Deep learning models for early detection of COVID-19 using radiological imaging: A comparative study
Author(s):
Article Type:
Research/Original Article (دارای رتبه معتبر)
Abstract:
Objectives
Artificial intelligence (AI) has the potential to transform various aspects of healthcare. A significant challenge lies in the development of reliable and cost-effective technologies for disease detection. This study aims to create a fully automated system for the early identification of COVID-19 and non-COVID-19 cases using radiological imaging.Methods
This research involved training, analyzing, and testing three existing pre-trained deep learning (DL) models alongside a novel approach for detecting radiographic images. The models used include InceptionV3, ResNet50, and VGG16, which are widely recognized in the field. We also developed a new model, FJCovNet2, designed to expedite disease detection. The proposed DL method follows a straightforward pipeline that includes preprocessing chest images and scan images. The classification was achieved through transfer learning, with the DL model being trained after the data preprocessing phase.Results
In this study, we developed a DL method that effectively extracts features and identifies COVID-19 from radiological images. Our findings indicate that DL models can address previously unrecognized nuances in evaluating radiological images, facilitating early disease identification. Notably, we employed the FJCovNet2 model, based on DenseNet121, to detect COVID-19 using CT scan and X-ray images. When comparing FJCovNet2 with InceptionV3, VGG16, and ResNet50-using the same dataset for training and testing-FJCovNet2 achieved the highest validation accuracy with the shortest training time. The proposed model attained an impressive accuracy of 98.23%.Conclusion
The results are encouraging, demonstrating that these models can accurately detect COVID-19 in radiological images. This suggests that deep learning will play a crucial role in combating the pandemic.Keywords:
Deep Learning , Healthcare , COVID-19 , CT Scan , X-Ray
Language:
English
Published:
International Archives of Health Sciences, Volume:11 Issue: 4, Oct-Dec 2024
Pages:
207 to 214
https://www.magiran.com/p2816257