AI Algorithm Set to Revolutionize Dental Crown Production

  • An international team of scientists led by the University of Hong Kong has developed an AI algorithm capable of designing personalized dental crowns in just three seconds, which are reportedly closer in resemblance and function to natural teeth compared to those created through conventional methods.
  • The team is working on potentially licensing the use of the AI algorithm within the Greater Bay Area, which produces 25 to 30 per cent of the world’s dental prostheses, and plans to conduct clinical trials to compare the performance of AI-designed crowns with conventionally produced ones.
  • The artificial intelligence algorithm developed by the University of Hong Kong-led team is revolutionizing dental prosthetics by designing personalized dental crowns in a mere three seconds, significantly reducing the conventional time of 5-15 minutes.

An international group of scientists led by the University of Hong Kong (HKU) have developed an artificial intelligence (AI) algorithm capable of designing personalized dental crowns in a matter of seconds. This groundbreaking development promises to improve the efficiency and precision of dental procedures worldwide.

The AI system, devised by a consortium of dental and computer scientists, generates dental crowns that closely mimic natural tooth structures, resulting in a better fit than conventionally produced crowns. The crowns are custom-made, drawing data from neighboring teeth to the tooth needing replacement.

Lead researcher James Tsoi Kit-hon of the HKU dental materials science team, flanked by study author and postdoctoral fellow Ding Hao (left) and PhD candidate Crystal Chen. Photo: Handout
Credit: South China Morning Post

Presently, the design process for a dental crown can take between five to 15 minutes, using software that fine-tunes crown templates to better fit patients, according to lead researcher James Tsoi Kit-hon, an associate professor in dental materials science at HKU. However, these conventional techniques often produce crowns that are either too big, leading to poor fit, or too thin, resulting in a high risk of breakage.

Tsoi explains that the AI, on the other hand, offers an optimal balance between the lifespan and functionality of the dental device, eliminating the common issues associated with traditional methods.

The findings of this ground-breaking research, conducted by scientists from HKU, ShanghaiTech University, Drexel University, the University of Minnesota, and the University of Manchester, were published in the peer-reviewed journal Dental Materials in March.

Postdoctoral fellow at the HKU faculty of dentistry and study author Ding Hao explains that the AI algorithm is comprised of a generator which produces crown designs, and a discriminator that can distinguish between real teeth and AI-generated ones. The generator was trained using 600 sets of natural and healthy dental casts, with the discriminator checking the designs for their likeness to real teeth.

AI-generated crowns were found to be closer in resemblance and function to natural teeth compared to crowns designed through conventional methods.

The team aims to leverage their technology to enhance the dental industry in the Greater Bay Area, which produces 25 to 30 per cent of the world’s dental prostheses. Collaborations with dental labs are underway to potentially license the use of the AI algorithm within the region.

The team’s next step includes using their software for the production of other dental prostheses such as bridges and dentures. They also plan to conduct clinical trials comparing the performance of AI-designed crowns with conventional ones, with two years of follow-up to ensure robust data collection.