RevisiĆ³n narrativa sobre el uso de inteligencia artificial en ortodoncia
No hay miniatura disponible
Archivos
Fecha
2023
Profesor/a GuĆa
Facultad/escuela
Idioma
es
TĆtulo de la revista
ISSN de la revista
TĆtulo del volumen
Editor
Universidad AndrƩs Bello
Nombre de Curso
Licencia CC
Licencia CC
Resumen
El desarrollo de los sistemas de Inteligencia Artificial (IA) se ha generalizado en el sector
sanitario y tiene el potencial de generar mejoras significativas en la atenciĆ³n y los
resultados de los pacientes. Este artĆculo tuvo como objetivo proporcionar informaciĆ³n
sobre las aplicaciones actualmente disponibles de IA en ortodoncia basƔndose en una
revisiĆ³n narrativa de la literatura actual. Fueron 146 trabajos encontrados en total en la
base de datos y luego de haber realizado los criterios de inclusiĆ³n y exclusiĆ³n, 54
publicaciones quedaron seleccionadas. Los resultados arrojaron que las Ɣreas
encontradas son: 1) extracciones dentarias 2) AnƔlisis cefalomƩtrico 3) Estado de
maduraciĆ³n cervical 4) PredicciĆ³n de crecimiento mandibular. Para la toma de decisiĆ³n
de extracciones dentarias, se encontrĆ³ que el modelo de aprendizaje automĆ”tico con
mayor precisiĆ³n fue el SVM. Sin embargo, el modelo NN y LR resultaron tener un poder
predictivo equivalente al del modelo SVM. El modelo menos preciso fue el RF. En el
estado de madurez cervical, se menciona el modelo de gemelo digital de CVM que se
basa en blockchain implementado en un Metaverso. En la proyecciĆ³n de crecimiento
mandibular se han aplicado varios modelos para comparar las velocidades medias
anuales de crecimiento con un modelo matemƔtico basado en la curva de crecimiento de
una poblaciĆ³n. En conclusiĆ³n, la aplicaciĆ³n de inteligencia artificial cada dĆa toma mĆ”s
fuerza en el Ɣrea de la ortodoncia, pero, aunque la IA ha demostrado ser una herramienta
para facilitar el trabajo al profesional, la falta de investigaciones con base de datos
amplios y significativamente representativos conllevan a finalmente la decisiĆ³n en cuanto
a diagnĆ³stico y tratamiento sean siempre responsabilidad del profesional.
The development of Artificial Intelligence (AI) systems has become widespread in the healthcare sector and has the potential to generate significant improvements in patient care and outcomes. This article aimed to provide information on currently available applications of AI in orthodontics based on a narrative review of current literature. There were 146 works found in total in the database and after having carried out the inclusion and exclusion criteria, 54 publications were selected. The results showed that the areas found are: 1) dental extractions 2) Cephalometric analysis 3) State of cervical maturation 4) Prediction of mandibular growth. For the decision making of tooth extractions, it was found that the machine learning model with the highest accuracy was the SVM. However, the NN and LR models were found to have equivalent predictive power to that of the SVM model. The least accurate model was the RF. In the state of cervical maturity, the CVM digital twin model is mentioned which is based on blockchain implemented in a Metaverse. In the projection of mandibular growth, several models have been applied to compare the average annual growth rates with a mathematical model based on the growth curve of a population. In conclusion, the application of artificial intelligence is gaining more strength every day in the area of orthodontics, but although AI has proven to be a tool to facilitate the professional's work, the lack of research with a large and significantly representative database leads to Finally, the decision regarding diagnosis and treatment is always the responsibility of the professional.
The development of Artificial Intelligence (AI) systems has become widespread in the healthcare sector and has the potential to generate significant improvements in patient care and outcomes. This article aimed to provide information on currently available applications of AI in orthodontics based on a narrative review of current literature. There were 146 works found in total in the database and after having carried out the inclusion and exclusion criteria, 54 publications were selected. The results showed that the areas found are: 1) dental extractions 2) Cephalometric analysis 3) State of cervical maturation 4) Prediction of mandibular growth. For the decision making of tooth extractions, it was found that the machine learning model with the highest accuracy was the SVM. However, the NN and LR models were found to have equivalent predictive power to that of the SVM model. The least accurate model was the RF. In the state of cervical maturity, the CVM digital twin model is mentioned which is based on blockchain implemented in a Metaverse. In the projection of mandibular growth, several models have been applied to compare the average annual growth rates with a mathematical model based on the growth curve of a population. In conclusion, the application of artificial intelligence is gaining more strength every day in the area of orthodontics, but although AI has proven to be a tool to facilitate the professional's work, the lack of research with a large and significantly representative database leads to Finally, the decision regarding diagnosis and treatment is always the responsibility of the professional.
Notas
Tesina (Programa de EspecializaciĆ³n en Ortodoncia y Ortopedia Dentomaxilofacial)
Palabras clave
Ortodoncia, Innovaciones TecnolĆ³gicas, Inteligencia Artificial