New Machine-learning Artificial Intelligence Can Help Identify Potential Risk for Tooth Loss

Investigators at Harvard School of Dental Medicine have compiled and researched a new study that suggests machine learning tools or artificial intelligence (AI) can help identify patients who may be at risk for potential tooth loss.

The study compared 5 algorithms using different combinations of variables to screen for tooth loss risk. Researchers used data comprising 12,000 adults from the National Health and Nutrition Examination Survey to design and test these machine learning algorithms. These algorithms were designed specifically to assess risk without a dental exam. Results demonstrated that medical characteristics and socioeconomic variables like race, education, arthritis, and diabetes, outperformed algorithms that relied on clinical indicators alone.

Analysis demonstrated that AI models are useful predictors of risk, and those that incorporate socioeconomic variables can supplement risk predictions even more. The AI approach can be applied in any healthcare profession, even non-dental. Tooth loss affects all aspects of life and can be extremely debilitating. Being able to delay or even prevent tooth loss with the help of machine learning could aid a great deal of people.