43-PUB: A Bibliometric Analysis of Machine Learning and Type 2 Diabetes from 2010 to 2023

Objective: Machine learning in the research on type 2 diabetes (T2D) has garnered attention in recent years. Nonetheless, few studies offer a thorough picture of the knowledge generation landscape in this field. This comprehensive bibliometric analysis was conducted to discover global research trends and networks and to emphasize the most prominent countries, institutions, and key topics in this domain.

Methods: Articles on the link between machine learning and T2D from 2010 to 2023 were searched in the Science Citation Index-Expanded of Web of Science Core Collection. The results were classified into categories of prediction, screening and management. CiteSpace software, Vosviewer, the R package "bibliometrix" and the Online Analysis Platform of Literature Metrology were used to analyze the data.

Results: A total of 611 articles were retrieved on the link between machine learning and T2D. The number of publications increased dramatically in the past 5 years, with a sharply increase in the last 3 years. The United States dominated the field until 2021, and the dominance of China started in 2022. The United States was the most frequently involved country in international cooperation. Harvard Medical School was the most productive university. The co-citation keyword clustering labels displayed eight main clusters: precision medicine, counterfactual prediction, risk scoring, diabetic retinopathy, predictive model, gut microbiome, clinical decision support system, and bariatric surgery. The keyword bursts analysis indicated that validation, risk prediction, predictive model, predictor, and metabolomics were the research hotspots with high strength.

Conclusion: Machine learning is now being increasingly applied in the field of prediction, screening and management of T2D, especially over the last years with exponential growth. This bibliometric study provided systematic information and guidance to help explore the application and practice of machine learning in T2D for future research.