Where do we, as radiologists, stand in our journey towards AI-enhanced radiological practice? Despite all the potential benefits AI could bring to radiology, being well-informed and unbiased is crucial, both for the models and for the radiologists. Undetected errors in the underlying data and the interaction with humans can have unintended consequences. This session will discuss where and how errors can occur and how to ensure ethical human-AI interaction.
Chairperson's introduction
Daniel Pinto Dos Santos, Mainz / Germany
How to handle data quality and data bias
Martin J. Willemink, San Diego / United States
How automation can lead astray
Mohammad Hosein Rezazade Mehrizi, Amsterdam / Netherlands
How to ethically use AI in practice
Paul Yi, Memphis / United States
Panel discussion: How can we be sure of what AI tells us?