Study: Artificial intelligence may help predict implant treatment outcomes
ANN ARBOR, Mich., U.S.: As dental implants become an increasingly common treatment modality for edentulous patients, understanding their potential side effects grows ever more important. An American research team has thus developed a novel machine learning algorithm that could help dental practitioners better predict the risk of their implant patients developing periimplantitis.
The study was conducted by an interdisciplinary team spread across the University of Michigan, Michigan State University and the Harvard School of Dental Medicine. According to the researchers, periimplantitis affects the long-term success rates of at least 25% of dental implants, as the inflammation leads to the loss of supporting bone. Further complicating the treatment of periimplantitis is the fact that there is currently no reliable method for accurately predicting how an individual might respond to treatment.
In an attempt to untangle this issue, the research team developed a machine learning algorithm that they titled Fast and Robust Deconvolution of Expression Profiles—FARDEEP, in short. FARDEEP was then used to investigate the clinical, microbial and immune profiles of a group of implant patients undergoing regenerative therapy to help correct advanced periimplant defects. In doing so, the team was able to measure the relative levels of certain deleterious bacteria and helpful immune cells in each tissue sample collected from the patients.
Overall, higher amounts of immune cell types associated with microbial control were found to be strongly correlated to better clinical outcomes. According to Dr. Jeff Wang, lead author of the study and clinical assistant professor at the University of Michigan School of Dentistry, the results greatly improved the research team’s understanding of the nature of periimplantitis and could help them “further understand how to provide precision care.”
“The most direct clinical application of this study will be to help predict the outcome of surgical regenerative therapy for periimplantitis,” Wang told Dental Tribune International.
“When a patient has severe periimplantitis, it is difficult to make a decision about whether to treat or remove the implant. Regenerative therapy is expensive yet unpredictable; rebuilding the bone and replacing the implant is also a challenge,” he added.
“Therefore, prognostic information can be very helpful in determining what the best course of treatment for each individual patient is.”
Though FARDEEP’s potential is promising, Wang acknowledged that further clinical trials would be needed before it could be used by dental professionals to help predict periimplantitis risk in patients.
“This was a pilot study, as we plan to conduct larger clinical trials for validation,” he noted.
The study, titled “Machine learning-assisted immune profiling stratifies peri-implantitis patients with unique microbial colonization and clinical outcomes,” was published on May 3, 2021, in Theranostics.