INTEGRATION OF ARTIFICIAL INTELLIGENCE AND OMICS DATA FOR PREDICTING THE COURSE OF CHRONIC PERIODONTITIS

Abstract

Chronic periodontitis remains one of the most prevalent inflammatory pathologies of the oral cavity; however, its clinical course is characterized by significant inter-individual variability. Traditional diagnostics—based on assessing periodontal pocket depth, bleeding, and attachment loss— effectively reflect established pathological changes but are less effective at predicting the future progression of the disease process. Consequently, the integration of artificial intelligence and omics data—particularly microbiomics and proteomics—is of particular interest. Microbiome profiles enable the assessment of the degree of dysbiosis and the identification of taxa associated with inflammation, whereas proteomics reflects the functional state of tissues as well as the intensity of the inflammatory response and tissue destruction. Modern machine learning algorithms are capable of integrating these layers of information with clinical parameters to construct models aimed not only at diagnosis but also at predicting therapeutic response and the likelihood of disease progression. Nevertheless, the clinical implementation of such solutions remains limited due to study heterogeneity, a lack of standardized protocols, and insufficient external validation of the models. Frontiers PMC PMC