Your browser doesn't support javascript.
Screening Covid-19 Infection from Chest CT Images using Deep Learning Models based on Transfer Learning
6th International Conference on Computing Methodologies and Communication, ICCMC 2022 ; : 1358-1363, 2022.
Article in English | Scopus | ID: covidwho-1840254
ABSTRACT
As the global epidemic of Covid19 progresses, accurate diagnosis of Covid19 patients becomes important. The biggest problem in diagnosing test-positive people is the lack or lack of test kits due to the rapid spread of Covid19 in the community. As an alternative rapid diagnostic method, an automated detection system is needed to prevent Covid 19 from spreading to humans. This article proposes to use a convolutional neural network (CNN) to detect patients infected with coronavirus using computer tomography (CT) images. In addition, the transfer learning of the deep CNN model VGG16 is investigated to detect infections on CT scans. The pretrained VGG16 classifier is used as a classifier, feature extractor, and fine tuner in three different sets of tests. Image augmentation is used to boost the model's generalization capacity, while Bayesian optimization is used to pick optimum values for hyperparameters. In order to fine-tune the models and reduce training time, transfer learning is being researched. Surprisingly, all of the proposed models scored greater than 93% accuracy, which is on par with or better than previous deep learning models. The results show that optimization improved generalization in all models and highlight the efficacy of the proposed strategies. © 2022 IEEE.
Keywords

Full text: Available Collection: Databases of international organizations Database: Scopus Language: English Journal: 6th International Conference on Computing Methodologies and Communication, ICCMC 2022 Year: 2022 Document Type: Article

Similar

MEDLINE

...
LILACS

LIS


Full text: Available Collection: Databases of international organizations Database: Scopus Language: English Journal: 6th International Conference on Computing Methodologies and Communication, ICCMC 2022 Year: 2022 Document Type: Article