 |
 |
SPM'ers, I have a couple questions regarding preprocessing for a multi-session, longitudinal protocol. In this study we are collecting 3 sessions of the same task on day 1, on day 2 (time one, pre-intervention), day 2 (time two, post-intervention), and day 3. A total of 12 functional acquisitions and 6 structural (2 on each day). How would it be best to treat these in preprocessing? We have considered the following options: Perform a separate preprocessing task for each day (separating pre and post on day 2), and using the mean functional from the realignment step on in coregistration. We would then normalize and smooth for each session separately. Perform one realignment procedure for all 12 sessions and use this mean image for coregistration. Each session would then be normalized and smoothed with this experiment-wide mean. Would it be best to just pick one structural image of the 6 structural images to be used in coregistration across each day, average all 6 structural images, average the two structural images from day 1 and use throughout? Would any of the above be ideal or is there another more appropriate method? Additionally, is oblique structural and functional acquisition contraindicated in any way in SPM? Best, Drew