Mapping the Landscape of Artificial Intelligence in Pneumonia Research: A Bibliometric Analysis (2016–2025)

To address this gap, this bibliometric study systematically maps the global research landscape of artificial intelligence (AI) applications in pneumonia, based on 3,468 publications retrieved from the Web of Science Core Collection (WoS). Using VOSviewer and CiteSpace for co-authorship, co-occurrence, co-citation, and burst detection, we found that annual output grew from 20 to 725 papers and citations from 12 to 19,503 between 2016 and 2025, with a marked acceleration after 2020.

China dominated in publication volume, followed by the United States and India, yet the United Kingdom and Australia achieved substantially higher citations per paper, revealing a productivity-impact divide. International collaboration was intensive, particularly between China and the US. Thematic evolution shifted from early work on radiation pneumonitis and fibrosis toward deep learning becoming central around 2019, then a COVID-19 surge after 2020, and more recently an expansion into clinical management and validation.