Most of my participants have 2 resting state fMRI files. There are two time points (pre and post), thus most participants have a total of 4 files/4 sessions. However, there are a few participants who have only 1 resting state fMRI file for a time point instead of 2. For these "1 file instead of 2" participants, they have the same amount of data as the 2 file participants (the total number of slices is the same, it's just in one file). I know that I can adjust the amount of sessions for each participant when setting up the design in CONN, but how can I tell CONN that the two post files are really post, because they will labelled as 2(pre2) and 3(post1).

For example, let's say sub 01 has 4 files. Pre has two files assigned to it (sub_01_pre1; sub_01_pre2) and post has the same amount (sub_01_post1; sub_01_post2). Sub 02 has 3 files. 1 file for pre (sub_01_pre) and 2 files for post. When I look at it in CONN, sub 02 now has 3 sessions, and the two post sessions are now session 2 and 3, when they should be 3 and 4. Do I treat session number 2 as 'missing?' or is there a better way to go about this?

Thanks for you help,
Brian

Hi Brian,

You can tell CONN any arbitrary association between each run and each condition either individually for specific subjects or jointly for groups of subjects in Setup.Conditions. In your example, you could go to Setup.Conditions and:

1) select Subject 2 and 'pre' condition, then select Session1 and mark 'condition spans entire session', last select Session2 and Session3 and mark 'condition is not present in this session' 

2) select Subject 2 and 'post' condition, then select Session2 and Session3 and mark 'condition spans entire session', last select Session1 and mark 'condition is not present in this session'

Hope this helps
Alfonso

ps. if this is too tedious or prompt to errors you may also specify this information through a configuration file; see "help conn_importcondition" for additional details about the different file formats that CONN currently supports

Originally posted by brian greeley: