Page 35 - AIH-2-1

Artificial Intelligence in Health

ORIGINAL RESEARCH ARTICLE
Deep learning on chest X-ray and computed

tomography scans for detection of COVID-19 as
a part of a network-centric digital health stack

for future pandemics

Ajay Kumar Gogineni 1 , Madapathi Hitesh 1 , Prashant Kumar Jha 2 ,
Soumya Suvashish Sen , Shreeja Das , and Kisor Kumar Sahu 2,4,5 *
2
3
1 School of Electrical Sciences, Indian Institute of Technology, Bhubaneswar, Odisha, India

Abstract

Developing a reliable rapid screening protocol for highly infectious diseases like
COVID-19 is of paramount interest since it facilitates the isolation of infected patients
from the rest of the population. Reverse-transcription polymerase chain reaction
(RT-PCR) test is presently the most widely accepted gold-standard test to detect
*Corresponding author: COVID-19. In this method, the RNA of the virus is duplicated by a process called reverse
Kisor Kumar Sahu
(kisorsahu@iitbbs.ac.in) transcription to form DNA for facilitating the copying process. Fluorescent dye is attached
to the viral genetic material and copied billions of times through the process called
Citation: Gogineni AK, Hitesh M,
Jha PK, Sen SS, Das S, Sahu KK. polymerase chain reaction. Enhanced fluorescence is used to identify the presence of
Deep learning on chest X-ray and genetic material of the virus. These tests are time-consuming and have significant false
computed tomography scans for negatives, i.e., a person with COVID-19 might be categorized as not having the virus.
detection of COVID-19 as a part of
network-centric digital health stack Large-scale RT-PCR testing has its own share of problems such as logistics, availability
for future pandemics. Artif Intell and affordability in underdeveloped nations, and reliability of the test results. Machine
Health. 2025;2(1):29-41. learning algorithms can act as a cheaper supplementary/alternative diagnostic tool
doi: 10.36922/aih.2888 for the testing process. In the current study, using publicly available chest X-ray image
Received: February 5, 2024 datasets, different convolutional neural network (CNN)-based models were developed
1st revised: April 17, 2024 for efficient identification of COVID-19 infected patients, and their efficacies were
compared. Key innovations in training the CNNs are discussed. Our results indicate
2nd revised: May 15, 2024 that EfficientNet, SeResNext, and ResNet are best at classifying normal, pneumonia and
3rd revised: July 3, 2024 COVID-19 cases, respectively. The ResNet architecture with transfer learning performed
Accepted: July 17, 2024 best at detecting COVID-19 with an accuracy of 94%, a rate far superior to that in the
RT-PCR test, which is typically in the range of 70 – 80%. This is particularly attractive as an
Published Online: October 7, 2024 additional noninvasive protocol since such technology-augmented detection is likely
Copyright: © 2024 Author(s). to help in reducing the psychological refractory period due to COVID-19 infections.
This is an Open-Access article Toward the healthy lung initiative in the post-COVID-19 era, we propose close coupling
distributed under the terms of the
Creative Commons Attribution of the present diagnostic protocols with digital approaches to ensure more reliable
License, permitting distribution, personal care within the ambit of large-scale pandemic control mechanisms. Such
and reproduction in any medium, integration with emerging technological tools can create a benchmark for the first line
provided the original work is
properly cited. of defense against future global pandemics.
Publisher’s Note: AccScience
Publishing remains neutral with Keywords: COVID-19; Machine learning; Deep learning; EfficientNet; ResNet; SeResNext;
regard to jurisdictional claims in
published maps and institutional Network-centric digital health stack
affiliations.

Volume 2 Issue 1 (2025) 29 doi: 10.36922/aih.2888