Healthcare analytics perspectives on artificial intelligence–based retinopathy of prematurity screening: a bibliometric mapping study

The integration of artificial intelligence (AI) into retinopathy of prematurity (ROP) screening represents an important advancement in neonatal ophthalmology, particularly in settings facing specialist shortages and screening infrastructure constraints. Despite the rapid expansion of research in this domain, a structured evaluation of its intellectual landscape, geographic distribution, and collaboration patterns remains limited. This study presents a bibliometric mapping analysis of global research on AI-based ROP screening. A total of 55 publications indexed in the Scopus database were identified through a reproducible search strategy and analyzed using Bibliometrix and VOSviewer. The retrieved publications span the period 2017–2025. The results demonstrate a marked upward trajectory in publications, with a peak of 16 publications in 2025 and an estimated annual growth rate of 41.42%. The analysis examines publication trends, geographic productivity, institutional collaboration networks, citation patterns, and thematic evolution. The findings indicate strong contributions from Asia and increasing participation from the Middle East, particularly Türkiye, Saudi Arabia, and Egypt. International co-authorship networks reveal structured South–North collaborations linking Middle Eastern countries with major research hubs in the United States and the United Kingdom. Citation analysis shows an average of 18.27 citations per publication. Thematic mapping highlights the dominance of deep learning and retinal image analysis, alongside emerging discussions on healthcare accessibility and deployment in resource-limited settings. These results provide a structured quantitative overview of the development of AI-based ROP screening research and may support future research prioritization and strategic planning in AI-assisted neonatal eye care.