A carregar...
Publicação
A inteligência artificial na análise do sorriso : revisão sistemática
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 1.46 MB | Adobe PDF |
Autores
Resumo(s)
Introdução: A análise estética do sorriso desempenha um papel central na medicina dentária contemporânea, refletindo a crescente importância atribuída à harmonia facial, à autoestima e à satisfação do paciente. A integração de ferramentas baseadas em inteligência artificial (IA) tem-se expandido, embora persistam dúvidas quanto à sua eficácia face aos métodos tradicionais conduzidos pelo médico-dentista. Objetivo: Avaliar a eficácia da IA na análise do sorriso em medicina dentária, em comparação com os métodos tradicionais, quanto à precisão, à eficácia clínica e aos resultados estéticos. Materiais e Métodos: Foi conduzida uma revisão sistemática segundo as recomendações PRISMA 2020, registada no PROSPERO. A pesquisa decorreu de Outubro a Dezembro de 2025 nas bases de dados MEDLINE (via PubMed), Scopus e Web of Science Core Collection, com filtros para publicações entre 2015 e 2025, em inglês e em português. A triagem e a avaliação do risco de viés (PROBAST e QUADAS-2) foram realizadas por dois investigadores independentes. Resultados: Foram incluídos 30 estudos publicados entre 2019 e 2025. Os modelos mais frequentemente descritos foram as redes neuronais convolucionais (CNN), seguidas de algoritmos clássicos de machine learning e, com menor ênfase, de redes generativas adversariais (GAN). Identificaram-se seis domínios de aplicação. A IA demonstrou desempenho comparável ao do clínico em tarefas geométricas e de identificação automática, contribuindo para a quantificação objetiva de mudanças estéticas pós-cirúrgicas. No design estético final de casos complexos, a preferência manteve-se favorável ao médico dentista. Conclusão: A IA constitui uma ferramenta promissora na análise do sorriso, com aplicações sólidas em segmentação, identificação de marcos e avaliação de atratividade, devendo ser usada como complemento ao julgamento clínico, com validação adicional do médico dentista e em conformidade com os requisitos éticos.
Introduction: Aesthetic smile analysis plays a central role in contemporary dentistry, reflecting the importance placed on facial harmony, self-esteem, and patient satisfaction. The integration of artificial intelligence (AI) tools is expanding, although doubts remain about their effectiveness compared with traditional clinician-led methods. Objective: To evaluate the effectiveness of AI in smile analysis in dentistry, compared with traditional methods, in terms of accuracy, clinical efficiency, and aesthetic outcomes. Materials and Methods: A systematic review was conducted in accordance with the PRISMA 2020 guidelines and registered with PROSPERO. The search was conducted between October and December 2025 in MEDLINE (via PubMed), Scopus, and Web of Science Core Collection, with filters for publications from 2015 to 2025 in English and Portuguese. Screening and risk-of-bias assessment (PROBAST and QUADAS-2) were conducted by two independent reviewers. Results: Thirty studies published between 2019 and 2025 were included. The most common models were convolutional neural networks (CNN), followed by classical machine learning algorithms and, less frequently, generative adversarial networks (GAN). Six application domains were identified. AI showed performance comparable to that of the clinician in geometric and automated identification tasks and contributed to the objective quantification of post-surgical aesthetic changes. For final aesthetic design in complex cases, preference remained with the dentist. Conclusion: AI is a promising tool for smile analysis, with solid applications in segmentation, landmark identification, and attractiveness assessment. It should complement clinical judgment and requires further validation and robust ethical safeguards.
Introduction: Aesthetic smile analysis plays a central role in contemporary dentistry, reflecting the importance placed on facial harmony, self-esteem, and patient satisfaction. The integration of artificial intelligence (AI) tools is expanding, although doubts remain about their effectiveness compared with traditional clinician-led methods. Objective: To evaluate the effectiveness of AI in smile analysis in dentistry, compared with traditional methods, in terms of accuracy, clinical efficiency, and aesthetic outcomes. Materials and Methods: A systematic review was conducted in accordance with the PRISMA 2020 guidelines and registered with PROSPERO. The search was conducted between October and December 2025 in MEDLINE (via PubMed), Scopus, and Web of Science Core Collection, with filters for publications from 2015 to 2025 in English and Portuguese. Screening and risk-of-bias assessment (PROBAST and QUADAS-2) were conducted by two independent reviewers. Results: Thirty studies published between 2019 and 2025 were included. The most common models were convolutional neural networks (CNN), followed by classical machine learning algorithms and, less frequently, generative adversarial networks (GAN). Six application domains were identified. AI showed performance comparable to that of the clinician in geometric and automated identification tasks and contributed to the objective quantification of post-surgical aesthetic changes. For final aesthetic design in complex cases, preference remained with the dentist. Conclusion: AI is a promising tool for smile analysis, with solid applications in segmentation, landmark identification, and attractiveness assessment. It should complement clinical judgment and requires further validation and robust ethical safeguards.
Descrição
Palavras-chave
Inteligência artificial Aprendizagem profunda Estética dentária Sorriso Diagnóstico por computador Artificial intelligence Deep learning Esthetics, dental Smiling Diagnosis, computer-assisted
Contexto Educativo
Citação
Editora
Licença CC
Sem licença CC