Dear ICN_atlas experts,

Thank you for sharing this new tool ! I have individual subjects Independent components z-maps obtained from back projection of my group ICA and I want to describe the engagement of different networks for each subject. I am hesitating whether to use a global metric giving one coefficient for each ICA component or to collapse ICN_i spatial involvement metric across all components. I read your 2017 NeuroImage paper and it looks like the combinations of each IC and atlas map is not very repeatable while repeatbility scores are high when collapsing across ICs. What is not clear for me is :

1- How to do a collapse after running an analysis with ICN_atlas. Is it just an averaging of the metric across all ICs ?
2- What about the repeatability of global spatial involvement ?

Any advice is welcome and thank you very much for your help
Salma