The design and fabrication of dentures often requires a great deal of time to collect dentition information and numerous fittings due to the limitations of the current technology, and can increase the time needed for patients to get proper fitting dentures.
If, for example, a missing tooth is not replaced, it not only affects one’s appearance and ability to chew but also causes jaw bone loss and displacement of nearby teeth, resulting in malocclusion, and affects the health of the remaining teeth, gums, jaw muscles, and jaw joints.
Therefore, when designing dentures, it is especially important that they resemble the original lost teeth in order to maintain the original appearance, chewing function, and oral and physical health. However, due to the lack of accuracy of the current methods, some denture products may not fit as well as expected.
The AI technology used in the study was based on 3D generative adversarial network (3D-GAN) algorithms that were tested on 175 student volunteers from HKU. The results showed that AI could reconstruct the shape of naturally healthy teeth and automate the design of dentures.
Co-researcher Dr. Reinhard Chau said that in a preliminary study, 3D-GAN was able to reconstruct the original, similar shape of teeth in 60 percent of patients. He expects the technology to improve as more data is collected.
Lead researcher Dr. Walter Yu Hang Lam said the study would facilitate the treatment workflow of dentists to replace missing teeth and reduce the need for patients to stay in the clinic for a long time as the preparation and installation of dentures would take less time than it presently does.
