Akademik Veri Yönetim Sistemi
Current Oral Health Reports, cilt.12, sa.1, 2025 (Scopus)
Purpose of Review: This study aims to review the current applications of artificial intelligence (AI) in periodontology, focusing on the studies related with diagnostic accuracy, disease classification, risk assessment, and treatment prediction. It seeks to identify strengths, limitations, and future directions for AI based-studies to improve periodontal healthcare and personalized treatment planning. Recent Findings: AI is transforming the field of periodontology by enabling efficient identification, diagnosis, classification, and prediction of periodontal and peri-implant diseases and treatment outcomes. Through machine learning models such as convolutional neural networks (CNNs), Support Vector Machines (SVMs), Random Forests (RFs), and others, AI has demonstrated high diagnostic accuracy using radiographic images, clinical photographs, and patient demographics. Notable advancements include segmentation of teeth, cementoenamel junction identification, and periodontal bone loss prediction. Despite their promise, AI models face limitations such as a lack of large datasets, inconsistency in methodologies, and insufficient integration of clinical parameters like gingival structure or color changes. Studies reveal that modifiable risk factors, including lifestyle and systemic health conditions, significantly impact disease progression. Furthermore, AI shows potential in predicting treatment outcomes using synthetic datasets and regression models. While retrospective studies dominate current research, there remains a need for larger, more standardized datasets and advanced deep learning approaches to further enhance AI's role in periodontal care. Summary: AI has shown significant potential in periodontology, utilizing machine learning models like CNNs, SVMs, and RFs to diagnose, classify, and predict periodontal diseases with high accuracy. Despite advancements, challenges such as inconsistent methodologies, limited datasets, and the exclusion of key clinical parameters highlight the need for further research and refinement in AI applications for personalized periodontal care.