Mapping the Landscape of Research on Artificial Intelligence, Machine Learning, and Deep Learning Applications in COVID-19 Detection: A Web of Science Analysis

Abstract

Purpose: The objective the present research study was to draw the landscape of AI, ML, and DL applications in COVID-19 detection by examining the quantity, quality, and structural indicators of publications listed in the ISI Web of Science. Design, Methodology, Approach: The study's objectives were attained by extracting data from the ISI Web of Science (WOS) database, followed by deep cleaning using Endnote and manual verification. Biblioshiny and VOSviewer tools were utilized to evaluate the refined data, covering both quantitative and qualitative dimensions. This comprehensive analysis included examining publication output, citation metrics, and country-specific productivity, as well as assessing citation impact through indicators like h-index, g-index, and m-index. Moreover, the study assessed conformity to Bradford's Law and Lotka's Law, and visualized research trends, collaborations, and thematic clusters. Findings: The study's findings show that a significant number of articles were co-authored. The most networked author belonged to Hong Kong. China leads in terms of output, influence, and collaboration. IEEE Access and Computers in Biology and Medicine remained the most prominent source for papers on the phenomenon under investigation. Originality Value: The research is the first to identify the amount (frequency), quality (impact), and structural indicators (correlations) of AI/ML/DL in COVID-identification. The results hold significant importance for scholars and nations seeking to enhance their comprehension of the phenomena of artificial intelligence, machine learning, deep learning, and COVID-19 detection. The results are also beneficial to the authors and nations hoping to increase international research collaboration.