Abstract
This study aimed to map the evolution and impact of artificial intelligence (AI) in diabetes prediction research from 2013 to 2023. Utilizing Scopus database records, a bibliometric analysis was conducted on documents featuring AI and diabetes prediction keywords. The analysis used the Bibliometrix and VOSviewer tools to evaluate research publication trends, author collaboration, and keyword co-occurrence in the application of AI in diabetes prediction. Data screening, focusing on specific terms in titles and abstracts, ensured the relevance of the documents. The study included diverse document types and subject areas, reflecting the field's multidisciplinary nature. The findings revealed a significant annual growth rate of 84.86% in AI applications for diabetes prediction, with 1 498 documents from 802 sources highlighting strong scholarly interest. A peak in citation impact in 2018 marks key contributions and diverse research themes in that year. International co-authorship, notably from the USA, India, China, and Saudi Arabia, underscores extensive collaboration. Thematic analysis points to focal areas like ophthalmology in diabetes-related complications and identifies central topics and emerging trends, including the Internet of Things. The biblio-metric review highlights a significant interdisciplinary expansion in AI research applied to diabetes prediction, with a marked increase in global collaborations and contributions. The study underscores the importance of AI in enhancing diabetes diagnostics and management, indicating a promising trajectory for future research, healthcare policy, and clinical practice. The evolution of AI, particularly machine learning, in diabetes prediction, demonstrates the potential for innovative solutions in managing this chronic condition.