Evolution of Research on Artificial Intelligence for Heart Failure: A Bibliometric and Visual Analysis.

PURPOSE

To investigate the role of artificial intelligence in enhancing precise diagnosis, personalized treatment, and efficient monitoring of heart failure over the past two decades and to predict future advancements of these investigations.

METHODS

A literature search was conducted using keywords from the Web of Science database from January 1, 2004, to August 31, 2024, and 684 articles were retrieved. Bibliometric and visual analysis was conducted to examine annual publication volume; and to analyze authors, institutions, countries, journals, references, and keywords. The following tools were utilized for the analysis: Citespace, SCImago Graphica, Microsoft Office Excel, VOSviewer, and Pajek.

RESULTS

The 684 retrieved studies comprised 70 countries/regions, 1550 institutions, and 4610 authors. The annual publishing output increased gradually between 2004 and 2016, and escalated significantly beyond 2017, particularly from 2021 to 2024. This upward trend is anticipated to persist in the future. Sengupta, Partho P., and Shah, Sanjiv J. were the most productive authors. The University of California and Harvard University were the leading institutions in the number of publications within this discipline. The primary nations conducting research in this domain are China and the United States; the United States predominates research impact and global collaboration. Moreover, Frontiers in Cardiovascular Medicine is the leading journal with the most articles published in this area, while Circulation ranks the highest in co-citations. The keywords include HF, machine learning, AI, and diagnosis.

CONCLUSION

The application of AI in HF is a global concern in research. Currently, investigations address AI-enhanced HF diagnosis and risk assessment; AI-powered personalized treatment strategies, remote patient monitoring, multi-omics data integration, and HF mechanisms. Predictably, optimizing the use of AI in the ICU and Multimodal data are future trends in research, with AI substantially facilitating effective management of HF.