The application of artificial intelligence in the diagnosis, prediction and monitoring of acute lung injury/acute respiratory distress syndrome: A web of science core collection-based bibliometric analysis of research trends from 2011 to 2026
Pengkun Yuan, Luyao Wang, Xiao Hu, Juanjuan DiaoObjectives
ALI/ARDS is a severe inflammatory lung disease with complex pathogenesis and limited specific therapeutic options. Artificial intelligence (AI) has shown increasing potential in the prediction, diagnosis, phenotypic identification, and monitoring of ALI/ARDS. However, the research landscape and evolving trends within this field remain insufficiently characterized. This study aimed to map the research landscape, major hotspots, and emerging trends in AI-related ALI/ARDS research indexed in the Web of Science Core Collection (WoSCC).
Methods
This WoSCC-based bibliometric study analyzed 210 SCIE-indexed articles on AI and ALI/ARDS using CiteSpace 6.4. R1 (Advanced), VOS Viewer 1.6.19, Scimago Graphica 1.0.39, Microsoft Excel 2021, and the Bibliometrix R package in RStudio, with results enriched by data visualizations.
Results
The results indicate that a total of 22 countries and 107 journals were included in the analysis, with a total of 210 papers published. The annual number of publications increased substantially in 2019 and peaked in 2025. China and the United States were identified as the countries with the highest publication volumes. The core journals included Scientific Reports and Plos One. The authorship network was highly collaborative, reflecting a dynamic balance between long-standing research groups and newly emerging contributors. The keywords analysis revealed that the research focus was concentrated on the following areas: ALI/ARDS prediction models, phenotype identification, deep learning, ICU mortality rate, and bioinformatics analysis. Reference list analysis revealed artificial intelligence and ALI/ARDS detection as the most popular current research topics.
Conclusion
Research activity in AI-related ALI/ARDS has increased, with recent work emphasizing early prediction, multimodal data integration, AI-assisted imaging, phenotype identification and continuous monitoring. These bibliometric patterns indicate growing research attention to clinically oriented applications, but they do not establish clinical effectiveness, implementation readiness or patient benefit. Future studies should prioritize external and prospective validation, transparent reporting, workflow evaluation, interoperability, regulatory considerations and cost-effectiveness.