Aplicações da inteligência artificial no diagnóstico precoce de doenças peri-implantares

: uma revisão de novas estratégias

Abstract

Introduction: Peri-implant diseases are significant complications in implantology, potentially compromising the longevity of dental implants. Early detection of these pathologies is essential to prevent their progression and improve clinical outcomes. This scoping review aims to map the most recent Artificial Intelligence (AI)-based strategies applied to the early identification of peri-implant diseases, assessing their potential for integration into clinical practice. Materials and methods: The review was conducted in accordance with PRISMA-ScR guidelines and registered in the Open Science Framework (doi: https://doi.org/10.17605/OSF.IO/BPU2X). A bibliographic search was carried out in the PubMed and Web of Science databases. Studies published between 2017 and 2024 addressing the application of AI in the detection or prognosis of peri-implant diseases were included. The selection process followed predefined inclusion and exclusion criteria, with independent screening by two reviewers and resolution of disagreements by a third. Article management was performed using the Rayyan platform. Results: A total of 122 articles were initially identified, of which 25 were excluded as duplicates. After screening titles and abstracts, 87 articles were excluded for not meeting the eligibility criteria. Ten studies were included in the final analysis, encompassing different AI methodologies such as machine learning and deep learning applied to clinical and radiographic data. The main outcomes reported were accuracy in the automated detection of marginal bone loss, risk prediction for peri-implantitis, and support for clinical decision-making. Conclusion: Artificial intelligence demonstrates significant potential to improve the early diagnosis and management of peri-implant diseases, promoting a more preventive and personalized approach in implantology. However, further studies are needed to standardize methodologies and validate the clinical applicability of these tools.

Date of Award	9 Jul 2025
Original language	Portuguese
Awarding Institution	Universidade Católica Portuguesa
Supervisor	Rita Bornes (Supervisor) & Nuno Rosa (Co-Supervisor)