A Deep Convolutional Neural Network for COVID-19 Chest CT-Scan Image Classification

• Published in: High Performance Computing and Networking Vol. 853; pp. 603 - 612
• Main Authors: Kumari, L. Kanya, Jagadesh, B. Naga
• Format: Book Chapter
• Language: English
• Published: Singapore Springer Singapore Pte. Limited 2022; Springer Singapore
• Series: Lecture Notes in Electrical Engineering
• Subjects: Artificial intelligence; Convolutional neural networks; CT-images; Deep learning; Machine learning
• ISBN: 9789811698842; 9811698848
• ISSN: 1876-1100; 1876-1119
• DOI: 10.1007/978-981-16-9885-9_49

A novel corona virus (COVID-19) is a new dangerous disease which affects the global economic growth and challenge to the doctors and scientists. This disease is escalating gradually which impacts world’s financial system at risk. Due to increase in COVID-19, the role of artificial intelligence, machine learning, and deep learning is crucial in this situation. Deep learning is a dominant tool to control this pandemic outbreak by predicting the disease in advance. Deep learning techniques deal with several types of data sources that put together to form the user-friendly platforms for physicians and researchers. The proposed methodology is based on convolutional neural network which classifies the COVID-19 chest CT-scan images into infected or not infected. We have done the experiment on publicly available dataset in GitHub which consists of 360 positive and 397 negative chest CT-images which are collected from 216 patients. In our proposed CNN model, we used Adam optimizer with learning rate 0.001 and obtained the classification accuracy 88.4%. The experimental results show that our methodology can handle current pandemic situation in a better manner.