UCSF Secures $4M NIH Grant to Train AI in Detecting Cavities

The UC San Francisco School of Dentistry has received a $4 million National Institutes of Health (NIH) grant to kickstart its clinical research program, in part by developing a new artificial intelligence tool to better detect cavities.

The grant will train School of Dentistry learners, residents and faculty to design and conduct clinical research, from enrolling patients to analyzing results, skills they will hone by embarking on research to evaluate AI technology. The training will enable learners and faculty to use this knowledge in their own careers, accelerating the advancement of new discoveries and treatments in patient care.

… UCSF is uniquely positioned to ensure the future of AI in oral health is patient-centered and equitable.”

STUART GANSKY, MS, DRPH

Education like this, said Dentistry Professor Lisa Berens, DDS, MPH, will also ensure that UCSF dental graduates are ready to assess and use new innovations such as AI in their own practices one day.

“By understanding how clinical research is conducted, we’re equipping our graduates to become lifetime educated consumers of science,” explained Berens, who is one of three principal investigators on the grant. “If they see a new technology at a conference, for instance, they’re going to know how to evaluate the evidence that shows whether that technology, app or device does what it’s claiming.”

Training provided through the grant will also enable dentistry faculty to apply for more than $5 million in annual funding for their own clinical research projects.

A window on dentistry’s future

The AI tool created as part of the grant will use dental images gathered from UCSF dental clinics in Northern California as the basis for “training” the AI.

“This research project looking at artificial intelligence, assisted diagnostics and treatment planning is a window into dentistry’s future,” said Eni Obadan-Udoh, DDS, MPH, DSc, associate professor and fellow principal investigator. “That future will either be shaped by the market, which doesn’t always have the public’s interest at heart, or by patients’ actual needs.”

AI is already being used to study collections of images like brain scans, “learning” what might signal cancer and revealing clues about how to treat it. UCSF researchers aim to develop a tool that will help dentists diagnose cavities earlier, sparing patients pain and more invasive and costly treatments like root canals or extractions.

Equitable care in the public interest

Another advantage, according to Associate Dean for Research and fellow principal investigator Stuart Gansky, MS, DrPH, is the diversity of patients served by UCSF dental clinics, which span urban areas like San Francisco and Sacramento to rural communities like Fresno and Grass Valley.

UCSF’s partner clinics, where dental students train, treat everyone regardless of ability to pay. By training UCSF’s future AI tool on dental scans from patients that represent the full spectrum of Californians, Gansky explained, the university hopes that its AI tool will be fit to serve a state as diverse as California.

“As a public institution with a long tradition in health equity research, UCSF is uniquely positioned to ensure the future of AI in oral health is patient-centered and equitable,” Gansky added.

The five-year grant is a partnership between the School of Dentistry – which has been the leading recipient of NIH funding among dental  schools for 30 years – and UCSF’s Center for Intelligent ImagingClinical & Translational Science Institute and Bakar Computational Health Sciences Institute, among others.