Loading...
Publication
Contributo da Inteligência Artificial para classificação de imagem em estudos de perfusão do miocárdio: uma revisão sistemática
| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 360.34 KB | Adobe PDF |
Advisor(s)
Abstract(s)
Introdução – Os métodos de Inteligência Artificial (IA) como o deep learning (DL) e, em particular, as convolutional neural networks (CNN) caracterizam-se por executar tarefas que normalmente requerem cognição humana. Nos estudos de perfusão do miocárdio, os sistemas de IA têm assumido importância como ferramenta auxiliar aos especialistas de medicina nuclear, particularmente para classificação de imagem no âmbito da doença arterial coronária (DAC). Objetivos – Avaliar o contributo dos métodos de DL para classificação de imagem em estudos de perfusão do miocárdio. Métodos – Realizou-se uma revisão sistemática, onde foram incluídos 11 artigos, pesquisados nas bases de dados PubMed, Web of Science e Scopus. Incluíram-se estudos publicados nos últimos cinco anos, que procuram avaliar o desempenho dos métodos de DL para classificação de imagem em estudos de perfusão do miocárdio em contexto de DAC. Resultados – Dos 11 artigos incluídos, 82% utilizaram CNN e os restantes 18% aplicaram outros métodos de DL. A arquitetura Red-Green-Blue Convolutional Neural Network (RGB-CNN) demonstrou maior aplicabilidade, correspondendo a 45% das CNN utilizadas e apresentou melhor desempenho (AUC=94,58%). As restantes arquiteturas utilizadas corresponderam a 11% cada (CNN 2D; Inception V3; ResNet152V2; CNN Hand-Crafted e ResNet50). Conclusão – Os métodos de IA, nomeadamente o DL com recurso a CNN, demonstram ser benéficos para a classificação de imagens de perfusão do miocárdio, com potencial de aplicação no diagnóstico precoce de DAC, embora com necessidade de investigação futura.
ABSTRACT Introduction – Artificial Intelligence (AI) methods such as deep learning (DL) and, in particular, convolutional neural networks (CNN), are characterized by performing tasks that normally require human cognition. In myocardial perfusion studies, AI systems have assumed importance as an ancillary tool for nuclear medicine (NM) specialists, particularly for image classification in the context of coronary artery disease (CAD). Objectives – Evaluate DL methods’ contribution to image classification in myocardial perfusion studies. Methods – A systematic review was carried out, which included 11 articles, searched in databases, PubMed, Web of Science, and Scopus. Studies published in the last five years that seek to evaluate the performance of DL methods for image classification in myocardial perfusion studies in the context of CAD were included. Results – Eleven articles were included in the systematic review, where 82% used CNN and the remaining 18% applied other DL methods. The Red-Green-Blue Convolutional Neural Network (RGB-CNN) architecture demonstrated greater applicability, corresponding to 45% of the CNN used, and presented better performance (AUC=94.58%). The remaining architectures corresponded to 11% each (CNN 2D; Inception V3; ResNet152V2; CNN Hand-Crafted, and ResNet50). Conclusion – AI methods, namely DL using CNN, prove to be beneficial for the classification of myocardial perfusion images, with potential application in the early diagnosis of CAD, although with the need for further investigation.
ABSTRACT Introduction – Artificial Intelligence (AI) methods such as deep learning (DL) and, in particular, convolutional neural networks (CNN), are characterized by performing tasks that normally require human cognition. In myocardial perfusion studies, AI systems have assumed importance as an ancillary tool for nuclear medicine (NM) specialists, particularly for image classification in the context of coronary artery disease (CAD). Objectives – Evaluate DL methods’ contribution to image classification in myocardial perfusion studies. Methods – A systematic review was carried out, which included 11 articles, searched in databases, PubMed, Web of Science, and Scopus. Studies published in the last five years that seek to evaluate the performance of DL methods for image classification in myocardial perfusion studies in the context of CAD were included. Results – Eleven articles were included in the systematic review, where 82% used CNN and the remaining 18% applied other DL methods. The Red-Green-Blue Convolutional Neural Network (RGB-CNN) architecture demonstrated greater applicability, corresponding to 45% of the CNN used, and presented better performance (AUC=94.58%). The remaining architectures corresponded to 11% each (CNN 2D; Inception V3; ResNet152V2; CNN Hand-Crafted, and ResNet50). Conclusion – AI methods, namely DL using CNN, prove to be beneficial for the classification of myocardial perfusion images, with potential application in the early diagnosis of CAD, although with the need for further investigation.
Description
Keywords
Inteligência artificial Deep learning Convolutional neural networks PET SPECT Medicina nuclear Classificação Doença arterial coronária Perfusão do miocárdio Artificial Intelligence Nuclear medicine Classification Coronary artery disease Myocardial perfusion
Citation
Cardoso M, Santos V, Figueiredo S. Contributo da Inteligência Artificial para classificação de imagem em estudos de perfusão do miocárdio: uma revisão sistemática. Saúde & Tecnologia. 2024;(30):e790.
Publisher
Escola Superior de Tecnologia da Saúde de Lisboa (Instituto Politécnico de Lisboa)