Do Pixel ao Diagnóstico: Inteligência Artificial no Reconhecimento de Lesões da Mucosa Oral — Evidências, Acurácia e Desafios Clínicos

Abstract

The incorporation of artificial intelligence (AI) into the field of dentistry has brought about significant changes in diagnostic methods, particularly through the analysis of digital images. The aim of this study is to analyse the scientific evidence regarding the application of artificial intelligence in the recognition of oral mucosal lesions, assessing its diagnostic accuracy, clinical potential and the challenges associated with its implementation in dental practice. The search engines used to select the articles included Google Scholar, PubMed, Scopus and Web of Science, using the following Portuguese-language search terms: “Artificial Intelligence”; “Deep Learning” “Representation-based Machine Learning”; “Oral Lesions”; “Oral Neoplasms” and “Diagnosis”, combined using the Boolean operators AND and OR, with the aim of increasing the sensitivity and specificity of the search strategy.The results show that AI systems possess a high capacity for pattern recognition, with accuracy comparable to or superior to that of conventional methods, thereby facilitating early detection and standardising diagnosis. However, limitations relating to data quality, algorithmic biases, clinical validation and ethical considerations still restrict their widespread application. Thus, AI emerges as a promising complementary tool in the diagnosis of oral mucosal lesions, acting as a support to the dental surgeon and requiring further research to enhance its clinical reliability.

Keywords: artificial intelligence; oral diagnosis; oral mucosal lesions; deep learning; digital dentistry.