Review of risk perception research by cognitive neuroscience technologies in high-risk industry

Risk perception failure is one of the primary causes of unsafe behavior in high-risk industries. While traditional studies hold value, they often fail to capture the real-time, dynamic cognitive processes involved. The emergence of cognitive neuroscience technologies has prompted a paradigm shift towards interdisciplinary research, yet a systematic synthesis of this evolving field is lacking. To address this gap, this review systematically examines the application of cognitive neuroscience technologies in risk perception research within high-risk industries, based on an analysis of 83 relevant studies. Results synthesize a multi-level analytical framework encompassing neural mechanisms, influencing factors, and computational modeling. Current research, from a cognitive neuroscience perspective, investigates: 1) the risk perception two-stage (hazard identification and risk assessment) model and the functions of the corresponding brain regions; 2) the effects of individual characteristics (e.g., expertise, traits, transient states) and external environmental factors, alongside pathways for targeted safety training interventions on risk perception; and 3) the role of machine learning models for risk perception in supporting theoretical construction and their translational potential for practical application. However, limitations persist, including an imbalance in research theme, insufficient exploration of mediating mechanisms, and a disconnect from real-world scenarios. Future research should prioritize causal analysis and the enhancement of ecological validity to better integrate risk perception theory with computational modeling, fostering the development of predictive models and intervention tools deployable in field settings.