Moraschini Vittorio, de Almeida Daniel Costa Ferreira, Louro Rafael Seabra, de Oliveira Silva Alice Maria, Neto Mario Pereira Couto, Dos Santos Gustavo Oliveira, Granjeiro José Mauro
Full Professor, Department of Oral Surgery, School of Dentistry, Fluminense Federal University (UFF), Niterói, RJ, Brazil.
Section Head, Digital Dentistry Section, Dentistry Division, Brazilian Air Force, Rio de Janeiro, Brazil.
J Prosthet Dent. 2025 Jun;133(6):1461.e1-1461.e10. doi: 10.1016/j.prosdent.2024.05.030. Epub 2024 Jul 9.
The use of artificial intelligence (AI) in dentistry has grown. However, the accuracy of clinical applications in implant dentistry is still unclear.
The purpose of this scoping review with systematic evidence mapping was to identify and describe the available evidence on the accuracy and clinical applications of AI in implant dentistry.
An electronic search was performed in 4 databases and nonpeer-reviewed literature for articles published up to November 2023. The eligibility criteria comprised observational and interventional studies correlating AI and implant dentistry. A bibliographic mapping and quality analysis of the included studies was conducted. Additionally, the accuracy rates of each AI model were evaluated.
Twenty-six studies met the inclusion criteria. A significant increase in evidence has been observed in recent years. The most commonly found applications of AI in implant dentistry were for the recognition of implant systems followed by surgical implant planning. The performance of AI models was generally high (mean of 88.7%), with marginal bone loss (MBL) prediction models being the most accurate (mean of 93%). Regarding the place of publication, the Asian continent represented the highest number of studies, followed by the European and South American continents.
Evidence involving AI and implant dentistry has grown in the last decade. Although still under development, all AI models evaluated demonstrated high accuracy and clinical applicability. Further studies evaluating the clinical efficacy of AI models in implant dentistry are essential.
人工智能（AI）在牙科领域的应用日益广泛。然而，其在种植牙科临床应用中的准确性仍不明确。
本系统证据图谱的范围综述旨在识别和描述关于AI在种植牙科准确性和临床应用的现有证据。
在4个数据库和非同行评审文献中进行电子检索，查找截至2023年11月发表的文章。纳入标准包括将AI与种植牙科相关联的观察性和干预性研究。对纳入研究进行文献图谱绘制和质量分析。此外，评估每个AI模型的准确率。
26项研究符合纳入标准。近年来证据数量显著增加。AI在种植牙科最常见的应用是识别种植系统，其次是手术种植规划。AI模型的表现总体较高（平均88.7%），其中边缘骨丢失（MBL）预测模型最为准确（平均93%）。就发表地而言，亚洲大陆的研究数量最多，其次是欧洲和南美洲大陆。
在过去十年中，涉及AI与种植牙科的证据有所增加。尽管仍在发展中，但所有评估的AI模型均显示出较高的准确性和临床适用性。进一步评估AI模型在种植牙科临床疗效的研究至关重要。
