Dear CONN community,

My paper was recently desk rejected, and one of the reasons of this decision was that "the use of resting-state group analyses has many problems. Modern methods usually study individual subjects."

I am very confused by this remark. Is there a new rs-fMRI method that does not average correlations across a group of participants? I have always believed that, because rs-fMRI is noisy, one should average across as many subjects as possible to get a valid result. Studying single subjects would be going in an opposite direction.

I know about approaches like Dual Regression, that allow identifying rs networks coming from group ICA in individual subjects, but this is still based on group analysis. 

I am planning to send an email asking for clarifications, but before I do that I would like to consult with you if I'm not missing anything obvious.

Thank you!