Artificial Intelligence in intensive care medicine: Bibliometric Analysis (Preprint)

Abstract

Interest in critical care-related Artificial Intelligence (AI) research is growing rapidly. However, the literature is still lacking in comprehensive bibliometric studies that measure and analyze scientific publications globally.

This study's objective is to assess the global research trends in artificial intelligence (AI) in intensive care medicine based on publication outputs, citations, co-authorships between nations, and co-occurrences of author keywords.

3619 documents published up to March 2022 were retrieved from the Scopus database. After selecting document type as articles, the title and abstract are checked for eligibility. For the final bibliometric study using Vosviwer, 1198 papers were included. The growth rate of publications, preferred journals, leading research countries, international collaboration, and top institutions was computed.

The number of publications increased steeply between 2018 and 2022, accounting for 72.54% (869/1198) of all included papers. USA and China contributed to about 55% of total publications. Nine out of fifteen most productive institutions were among top 100 universities worldwide. Detecting clinical deterioration, monitoring, predicting disease progression, mortality, prognosis, and classifying disease phenotype or subtype are some of the research hotspots for AI in critically ill patients. Neural networks, decision support systems, machine learning, and deep learning were all commonly utilized AI technology.

This study highlights popular research areas in AI research aimed at improving health care in ICUs, offers a comprehensive look at the research trend in AI application in the ICU, and provides insight into potential collaboration and prospects for future research. The 30 articles that received the most citations were listed in detail. For AI-based clinical research to be convincing enough for routine critical care practice, collaborative research efforts are needed to increase the maturity and robustness of AI-driven models.

N.A.