Abstract
The risk of violence faced by healthcare workers is a widespread challenge in the global healthcare industry. This article examines the feasibility and practical value of developing a spatial prediction and early warning model for hospital violent injury incidents using Geographic Information System(GIS) technology. By integrating multi-source spatiotemporal data and employing spatiotemporal analysis, machine learning, and spatial interpolation methods, the model can identify hotspots and predict the risk of violent events at a micro scale, shifting from passive response to proactive prevention and control. This study not only offers a novel approach for the refined and scientific management of hospital safety but also provides valuable insights for the development of smart hospitals, the application of artificial intelligence, and interdisciplinary integration. Although challenges remain in data integration, privacy protection, and model generalization, advancements in technology and management optimization are expected to establish this technology as a core tool for security prevention and control in smart hospitals. This study is a preliminary exploration in this field.