Artificial intelligence for cardiac arrest in the digital health era: From algorithmic performance to workflow-integrated, outcome-oriented digital health systems.
Digital health · 2026-01-01 · Other / unclear
Abstract
OBJECTIVES: To characterize the global research landscape, collaboration patterns, and thematic evolution of artificial intelligence (AI) in cardiac arrest care. METHODS: We conducted a bibliometric analysis of AI and cardiac arrest research using the Web of Science Core Collection and Scopus. We retrieved records on October 29, 2025, limited to English-language articles and reviews. After deduplication, 1,228 publications were included. CiteSpace, VOSviewer, and bibliometrix (R) assessed publication growth, country/institution contributions, collaboration networks, co-citation structures, and keyword dynamics. RESULTS: We included 1,228 publications (2012-2025) with a compound annual growth rate (CAGR) of 22.05%. The corpus comprised 6,994 authors (mean of 7.46 per paper), and international co-authorship accounted for 13.36% of all documents. Forty-eight countries contributed, with the United States leading (134 publications; 10.9%), followed by South Korea (101) and China (77). Seoul National University was the most productive institution (92), followed by Harvard University (60). Aramendi E. ranked first among authors (21), followed by Park J. (17) and Ong M.E.H. (16). Resuscitation ranked first among journals, with 54 publications and 1,350 citations. Keywords shifted from method-focused topics to more clinical, system- and outcome-focused themes. Burst and clustering analyses emphasized early warning/prediction and neurological outcomes, with recent burst terms including "Cerebral Performance Category (CPC)," "brain injury," and "emergency medical services (EMS)." CONCLUSIONS: AI-cardiac arrest research is entering a maturing expansion phase characterized by interdisciplinary linkages and a multipolar collaboration structure. The field is shifting beyond algorithmic performance toward transportable, workflow-aware, and outcome-oriented digital health systems.