Loading...
Publication
Análise cefalométrica em ortodontia: avanços e impacto da inteligência artificial
| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| Projeto de pós-graduação_2022100616 | 1.65 MB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
A análise cefalométrica desempenha um papel crucial na ortodontia, fornecendo informações essenciais para o diagnóstico e planeamento dos tratamentos. Com a evolução da tecnologia, técnicas baseadas em Inteligência Artificial (IA), como Deep Learning e Redes Neurais Convolucionais, têm sido aplicadas para superar limitações dos métodos tradicionais, principalmente a variabilidade das marcações manuais.
Esta revisão narrativa tem como objetivo analisar a evolução e o impacto da IA na análise cefalométrica, fornecendo um panorama dos avanços recentes e das suas aplicações na prática ortodôntica. Foi realizada uma pesquisa na base de dados Medline PubMed, considerando apenas artigos em inglês dos últimos 10 anos. Foram selecionados 32 artigos para análise detalhada.
Os resultados demonstram que a IA, especialmente através de modelos baseados em Deep Learning, pode alcançar níveis elevados de precisão na detecção automática de marcos cefalométricos, com desempenho comparável ou superior ao de médicos dentistas iniciantes. Estudos mostram que a colaboração entre IA e profissionais pode melhorar significativamente a taxa de detecção de pontos cefalométricos e reduzir o erro radial médio.
Apesar do grande potencial, ainda existem desafios éticos e técnicos a considerar, como a necessidade de validação contínua, o risco de viés e a dependência de grandes conjuntos de dados. A IA surge, assim, como uma ferramenta promissora para aumentar a precisão diagnóstica, otimizar a formação em ortodontia e transformar a prática clínica.
Cephalometric analysis plays a crucial role in orthodontics, providing essential data for diagnosis and treatment planning. With technological advancements, Artificial Intelligence (AI) techniques such as Deep Learning and Convolutional Neural Networks have been applied to overcome the limitations of traditional methods, particularly the variability in manual landmark annotations. This narrative review aims to analyze the evolution and impact of AI on cephalometric analysis, offering an overview of recent developments and their clinical applications in orthodontics. A literature search was conducted in the Medline PubMed database, including only English articles from the last 10 years. A total of 32 articles were selected for in-depth analysis. Findings indicate that AI, especially Deep Learning-based models, can achieve high accuracy in automatic cephalometric landmark detection, often outperforming beginner dental professionals. Studies show that collaboration between AI and humans significantly improves detection rates and reduces radial errors. Despite its potential, challenges remain, including ethical and technical concerns such as the need for ongoing validation, potential bias, and dependence on large datasets. AI thus represents a promising tool to enhance diagnostic accuracy, support orthodontic education, and transform clinical practice.
Cephalometric analysis plays a crucial role in orthodontics, providing essential data for diagnosis and treatment planning. With technological advancements, Artificial Intelligence (AI) techniques such as Deep Learning and Convolutional Neural Networks have been applied to overcome the limitations of traditional methods, particularly the variability in manual landmark annotations. This narrative review aims to analyze the evolution and impact of AI on cephalometric analysis, offering an overview of recent developments and their clinical applications in orthodontics. A literature search was conducted in the Medline PubMed database, including only English articles from the last 10 years. A total of 32 articles were selected for in-depth analysis. Findings indicate that AI, especially Deep Learning-based models, can achieve high accuracy in automatic cephalometric landmark detection, often outperforming beginner dental professionals. Studies show that collaboration between AI and humans significantly improves detection rates and reduces radial errors. Despite its potential, challenges remain, including ethical and technical concerns such as the need for ongoing validation, potential bias, and dependence on large datasets. AI thus represents a promising tool to enhance diagnostic accuracy, support orthodontic education, and transform clinical practice.
Description
Keywords
Análise cefalométrica Aprendizagem profunda Inteligência artificial Cephalometric analysis Deep learning Artificial intelligence