A carregar...
Publicação
PrepAIze : aplicação digital para análise e feedback de preparações dentárias 3D/IA
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 1.83 MB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
Introdução: A Medicina Dentária passa por um período de transformação incentivado não só pela digitalização, mas, também, pela inteligência artificial (IA).
No ensino pré-graduado, a avaliação convencional de preparos dentários apresenta algumas limitações estruturais, tais como a variabilidade inter-avaliador, a escassez de feedback imediato e individualizado e a ausência de métricas padronizadas, que comprometem a equidade e a eficácia da aprendizagem.
Objetivos: Desenvolver e, posteriormente, averiguar o funcionamento de uma aplicação digital capaz de analisar modelos tridimensionais de preparos para coroa total em zircónia de pré-molares superiores, capaz de calcular automaticamente parâmetros geométricos clinicamente relevantes e fornecer feedback educativo estruturado, com o objetivo de melhorar do desempenho do estudante.
Materiais e Métodos: Estudo de investigação aplicada e desenvolvimento tecnológico, de natureza descritiva, em contexto pré-clínico simulado. Após calibração de dois avaliadores especialistas de Reabilitação Oral, foram realizados e digitalizados 30 preparos em tipodentes, sendo o conjunto de referência. A avaliação baseia-se numa grelha de 11 critérios ponderados (reduções, parâmetros técnicos e acabamento). Os thresholds foram definidos a partir do Color Atlas of Fixed Prosthodontics e a ferramenta foi demonstrada num preparo prático.
Resultados: Confirmou-se que os critérios dos especialistas são convertíveis em parâmetros geométricos mensuráveis. A PrepAIze demonstrou capacidade de importar ficheiros STL, calcular os 11 parâmetros, verificar os intervalos de aceitação, gerar a Pontuação Clínica e o IA Score Final, classificando o preparo num dos quatro níveis qualitativos, finalizando com a elaboração de um relatório de feedback estruturado.
Conclusão: As três hipóteses foram confirmadas, suportando a viabilidade técnica e funcional do sistema. A PrepAIze revela-se uma ferramenta promissora para a objetivação e automação da avaliação de preparos, complementando o papel do docente e promovendo a aprendizagem autodirigida.
Introduction: Dentistry is undergoing a period of transformation driven not only by digitalization but also by artificial intelligence (AI). In undergraduate education, the conventional assessment of tooth preparations presents several structural limitations, such as inter-rater variability, the scarcity of immediate and individualized feedback, and the absence of standardized metrics, which compromise the equity and effectiveness of learning. Objectives: To develop and subsequently demonstrate the functioning of a digital application capable of analysing three-dimensional models of full-coverage zirconia crown preparations on maxillary premolars, able to automatically compute clinically relevant geometric parameters and provide structured educational feedback, with the aim of improving student performance. Materials and Methods: An applied research and technological development study, descriptive in nature, conducted in a simulated preclinical context. Following the calibration of two specialist examiners in Oral Rehabilitation, 30 frasaco preparations were performed and digitized, constituting the reference set. The assessment is based on a grid of 11 weighted criteria (reductions, technical parameters, and finishing). The thresholds were defined from the Color Atlas of Fixed Prosthodontics, and the tool was demonstrated on a practical preparation. Results: It was confirmed that teaching criteria are convertible into measurable geometric parameters. PrepAIze demonstrated the ability to import STL files, compute the 11 parameters, verify the acceptance ranges, generate the Clinical Score and the Final AI Score, classify the preparation into one of four qualitative levels, and conclude with the production of a structured feedback report. Conclusion: The three hypotheses were confirmed, supporting the technical and functional feasibility of the system. PrepAIze proves to be a promising tool for the objectification and automation of preparation assessment, complementing the role of the teacher and promoting self-directed learning.
Introduction: Dentistry is undergoing a period of transformation driven not only by digitalization but also by artificial intelligence (AI). In undergraduate education, the conventional assessment of tooth preparations presents several structural limitations, such as inter-rater variability, the scarcity of immediate and individualized feedback, and the absence of standardized metrics, which compromise the equity and effectiveness of learning. Objectives: To develop and subsequently demonstrate the functioning of a digital application capable of analysing three-dimensional models of full-coverage zirconia crown preparations on maxillary premolars, able to automatically compute clinically relevant geometric parameters and provide structured educational feedback, with the aim of improving student performance. Materials and Methods: An applied research and technological development study, descriptive in nature, conducted in a simulated preclinical context. Following the calibration of two specialist examiners in Oral Rehabilitation, 30 frasaco preparations were performed and digitized, constituting the reference set. The assessment is based on a grid of 11 weighted criteria (reductions, technical parameters, and finishing). The thresholds were defined from the Color Atlas of Fixed Prosthodontics, and the tool was demonstrated on a practical preparation. Results: It was confirmed that teaching criteria are convertible into measurable geometric parameters. PrepAIze demonstrated the ability to import STL files, compute the 11 parameters, verify the acceptance ranges, generate the Clinical Score and the Final AI Score, classify the preparation into one of four qualitative levels, and conclude with the production of a structured feedback report. Conclusion: The three hypotheses were confirmed, supporting the technical and functional feasibility of the system. PrepAIze proves to be a promising tool for the objectification and automation of preparation assessment, complementing the role of the teacher and promoting self-directed learning.
Descrição
Dissertação para obtenção do grau de Mestre no Instituto Universitário Egas Moniz
Palavras-chave
Inteligência artificial Educação dentária Preparo dentário Avaliação automatizada