Application of artificial intelligence in heart diseases: A bibliometrics review based on CiteSpace analysis

摘要

This study analyzes the research status and trends concerning the global application of AI in heart diseases within the past 36 years, to aid researchers navigate more studies in the future. We acquired publications from to 1986–2021 from PubMed using terms from Medical Subject Headings (MeSH) thesaurus. Subsequently, we filtered the type and language of publications in Web of Science and analyzed the data for descriptive statistics, fitting curves, and CiteSpace for visualization analysis.A total of 2038 records related to the application of AI in heart diseases were imported into CiteSpace. There has been a growing number of publications over the past 36 years roughly. Two author clusters focused on application of AI in heart diseases for clinical practice and data-driven medicine.And USA produced the most academic achievements with high impact. Machine learning, neural network, classification, diagnosis, heart failure, atrial fibrillation, and myocardial infarction have become the principal research hotspots and trends in recent years. Generally, AI techniques are increasingly applied to heart diseases in areas ranging from studies on medical data to clinical practice-related robotics-assist. Compared with conventional approaches, AI has been outperforming for heart diseases in terms of surgery operation, identification, detection, prediction, classification, and risk stratification. It has the potential to make healthcare feasible, safe, and accessible.