Development of an AI model to predict root canal treatment success using CBCT, clinical findings, and patient risk factors

Abstract

Root canal treatment (RCT) is a widely performed endodontic procedure aimed at preserving natural teeth by eliminating infection and promoting periapical healing. Despite advancements in endodontic techniques and imaging technologies, accurately predicting treatment success remains challenging because outcomes are influenced by multiple interrelated factors, including anatomical characteristics, clinical findings, treatment quality, and patient-specific risk factors. The growing adoption of artificial intelligence (AI) in healthcare presents an opportunity to improve prognostic accuracy through the integration of diverse clinical and imaging data. This study proposes the development of an AI-based predictive model that combines cone-beam computed tomography (CBCT) imaging, clinical findings, and patient risk factors to estimate the likelihood of root canal treatment success. The proposed framework involves collecting retrospective clinical records and CBCT datasets, preprocessing and extracting relevant imaging and clinical features, and training machine learning algorithms to identify patterns associated with favorable and unfavorable treatment outcomes. Model performance will be evaluated using standard predictive metrics, including accuracy, sensitivity, specificity, precision, recall, F1-score, and the area under the receiver operating characteristic curve. By integrating radiographic, clinical, and patient-related variables into a single predictive system, the proposed model is expected to enhance clinical decision-making, support personalized treatment planning, facilitate early identification of high-risk cases, and improve long-term treatment outcomes. The study further highlights the potential of AI-assisted decision support systems to advance evidence-based endodontic practice and contribute to more efficient, accurate, and patient-centered dental care.