Retraction: Deep learning methods to analyse and detect the presence of COVID-19 using X-rays

Abstract

The COVID-19 was discovered in December 2019 in China's Wuhan. and resulted in a major outbreak in several cities in China and spread globally, continuing to have a devastating influence on the world wide population's health and well-being. The virus affects the respiratory system and it is transmitted through close contact of individuals. Efficient screening of infected patients is a crucial step in standing up to COVID-19. As the virus affects the respiratory system, images of chest X-rays are analyzed using deep learning techniques for early detection of COVID-19. The existing deep learning models identify COVID-19 with an accuracy of 79%. The proposed model focuses on detecting COVID-19 in an effective manner, the model includes the following phases: Preprocessing the image using data augmentation and infusing the trained model with different Convolutional Neural Network architectures. The proposed model uses Residual Neural Network architecture (ResNet-152v2), NASNetLarge, and Visual Geometry Group architecture (VGG16), a Convolutional Neural Network (CNN) which are faster when compared to the already existing systems and have an accuracy of 87.50%, 87.50% and 82% respectively.