 |
 |
On Tue, Mar 27, 2012 at 8:50 AM, Modestino, Edward J *HS < [log in to unmask]> wrote: > Dear SPM experts, > I have a simple question. I am doing a parametric fMRI design. I have > three conditions. (1) On (2) Residual, and (3) Off. I expect the most > activation to be in condition 1, lesser activation in condition 2 and close > to baseline for condition 3. A couple of comments: (a) parametric modulation refers to modulation within a single condition with a parametric modulator as a regressor; and (b) it is unclear what baseline is in your study since you have an "Off" condition. If you don't have any timepoints without one of the three conditions, then there is no baseline. How do I model this in the contrasts? In fMRI everything is relative. You are asking several questions within one contrast, which is hard to do. There are several options: (a) Use a linear contrast [1 0 -1] where condition1 > condition3; although this doesn't tell you where condition 2 falls or if condition 3 is close to baseline; (b) Use two contrasts and look at where they overlap [1 -1 0] and [0 1 -1]; although if the differences aren't significant between c1/c2 or c2/c3 you will have a problem and it also doesn't tell you if c3 is close to baseline. (c) Use an F-contrast to determine where the conditions are different, then with in these regions look at where (b) overlaps. This is probably the best option. Again you don't know if c3 is close to baseline. (d) Use three contrasts to identify only voxels that are greater than 0 in all conditions; then use one of the above options to test the relationships in those voxels. c3 could be significantly different than 0 though. ** There is no way to test if c3 is close to baseline because there is no way to test if something is significantly the same as a specific value, you are only able to test if something is significantly different. > Oh, and I am also using the realignment parameter file with this design. > Thanks, > Ed Modestino > > Edward Justin Modestino, Ph.D. > Postdoctoral Research Associate > Ray Westphal Neuroimaging Laboratory > Division of Perceptual Studies > Department of Psychiatry and Neurobehavioral Sciences > University of Virginia > Email: [log in to unmask] > > >