Dear CONN experts,

first, thank you for the excellent work in adding functionalities to CONN!

If I'm correct, by now it should be possible to run in CONN a group ICA + dual-regression as it is canonically done in FSL. I have found some information in the manual and threads in the forum that led me to the following pipeline. Could you confirm that this is a proper dual regression analysis? Thank you!

1) Run Voxel 2 voxel analysis
2) Choose option GICA3 + backpropagation
3) Identify meaningful ICAs (DMN, etc.) from second level analysis
4) Extract their time series from ICATimeseries.mat (first level folder) for each individual -> FIRST STEP OF THE DUAL REGRESSION
5) Run a Seed 2 voxel analysis using those time series as covariates -> SECOND STEP OF THE DUAL REGRESSION