Page 107 - IJEEE-2022-Vol18-ISSUE-1

Received: 2 March 2022 Revised: 31 March 2022 Accepted: 31 March 2022
DOI: 10.37917/ijeee.18.1.12
Vol. 18| Issue 1| June 2022
? Open Access

Iraqi Journal for Electrical and Electronic Engineering

Original Article

Expanding New Covid-19 Data with Conditional
Generative Adversarial Networks

Haneen Majid*, Khawla Hussein Ali
Computer Science, College of Education, Basrah, Iraq

Correspondence
*Haneen Majid
Computer Science, College of Education,
University of Basrah, Basrah, Iraq
Email: hneenaltmemi@gmail.com

Abstract
COVID-19 is an infectious viral disease that mostly affects the lungs. That quickly spreads across the world. Early detection
of the virus boosts the chances of patients recovering quickly worldwide. Many radiographic techniques are used to diagnose
an infected person such as X-rays, deep learning technology based on a large amount of chest x-ray images is used to diagnose
COVID-19 disease. Because of the scarcity of available COVID-19 X-rays image, the limited COVID-19 Datasets are
insufficient for efficient deep learning detection models. Another problem with a limited dataset is that training models suffer
from over-fitting, and the predictions are not generalizable to address these problems. In this paper, we developed Conditional
Generative Adversarial Networks (CGAN) to produce synthetic images close to real images for the COVID-19 case and
traditional augmentation that was used to expand the limited dataset then used to train by Customized deep detection model.
The Customized Deep learning model was able to obtain excellent detection accuracy of 97% accurate with only ten epochs.
The proposed augmentation outperforms other augmentation techniques. The augmented dataset includes 6988 high-quality
and resolution COVID-19 X-rays images. At the same time, the original COVID-19 X-rays images are only 587.
KEYWORDS: deep learning, COVID-19, Augmentation, Generative Adversarial Network, Conditional Generative
Adversarial Network, image synthetic.

I. INTRODUCTION COVID-19 uses fewer resources than classic RRT-PCR [6],
[7]. PCR has limited sensitivity for COVID-19 detection.
The people of China's regions were contaminated in
2003 by (severe acute respiratory syndrome) a particular type Because COVID-19 virus's widespread change in
of severe disease of the lungs (SARS) later renamed (SARS- approaches, medical practitioners may lack the necessary
COV1) [1]. In December of 2019, COVID-19 was detected knowledge to execute proper diagnoses based on medical
for the first time in Wuhan. It soon spread throughout the field images; chest radiography (X-rays) with computed
world, killing millions of people. The pandemic was tomography (CT) are two examples. Consequently,
announced in late March of 2020 by the Organization of the practitioners may use artificial intelligence (AI) and deep
Health World, and on 21 July, across 188 nations worldwide, learning technologies to help automate COVID-19
there have been nearly 14 million verified infections and diagnostic processes. Deep learning techniques play an
609,198 fatalities [2]. Coronavirus illness is a respiratory essential role in medical image Developing based on
infection caused by (SARS-CoV2). The coronavirus family diagnostic tools for COVID-19. The diagnostics of medical
includes alpha (a) followed by beta (ß), gamma (?), and delta images can deliver accurate and concise analytical findings
(d) coronaviruses, as well as omicron, until now. In 2003, a using deep learning techniques. The lack of COVID-19-
SARS coronavirus epidemic afflicted 26 nations, resulting in related medical imaging data, like X-rays and CT images, is
over 8000 cases. SARS-CoV-2 (COVID-19) has infected one of the most critical issues; the dataset that is accessible
about 1.5 million people in 150 countries, with a fatality rate is often limited in size. So, we need to extend these data
of 6% as of this writing. SARS-CoV-2 has a higher safely to get excellent and accurate detection results. Deep
transmission rate than SARS coronavirus [3][4]. The most learning models may be used to detect COVID-19[8]. Deep
common symptoms of Covid-19 are a dry cough, a loss of learning algorithms for detecting COVID-19 show
appetite, liver damage, and weariness [5][6]. promising results; such models should be viewed cautiously
COVID-19 may be identified in different ways, like PCR [6], since they are based on a small sample of data. When the
Chest computed tomography (CT) imaging, and chest deep learning models are trained on a limited dataset, the
radiography (X-rays). These methods are more sensitive to critical problem is that they are prone to overfitting. To

This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and
reproduction in any medium, provided the original work is properly cited.
© 2022 The Authors. Iraqi Journal for Electrical and Electronic Engineering by College of Engineering, University of Basrah.

https://doi.org/10.37917/ijeee.18.1.12 https://www.ijeee.edu.iq 103