Objective. To evaluate current artificial intelligence (AI) and machine learning (ML) models used for predicting dental implant outcomes and to identify key risk factors and performance metrics based on a systematic literature review. Materials and methods. A systematic review was conducted in accordance with PRISMA 2020 guidelines. Literature searches were performed in PubMed and eLIBRARY databases, supplemented by manual searches of publications from 2015 to 2025. Original clinical studies applying AI/ML methods to predict dental implant success or complications were included. Risk of bias was assessed using RoB 2.0, PROBAST, and QUADAS-2 tools. Due to methodological heterogeneity, a qualitative synthesis was performed. Results. Twenty-two studies were included in the final analysis. Deep learning models, particularly convolutional neural networks applied to CBCT and panoramic radiographic data, demonstrated the highest predictive performance (AUC up to 0.93; accuracy >90%). Ensemble ML methods showed stable predictive capability and facilitated identification of clinically relevant predictors. Smoking, oral hygiene status, bone quality, implant dimensions, and systemic diseases consistently emerged as major risk factors across models. Conclusion. AI and ML methods show strong potential for improving personalized prediction of dental implant outcomes. Future progress depends on integrating multimodal clinical and radiological data, improving interpretability, and conducting external validation to support clinical implementation.
Semenova et al. (Tue,) studied this question.