Is there a way by which you can compare the pattern of activation between two fMRI scans where the subject performs one task (say condition 1) on the first scan and another task (condition 2) on the second task and account for the scanner baseline signal difference between these two scans or is the only solution to get subjects to perform the two conditions during the same scan.

Also is there an issue with having the paradigm subject paced (same scan duration for everyone but different number of trials per subject) if I am interested in the sustained signal change over the entire condition (i.e individual events are not of interest).