TY - GEN
T1 - Deep Transfer Learning Model-Based Automated Detection of COVID-19 from X-ray Images and Interpretation of COVID-19 Images Using GLCM Texture Features
AU - Ankalaki, Shilpa
AU - Shorya, Kartikeya
AU - Majumdar, Jharna
N1 - Publisher Copyright:
© 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
PY - 2022
Y1 - 2022
N2 - The novel coronavirus 2019 (COVID-2019), which initially proved its existence in Wuhan city of China in December 2019, spread quickly around the globe and turned into a pandemic. It has caused a staggering impact on all the sectors of the world like public health, global economy and daily lives. COVID-19 positive cases and death due to COVID-19 are rapidly increasing day by day. It is crucial and essential for fast and accurate automatic detection of COVID-19 infection to make better decisions and to provide appropriate treatment for the patients that can hopefully save their lives. The current has employed the VGG19, RESNET50 and DesNet121 deep learning convolutional neural network with transfer learning to identify and classifies the X-ray images into COVID-19 and non-COVID-19 classes. This study has been extended to analyse the factors which distinguishes COVID-19 and non-COVID-19 images. To accomplish this task, we have employed GLCM features and determined that variance is the best feature for this purpose. GRAD-CAM algorithm has been used to interpret the decision of CNN architecture. In this study, VGG19 and DenseNet121 achieved the classification accuracy of 98.80%, and ResNet50 achieved the accuracy of 97.65% for binary classes (COVID-19 and non-COVID-19 classes).
AB - The novel coronavirus 2019 (COVID-2019), which initially proved its existence in Wuhan city of China in December 2019, spread quickly around the globe and turned into a pandemic. It has caused a staggering impact on all the sectors of the world like public health, global economy and daily lives. COVID-19 positive cases and death due to COVID-19 are rapidly increasing day by day. It is crucial and essential for fast and accurate automatic detection of COVID-19 infection to make better decisions and to provide appropriate treatment for the patients that can hopefully save their lives. The current has employed the VGG19, RESNET50 and DesNet121 deep learning convolutional neural network with transfer learning to identify and classifies the X-ray images into COVID-19 and non-COVID-19 classes. This study has been extended to analyse the factors which distinguishes COVID-19 and non-COVID-19 images. To accomplish this task, we have employed GLCM features and determined that variance is the best feature for this purpose. GRAD-CAM algorithm has been used to interpret the decision of CNN architecture. In this study, VGG19 and DenseNet121 achieved the classification accuracy of 98.80%, and ResNet50 achieved the accuracy of 97.65% for binary classes (COVID-19 and non-COVID-19 classes).
UR - https://www.scopus.com/pages/publications/85121749080
UR - https://www.scopus.com/inward/citedby.url?scp=85121749080&partnerID=8YFLogxK
U2 - 10.1007/978-981-16-1342-5_45
DO - 10.1007/978-981-16-1342-5_45
M3 - Conference contribution
AN - SCOPUS:85121749080
SN - 9789811613418
T3 - Lecture Notes in Electrical Engineering
SP - 581
EP - 598
BT - Emerging Research in Computing, Information, Communication and Applications, ERCICA 2020
A2 - Shetty, N. R.
A2 - Patnaik, L. M.
A2 - Nagaraj, H. C.
A2 - Hamsavath, Prasad N.
A2 - Nalini, N.
PB - Springer Science and Business Media Deutschland GmbH
T2 - 6th International Conference on Emerging Research in Computing, Information, Communication and Applications, ERCICA 2020
Y2 - 25 September 2020 through 26 September 2020
ER -
