AI may automate design of biomimetic single-tooth protheses

Search Dental Tribune

Artificial intelligence may automate design of biomimetic single-tooth protheses

E-Newsletter

The latest news in dentistry free of charge.

  • This field is for validation purposes and should be left unchanged.
In a recent experimental study, Hong Kong researchers demonstrated that their AI system could generate the design of a molar (red) based on the features of the remaining dentition (dark grey). (Image: HKU)

HONG KONG: Even with the support of modern CAD/CAM technology, creating a dental prosthesis is still rather time-consuming, resulting in more chair time and high costs for patients. To facilitate the design of molar crowns, researchers from the Faculty of Dentistry at the University of Hong Kong (HKU) and the Department of Computer Science of Chu Hai College of Higher Education in Hong Kong collaborated to develop a novel approach using artificial intelligence (AI).

When asked what inspired the research, lead author Dr Walter Yu Hang Lam, clinical assistant professor in prosthodontics at the Division of Restorative Dental Sciences at HKU, told Dental Tribune International: “Some patients sense a very subtle hair-thin high spot on their dental prosthesis. Therefore, in the dental curriculum, a significant proportion of time is dedicated to occlusion theory and clinical training to provide a dental prosthesis that fits the patient’s mouth. My colleagues and I hoped to figure out a solution for improved treatment efficiency and patient experience.”

In order to restore the patient’s original appearance, masticatory function and general oral health, dental protheses should have the same occlusal morphology and 3D position of the natural teeth. These can be deduced for a missing tooth from those of the surrounding dentition because the teeth of an individual are all controlled by the same set of genes and exposed to the same oral environment. The researchers hypothesised that AI could thus generate the design for a single-tooth prothesis based on the characteristics of the remaining dentition.

The research team used a machine learning approach called a generative adversarial network (GAN) to train and validate their AI system and have tested it on 175 participants. The system was able to reconstruct the shape of a natural tooth and automate the process of dental protheses design based only on the digital model of the patient’s dentition.

“The 3D GAN algorithm was selected due to its superior performance on 3D object reconstruction compared with other AI algorithms. In the preliminary study, 3D GAN was able to rebuild similar shapes to the original teeth for 60% of the cases. It is expected to mature with more AI training data,” commented co-author Dr Reinhard Chun Wang Chau, research assistant in the Division of Restorative Dental Sciences and of Applied Oral Sciences and Community Dental Care at HKU, in a press release. For future research, the team proposes to investigate whether the presence of opposing teeth will help the AI to generate a more natural tooth.

Asked about the advantages of this method for dental professionals and patients, Dr Lam said: “It’s less time-consuming for both of them. Dentists will spend less time on registering jaw relationships and chairside adjustment, greatly facilitating the entire treatment process and enabling them to take on more cases.”

He continued: “Patients will spend less time and money on the treatment. In addition, the dental prostheses they receive will fit better to their remaining dentition and are thus less likely to cause jaw problems.”

According to Dr Lam, the research group hopes to make the AI technology available for dental professionals within the next five years, after having tested its accuracy further in simulated and clinical scenarios. Moreover, the researchers believe that the method may be applied to the fabrication of crowns for other teeth and of multi-unit restorations in the future.

The study, titled “Artificial intelligence-designed single molar dental prostheses: A protocol of prospective experimental study”, was published online on 2 June 2022 in PLOS ONE.

Leave a Reply

Your email address will not be published. Required fields are marked *

advertisement
advertisement