A Literature Review: Detection of COVID-19 in Computed Tomography Images Using Deep Learning

Abstract

Through the development of the COVID-19 disease, various diagnosis methods have been studied. One of them is the computed tomography (CT), which has the best level of detail among medical image exams. The CT generates a repeatable and tiring workload, in addition to needing a team that is familiar with the findings that indicate pneumonia caused by COVID-19. To reduce this manual work and collaborate with these teams, several studies have been carried out using deep learning techniques. In this way, this study presents a review of the literature regarding the detection of COVID-19 in CT that uses deep learning to collaborate with a theoretical basis for future works.

Classificação e Segmentação de COVID-19 em Imagens de Tomografia Computadorizada Usando Aprendizado Profundo

“…Com isso, o erro obtido pela rede é menor a cada vez que uma mesma imagem passa pela rede, demonstrando então que a rede está aprendendo a classificar corretamente as imagens de treinamento. e [Marques et al 2022], que mostram que mesmo possuindo arquiteturas mais simples possuem bons resultados. A LeNet-5 foi uma das primeiras redes neurais convolucionais e teve um grande papel para a propagac ¸ão do aprendizado profundo.…”