If your statistical analysis will use both sessions (runs) concatenated, then I guess that you would realign all the images together, and hope for the best. You could run the stats with a design matrix that models the different sessions. If you do it this way, then the realignment may be better is it could model systematic differences between sessions. The size of such effects depends on whether the scanner is re-shimmed etc before each run. Best regards, -John ________________________________ From: SPM (Statistical Parametric Mapping) [mailto:[log in to unmask]] On Behalf Of Tracy Wang Sent: Friday, September 14, 2007 3:51 AM To: [log in to unmask] Subject: [SPM] realigning multiple sessions Hi all, If you are concatenating two sessions (which I am aware is not recommended) into a single model, do you preprocess the data as one session (i.e. realigning two sessions as one, so that the movement parameters will result in one file instead of one for each session), or would you preprocess the data as two sessions and concatenate the movement parameters to input as a regressor for the GLM? Thanks for the help! Tracy Tracy H. Wang Functional Neuroimaging of Memory Laboratory Center for the Neurobiology of Learning and Memory Department of Neurobiology and Behavior University of California at Irvine http://fnim.bio.uci.edu [log in to unmask]