An Enhanced Deep Network for Recognizing the Coronavirus Disease Using X-ray Images

Abstract

The spreading of Coronavirus (covid-19) is pushing the healthcare organizations under exceptional strain over the universe and increasing pressure concurring to the World Health Organization (WHO). With advancement of Artificial Intelligence, the discovery of this type of infection during the initial stage offers assistance in quick recuperation and in discharging the pressure on healthcare organizations. This paper presents the deep convolution network to detect the coronavirus in chest x-ray images. Due to the lack of benchmark datasets for covid-19 specifically in chest x-ray images, this work presents the framework that adopts the Generative Adversarial Network of Deep Convolution (DC-GAN). The proposed model of DC-GAN can generate a maximum of 30 different patterns for a single image and thus increases the size of dataset. It also offers assistance in overwhelming the overfitting issue and building the proposed framework more robust. The proposed model implements the transfer learning updated by the loss function using the ImageNet to attain the finest parameters from the pre-trained model. The publically available dataset is utilized to validate the proposed work. This work is validated by conducting the various experiments in various perspectives and also its performances are recorded using the measures namely accuracy, recall, precision and, F1score. These recorded results compared with the various existing methods..

KEYWORDS:

• Augmentation
• Chest-X-ray
• COVID-19
• Deep convolution
• Transfer learning

DISCLOSURE STATEMENT

No potential conflict of interest was reported by the author.

Additional information

Notes on contributors

D. Vaishnavi

D Vaishnavi is currently working as assistant professor in the Department of Computer Science and Engineering at SRC, SASTRA Deemed University, Tamil Nadu, India. She obtained her doctoral degree in computer science and engineering from Annamalai University in 2017. Her area of doctoral study  is digital image processing. She received her Bachelor of Engineering in information technology and master of engineering in computer science and technology from the same university in 2009 and 2011, respectively. Also, she has published her research works in 35 international journals and conferences in Elsevier, IEEE, Springer, and Inderscience publishers.