Loading...
Publication
A inteligência artificial para a análise automática de imagens CBCT/TC: aplicações no planeamento de implantes dentários – estudo comparativo de um caso clínico
| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| Projeto de pós-graduação_40413 | 5.56 MB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
O plano de tratamento para a colocação de implantes dentários é uma etapa essencial para o sucesso da reabilitação de pacientes parcialmente ou totalmente desdentados. O planeamento preciso dos implantes é facilitado pela utilização de imagens CBCT (Tomografia Computorizada de Feixe Cónico) ou TC, que fornecem uma visão tridimensional detalhada das estruturas maxilares. No entanto, a análise meramente observacional destas imagens está sujeita a erros de interpretação. O objetivo principal desta tese foi comparar as decisões e os resultados da colocação de implantes dentários orientada por IA com aqueles realizados por um profissional sem o recurso à IA, de modo a avaliar o potencial da IA no planeamento de implantes. Para tal, procurámos utilizar técnicas de IA capazes de automatizar a análise das imagens CBCT ou TC, detetando as estruturas anatómicas, avaliando a densidade e a disponibilidade óssea e simulando virtualmente o posicionamento ideal dos implantes.
No entanto, apesar de este objetivo principal ter sido estabelecido, não foi totalmente alcançado devido à falta de software capaz de realizar esta tarefa de forma completa até ao momento. Os softwares de IA utilizados permitem atualmente apenas a segmentação das imagens, limitando a possibilidade de um plano de tratamento totalmente autónomo e impedindo a comparação direta com as decisões humanas. Dado que o objetivo principal não era realizável, definimos um objetivo secundário: utilizar um dos softwares de IA mais desenvolvidos atualmente para a análise das imagens CBCT, avaliando as suas capacidades e limitações. Para tal, utilizámos o software Diagnocat, que permitiu uma segmentação automatizada das imagens CBCT e possibilitou uma comparação da análise das imagens CBCT realizada pelo dentista e pelo software de IA.
Em conclusão, o uso da inteligência artificial para a análise automática de imagens CBCT ou TC no planeamento de implantes dentários representa um avanço significativo para a odontologia, abrindo caminho para práticas mais eficazes e seguras. No entanto, embora a IA traga inovações significativas, ainda apresenta limitações, nomeadamente na capacidade de realizar um plano de tratamento completo sem intervenção humana. É, portanto, crucial salientar que os softwares de IA utilizados aqui permitem apenas a segmentação automática, o que limita a concretização do nosso objetivo principal.
The treatment plan for dental implant placement is an essential step for the successful rehabilitation of partially or fully edentulous patients. Accurate implant planning is facilitated by the use of CBCT (Cone Beam Computed Tomography) or CT images, which provide a detailed three-dimensional view of maxillary structures. However, a purely observational analysis of these images is subject to interpretative errors. The primary objective of this thesis was to compare the decisions and outcomes of dental implant placement guided by AI with those performed by a professional without the use of AI, to assess AI's potential in implant planning. To achieve this, we sought to employ AI techniques capable of automating the analysis of CBCT or CT images, detecting anatomical structures, evaluating bone density and availability, and virtually simulating the optimal positioning of implants. However, despite establishing this primary objective, it was not fully achieved due to the lack of software capable of performing this task comprehensively to date. The AI software currently available only allows for image segmentation, limiting the possibility of a fully autonomous treatment plan and preventing a direct comparison with human decisions. Given that the primary objective was not achievable, we defined a secondary objective: to use one of the most advanced AI software currently available to analyze CBCT images, assessing its capabilities and limitations. For this purpose, we used the Diagnocat software, which enabled automated segmentation of CBCT images and allowed for a comparison between the CBCT image analysis conducted by the dentist and that performed by the AI software. In conclusion, the use of artificial intelligence for the automatic analysis of CBCT or CT images in dental implant planning represents a significant advancement in dentistry, paving the way for more effective and safer practices. However, although AI brings significant innovations, it still presents limitations, particularly in its ability to perform a complete treatment plan without human intervention. It is, therefore, crucial to highlight that the AI software used here only allows for automated segmentation, which limits the achievement of our primary objective.
The treatment plan for dental implant placement is an essential step for the successful rehabilitation of partially or fully edentulous patients. Accurate implant planning is facilitated by the use of CBCT (Cone Beam Computed Tomography) or CT images, which provide a detailed three-dimensional view of maxillary structures. However, a purely observational analysis of these images is subject to interpretative errors. The primary objective of this thesis was to compare the decisions and outcomes of dental implant placement guided by AI with those performed by a professional without the use of AI, to assess AI's potential in implant planning. To achieve this, we sought to employ AI techniques capable of automating the analysis of CBCT or CT images, detecting anatomical structures, evaluating bone density and availability, and virtually simulating the optimal positioning of implants. However, despite establishing this primary objective, it was not fully achieved due to the lack of software capable of performing this task comprehensively to date. The AI software currently available only allows for image segmentation, limiting the possibility of a fully autonomous treatment plan and preventing a direct comparison with human decisions. Given that the primary objective was not achievable, we defined a secondary objective: to use one of the most advanced AI software currently available to analyze CBCT images, assessing its capabilities and limitations. For this purpose, we used the Diagnocat software, which enabled automated segmentation of CBCT images and allowed for a comparison between the CBCT image analysis conducted by the dentist and that performed by the AI software. In conclusion, the use of artificial intelligence for the automatic analysis of CBCT or CT images in dental implant planning represents a significant advancement in dentistry, paving the way for more effective and safer practices. However, although AI brings significant innovations, it still presents limitations, particularly in its ability to perform a complete treatment plan without human intervention. It is, therefore, crucial to highlight that the AI software used here only allows for automated segmentation, which limits the achievement of our primary objective.
Description
Keywords
Inteligência artificial em medicina dentária CBCT Diagnóstico Implantes Diagnocat Artificial intelligence in dentistry Diagnostic Implants