I'm looking for some clarification on doing ICA based network identification when there are two conidtions in an analysis. Ideally, I'd like to run the ICA seperatley for sessions 1 and 2 as the conditions are quite different, and I don't expect the patterns of brain activity to be the same. I can't figure out a good way to do this however. Could someone please clarify if running across all sessions is what CONN is doing by default when you run a group ICA analysis? If so, how do I set it up to only run ICA across one session at a time? 

For example, let's say I'm interested in a between subjects contrast for groups A and B. I'm also interested in the difference between Default Mode (DMN) to whole brain connectivity between conditions 1 and 2 for those two groups. If I didn't care about ICA being run across both conditions, I'd run a group ICA in 1st level, identify my DMN component in the summary tab of second level, and then select that DMN component in the spatial components tab of second level. I don't know how to do something like that without running ICA across all the conditions as one big analysis. 

Any suggestions or pointers towards the correct part of the maual would be much appreciated!!