AI Predictive Models Show Promise in Tooth Loss Prevention

Using Artificial Intelligence to Predict Tooth Loss

Artificial intelligence (AI) is making significant strides in healthcare, particularly in predicting tooth loss, a critical indicator of overall oral health. Dr. Seyedmisagh Imani from Marquette University School of Dentistry has highlighted the potential of AI to assess a variety of risk factors, such as age, income, smoking habits, and frequency of dental visits. AI tools could help flag individuals at a higher risk of tooth loss even before they seek a dental examination.

“This technology may aid healthcare providers in emphasizing preventive care, particularly for those with limited access to dental services,” explained Dr. Imani, underscoring the role AI could play in transforming oral healthcare.

Machine Learning Techniques for Oral Health Insights

Dr. Imani, alongside his colleagues Meisam Omidi and Lobat Tayebi from Marquette University, presented their research at the 2024 IADR/AADOCR/CADR General Session in New Orleans. The study aimed to develop AI models to forecast the likelihood of permanent tooth loss, relying on a diverse set of behavioral and lifestyle factors.

The research utilized data from the CDC’s 2022 Behavioral Risk Factor Surveillance System (BRFSS), encompassing 293,398 individuals. Factors such as age, gender, income, smoking history, and dental care visits were evaluated using five different machine learning techniques, including K-nearest neighbor, logistic regression, decision trees, random forests, and extreme gradient boosting trees.

Key Findings: Socioeconomic Factors Matter

While age and routine dental visits emerged as the most significant predictors of tooth loss, the study found that socioeconomic conditions also played an essential role. The inclusion of socioeconomic characteristics, such as education and income, improved the accuracy of AI models, indicating that dental health disparities are deeply intertwined with broader societal factors.

“Models incorporating socioeconomic characteristics outperformed those relying solely on clinical dental indicators,” noted Dr. Imani, emphasizing the broader implications of the findings. These results suggest that AI models could be used in real-world settings to better identify high-risk individuals, guiding healthcare providers in prioritizing those in need of preventive interventions.

The Best Model: Extreme Gradient Boosting Trees

Among the AI techniques employed, the extreme gradient boosting trees model exhibited the highest predictive performance, achieving an area under the curve (AUC) score of 81.2%. This method stood out as the most effective at predicting tooth loss based on a combination of behavioral risk factors and socioeconomic conditions.

The study underscores the importance of integrating non-clinical factors into AI-driven healthcare solutions, as it significantly enhances their effectiveness. Dr. Imani’s team hopes these models will eventually be deployed in clinical settings, helping to detect at-risk individuals earlier and encouraging a shift toward preventive care in dentistry.

Implications for Preventive Dental Care

The broader application of these AI models could help address the challenges faced by populations with limited access to dental services, enabling more proactive healthcare interventions. By identifying high-risk individuals before they experience severe dental issues, healthcare systems could focus on preventive care, reducing the overall burden of tooth loss.

In conclusion, Dr. Imani and his team’s research highlights the growing potential of AI in the field of dentistry. By leveraging machine learning to predict tooth loss, healthcare providers could significantly improve oral health outcomes, especially for underserved communities.

“The findings suggest that incorporating socioeconomic factors into predictive models can enhance their accuracy and effectiveness,” concluded Dr. Imani, pointing to a future where AI-driven tools play a central role in improving dental care.