APPLICATION OF AI AND MACHINE LEARNING IN PREDICTING DENTAL DISEASES

Abstract

The predictive power of artificial intelligence (AI) and machine learning (ML) techniques for dental problems is examined quantitatively in this research paper. The study carefully reviews and evaluates the body of current literature to identify developments, methodologies, and findings on the application of AI and ML in dental diagnostics and sickness prediction.
Methodology: A systematic review methodology was utilized to locate appropriate studies published between 2010 and 2023 in major academic databases such as PubMed, IEEE Xplore, and Scopus. We integrated keywords related to artificial intelligence, machine learning, dental problems, diagnostics, and prediction into our search strategy. The eligibility requirements were satisfied by peer-reviewed English studies that focused on AI and ML applications for dental disease prediction.
Full-text publications were screened for inclusion using predefined criteria after their titles and abstracts were approved. A final selection of 35 publications for quantitative analysis was made using this procedure. The data extraction process took into account the study's features, the AI/ML techniques used, the size of the dataset, evaluation metrics, and the declared performance measurements (specificity, accuracy, sensitivity, and AUC).
Results: The results showed that a wide range of AI and ML methods were applied to predict dental problems, with impressive results being reported across studies. For the prediction of caries, deep learning techniques yielded an average accuracy of 85%, sensitivity of 88%, specificity of 82%, and AUC of 0.91. When it came to periodontal disease prediction, ensemble learning algorithms demonstrated an average of 78% accuracy, 82% sensitivity, 76% specificity, and 0.85 AUC.
Furthermore, AI-based models demonstrated promising results for the identification of oral cancer, with an average accuracy of 90%, sensitivity of 92%, specificity of 88%, and AUC of 0.94. These findings demonstrate how AI and ML can accurately identify dental problems, allowing for early intervention and personalized treatment regimens.
However, there are still problems that need to be solved, such as heterogeneity in the data, generalization of models, and clinical applicability, which calls for more research into the most effective applications of AI in dentistry. This quantitative study provides useful information on the efficacy and potential of AI/ML technologies in enhancing the prediction of dental disorders, which paves the way for more precise and effective clinical decision-making.