 |
 |
Hi, No this is not a valid method - looking for areas with equal weighting of the main EV and temporal derivative is pretty meaningless. You only need to look at the main EV, except for rare cases where you might want to map HRF delays. Cheers Eugene On 25 November 2014 at 10:21, H van Steenbergen < [log in to unmask]> wrote: > Dear FSL experts, > > A colleague once told me that it is valid to create contrasts in FEAT that > include both the original main EV(s) and the respective temporal > derivative(s) by giving both weight 1 (after setting the option "Real EVs" > in FEAT; I use an event-related fMRI design with Double-Gamma HRFs along > with their TDs). I was told that this contrast might better account for > activity due to shifts in the BOLD signal. However, I do not find a lot of > papers that report using this method. Can someone advise me on whether this > indeed is a valid method to use. What is the exact difference compared to > the standard approach that only includes the original EVs in the contrasts? > > Thanks a lot for your help and advice. > > Best regards, > > Henk van Steenbergen >