Home > Published Issues > 2024 > Volume 15, No. 5, 2024 >
JAIT 2024 Vol.15(5): 622-629
doi: 10.12720/jait.15.5.622-629

Detection of COVID-19 Infection Using Deep Neural Network and Machine Learning Technique

M. Hema * and T. S. N. Murthy
Department of Electronics and Communication Engineering, Jawaharlal Nehru Technological University, Gurajada, Vizianagaram, India
Email: mhema.jntugvcev@gmail.com (M.H.); tsnmurthyece.jntuk@ieee.org (T.S.N.M.)
*Corresponding author

Manuscript received July 14, 2023; revised August 2, 2023; accepted November 7, 2023; published May 16, 2024.

Abstract—In the growing population expansion, automated illness identification is one of the critical subjects in the field of medical. Recently new virus named Coronavirus (COVID-19) has emerged and created a severe threat to lives and the rate of spreading is severe around the world. As the quickest diagnostic alternative, an automatic detection system should be built to detect the virus and restrict the persons to isolation and stop spreading. The objective of the work is to present a deep learning and machine learning approach combined to autonomously recognize the presence of COVID-19 using the chest X-ray images. In this system, Convolution Neural Network (CNN) with MobileNet V2 network and DesNet is used for extraction of features deeply. The extracted features are fed to classifier. The classification is performed using Multilayer Perceptron (MLP) and Support Vector Machine (SVM). The results obtained using CNN-MLP and CNN-SVM is compared. The parameters like Accuracy, F1-score, recall and precision are evaluated. The accuracy using proposed CNN-SVM is 99.18% whereas for CNN-MLP the rate of accuracy is 98.68% using MobileNet-V2. The experimental results are evaluated using Matlab tool.
Keywords—X-ray data, COVID-19, neural networks, Convolution Neural Network (CNN), Support Vector Machine (SVM)

Cite: M. Hema and T. S. N. Murthy, "Detection of COVID-19 Infection Using Deep Neural Network and Machine Learning Technique," Journal of Advances in Information Technology, Vol. 15, No. 5, pp. 622-629, 2024.

Copyright © 2024 by the authors. This is an open access article distributed under the Creative Commons Attribution License (CC BY-NC-ND 4.0), which permits use, distribution and reproduction in any medium, provided that the article is properly cited, the use is non-commercial and no modifications or adaptations are made.