By An-Li Herring
Photography by Rayni Shiring, University of Pittsburgh
Should a damaged tooth be saved or removed? Is a subtle shadow on a dental X-ray cause for concern?
Those are the kinds of decisions dentists make every day, whether they practice on their own or as part of a larger team. At the University of Pittsburgh School of Dental Medicine, faculty members are exploring how artificial intelligence (AI) can help students answer those questions with greater confidence and develop the judgment that remains at the heart of patient care.
With support from Pitt’s Innovation in Education Awards program, two faculty-led projects are putting that approach into practice.
Kevin Pursel, assistant professor of restorative dentistry and comprehensive care, is leading a team using an AI-powered platform to help students learn one of dentistry's most fundamental skills: interpreting dental X-rays.
Rather than turning to AI first, students interpret each X-ray on their own before reviewing the AI’s assessment and then discussing any differences with faculty.
“What we really don’t want to have students doing is going straight to the AI,” Pursel said. Instead, the technology becomes a point of comparison students can use to sharpen their own observations and strengthen their diagnostic skills.
The AI gives students immediate feedback while they practice interpreting dental X-rays, helping them build the diagnostic skills and confidence they will rely on when caring for patients independently. Comparing their own interpretation with both the AI’s assessment and their instructor’s feedback helps them recognize where their thinking aligns—and where it can improve.
A second project focuses on another of dentistry’s most consequential decisions: whether a damaged tooth can be saved or should be extracted. While removing a damaged tooth may seem like the simplest solution, dentists also have to consider whether that decision could leave a patient facing a more extensive treatment plan years down the road. Some cases are straightforward, but others fall into a gray area where clinicians must weigh a patient’s overall health and other clinical findings before recommending a course of treatment.
Working with faculty mentor Andrea Ravidà, assistant professor of periodontics and preventive dentistry and director of Pitt’s Advanced Education Program in Periodontics, third-year periodontics resident Pablo Lopez Galindo is helping bring years of AI research into dental education. By analyzing clinical measurements along with patient factors such as age, smoking and diabetes, the team’s AI model estimates the likelihood that a tooth will survive for 10 years with treatment that does not involve extraction. Students first record their own prognosis, then compare it with the AI’s assessment, encouraging them to reflect on their own clinical reasoning.
“Dentists have to make decisions that can affect their patients irreversibly,” Lopez Galindo said. “One of them is whether to extract or to retain a tooth.”
Although the projects led by Ravidà and Pursel focus on different aspects of dentistry, they share a common educational approach. In both cases, students make their own assessment before consulting AI, using the technology to compare, question and refine their thinking. The confidence students gain by learning to think alongside AI will help prepare them to adapt as new technologies continue to shape patient care across a wide range of practice settings.