions can produce inflammation, disrupting bone homeostasis and leading to increased bone resorption.

Most dentists attempt to diagnose the condition by determining the type, degree and extension of alveolar bone loss using conventional radiographs; in specialist practices, the use of cone-beam computed tomography (CBCT) for diagnosis is also prevalent, while panoramic radiographs are also used to screen for apical pathosis. CBCT allows dentists to treat their patients more effectively by producing undistorted, three-dimensional images, enabling the evaluation of anatomical structures with no superimposition. The novel possibility of enhancing conventional radiographs and CBCT with AI or machine-based learning is a significant step forward to aid in diagnosing and treating complicated dentoalveolar pathologies misdiagnosed previously.

Current literature supports the acceptable performance of some AI algorithms in assessing periapical lesions based on CBCT, panoramic and periapical radiographs. For instance, a new deep learning algorithm improved the sensitivity and specificity of CBCT segmentation and pathosis detection. Setzer et al. reported a detection accuracy of 93% with a specificity of 88%, a positive predictive value (PPV) of 87% and a negative predictive value (NPV) of 93% on CBCT images for a deep learning algorithm. 

Endres et al. used a DL algorithm to detect periapical pathoses on panoramic radiographs. They reported precision and F1 scores of 60% and 58%, respectively, compared to the performance of an oral and maxillofacial surgeon. Li et al. found that a modified deep-learning model with a large dataset of periapical radiographs could detect periapical periodontitis with a score of 82%. Based on the results, deep learning could enhance diagnostic accuracy and improve inter-observer agreement.

Future of AI

Due to the need for precision and instant information exchange in dentistry, AI will continue to connect with the dental profession in every aspect. The authors believe that, with the current trend and recent rapid development of AI, we can expect to see its impact on dentistry in the very near future. ML, especially deep learning, will help researchers understand certain multifactorial diseases better.