In the context of the Artificial Intelligence for Health Imaging cluster (AI4HI), a survey has been developed to achieve a better understanding on how the AI validation is conducted in the domain of health imaging, whether there are common practices or a variety of strategies and approaches.

The AI4HI projects are a network of five EU-funded research projects currently working on developing cancer imaging data repositories and AI solutions based on medical images and related clinical and molecular data, to improve clinical practice). These projects are Primage (GA 826494), Chaimeleon (GA 952172), EuCanImage (GA 952103), Incisive(952179), and Pro-Cancer-I (GA 952159).

The questions of the survey address the overall strategy for model building and validation, as well as a number of more specific areas, technical validation, statistical analysis and evaluation metrics, and a special section on Bias & Fairness. This is meant to be collected by all those who have a role in an AI project somewhere in the process from model generation to evaluation, from both technical and clinical sides.

The survey includes 6 user profile questions and 29 model validation questions. It will not take more than 30 minutes to fill it in.

For more information or to answer the survey, please click on the following link:

https://docs.google.com/forms/d/e/1FAIpQLSds376e6gm1YB2o_z-YHX0Zxj3qpi3g1FhU8Hg20MR1PIMn4Q/viewform