Clinical readiness and generalizability of artificial intelligence in implant dentistry: a systematic review of diagnostic accuracy, treatment planning, prognosis and external validation

Vollständiger Abstract

Worum geht es in dieser Arbeit?

Artificial intelligence (AI) is increasingly being incorporated into digital dentistry. In implant dentistry, machine-learning and deep-learning systems have been investigated for radiographic diagnosis, anatomical segmentation, treatment planning, prosthetically driven implant positioning, and prediction of implant-related outcomes. To systematically evaluate the clinical applications of AI in implant dentistry, with particular emphasis on diagnostic accuracy, treatment planning, prognosis, external validation, generalizability, and clinical applicability. A systematic review will be conducted according to PRISMA 2020. PubMed/MEDLINE, Scopus, Web of Science, Embase, and Cochrane Library will be searched from inception to the final search date. Studies evaluating AI, machine learning, deep learning, neural networks, or related computational approaches in human implant dentistry will be considered. Diagnostic studies will be assessed using QUADAS-2, prognostic studies using QUAPAS where applicable, and prediction-model studies using PROBAST+AI. Reporting quality of prediction models will be assessed using TRIPOD+AI. Current evidence demonstrates promising AI performance in anatomical segmentation, CBCT analysis, implant treatment planning, prosthetically driven positioning, and prediction of implant-related outcomes. However, the literature is heterogeneous with respect to datasets, algorithms, reference standards, outcome definitions, validation procedures, and performance measures. Many studies remain retrospective, and external validation and evidence of clinical utility are comparatively limited. AI has considerable potential as a decision-support technology in implant dentistry. The principal research priority is now progression from technical accuracy toward externally validated, explainable, calibrated, clinically useful, and patient-centered AI systems. Keywords AI; machine learning; deep learning; dental implants; implant dentistry; CBCT; implant planning; implant prognosis; peri-implantitis; digital dentistry.