Exploring the impact of patients' risk-benefit and knowledge perceptions on trust and intention to use AI-based medical imaging tools in radiology

The integration of artificial intelligence (AI) with medical imaging tools has enabled faster and more accurate diagnostic processes, transforming radiology into a more precise, efficient, and data-driven medical discipline. However, the successful implementation of AI-based medical imaging tools in emotionally sensitive, life-critical domains such as radiology depends heavily on public trust and acceptance. This study examines how value perceptions and trust shape behavioral intentions to adopt AI-based tools in radiology by extending Esmaeilzadeh's Value-Based Model, which is conceptually aligned with Privacy Calculus Theory. To enhance the model's explanatory power, additional variables were incorporated, including perceived knowledge as a predictor and trust as a mediating factor. A cross-sectional online survey (N = 961) was conducted, and data was analyzed through structural equation modeling. The findings indicate that perceived risk, perceived benefit, and perceived knowledge significantly influence trust perception. Importantly, trust served as a key mediating variable, partially mediating the effects of these factors on the intention to use AI-based medical imaging tools. The inclusion of trust increased the model's explanatory power from R 2 = 0.68 to R 2 = 0.74. Multigroup analysis based on gender, age, and education level revealed significant differences in certain pathways; however, the effect sizes were small. These f indings highlight the importance of developing inclusive and targeted strategies that address both technical and emotional concerns, enhance perceived benefits, foster public trust, and strengthen the intention to use AI-based tools in radiology.