AI Validation

The scope of this WP is to define a strategic position on AI validation, encompassing aspects such as fairness, cost effectiveness, and clinical effectiveness. Additionally, the group engages in discussions to address common questions related to AI validation. This WG has developed a survey titled ‘How are you performing AI validation?’ with the aim of gaining a deeper understanding of the practices and approaches used for AI validation in the field of health imaging and to determine whether there are common practices or a diverse range of strategies and approaches employed in this domain. The results of the survey aim to map and report on the current state of AI method validation in Cancer Imaging, and utilize the findings to establish a consensus on best practices for AI validation in Health Imaging Research. The ultimate objective is to widely adopt these best practices and contribute to the community by promoting standardized and effective AI validation processes in health imaging research. Furthermore, the survey findings will be summarized and presented in an academic paper that is currently being planned.