Abstract / Summary
Stroke remains a leading cause of mortality and long-term disability worldwide, creating an urgent need for accurate risk prediction and early detection. Traditional statistical methods often fail to capture the complex and nonlinear relationships in stroke data, which has led to the rapid adoption of artificial intelligence (AI) and machine learning (ML). Despite significant growth in the field, a comprehensive synthesis of its intellectual structure, collaboration networks, subject evolution, and translational challenges is still lacking. This bibliometric study analyzes 2378 publications indexed in the Web of Science from 1997 to June 2026. The field is expanding rapidly, with an annual growth rate of 21.15% and significant international collaboration. Keyword co-occurrence and thematic analyses reveal dominant research clusters focused on machine learning and deep learning approaches to predict, diagnose, and assess stroke outcome, while emerging directions include explainable artificial intelligence and multimodal data fusion. Significant challenges remain, particularly limited external validation and the gap between high-performance research models and real-world clinical deployment. Mapping nearly three decades of research, this study provides researchers, clinicians, and policymakers with an evidence-based overview of publication trends, influential contributors, evolving themes, and key opportunities for developing robust, clinically applicable stroke prediction systems.