You should really only need a single T1 image (if any). Personally, I would simply realign the data as multiple sessions, coregistering all the data to the same T1 image. You could use both T1s though - averaging them in order to improve your signal-to-noise. Best regards, -John > In the current project I am running, I have four runs of a task. > Normally, I coregister all four sets of EPI data to my t1 taken at the > beginning. However, in one case, the subject had to be removed from the > scanner after 2 runs and put back in. Upon putting the subject back in > the scanner, I ran another T1, and the remaining two runs of the task. > In order to deal with analyzing all four runs of the data, I thought I > would do the following: > 1) Realign all four runs > 2) Coregister my second T1 to my first T1 (in other words, make the > first T1 the "target" and the second one the "source" and selecting no > images when it asks for "others") > 3) Coregister runs 1 and 2 to my first T1 > 4) Coregister runs 3 and 4 to my second T1 > > Is this correct? And if so, what T1 should I choose for my > normalization step? Also, when I coregister my T1's to eachother, > should I select coregister only? or coregister and reslice? > > Is there anything else to be done in this situation?