Despite significant advancements in artificial intelligence (AI) and its unquestionable potential to improve healthcare, the widespread adoption of AI-driven tools is often hindered by perceptions of mistrust. To address these concerns, the field of trustworthy AI has emerged, focusing on understanding the factors that lead to mistrust among various stakeholders involved in the design, development, and implementation of AI tools in healthcare. Therefore, a key strategy to enhance trustworthiness is the engagement of all relevant actors from the early stages of clinical conceptualisation. This chapter details the process and outcomes of a multi-stakeholder engagement strategy grounded in a social innovation approach and framed around the FUTURE-AI guidelines for trustworthy AI, implemented within the RadioVal project. RadioVal is an international, multi-disciplinary project that conducts the first international clinical validation study of radiomics-based prediction for neoadjuvant chemotherapy treatment response in breast cancer. We outline the stakeholder mapping process and describe three rounds of social innovation sessions conducted throughout the project’s progression. Additionally, we describe a survey and a workshop focused on specific principles of the FUTURE-AI guidelines. The methods and results of this work provide a transferable framework for the design and implementation of trustworthy AI-driven tools in healthcare.