Dear all,
I would be able to axclude learning or habituation effects from my fMRI
data so I tried using the time parametric linear modulation but I’m not
sure of the meaning of this.
Let’s say I have 2 tasks in a block design. In the design matrix I will
have 2 more columns:
Task1, task1*(time modulation), task2, task2**(time modulation).
I made contrasts between task1 and 2 (odd columns) without considering
the even columns and I thought that this would have been sufficient to
eliminate linear trend contribution. Is this right?
My doubt arose because the slope of the trend furnished from SPM is
positive. Do the program make a fit of the linear trend considering the
data? I don’t think so, so maybe my approach is totally wrong and I
should at least consider a negative slope using "other" in parametric
modulation.
Could anybody explain to me what exactly does the parametric modulation?
If I made wrong, Is there another way to get rid of any kind of linear
trends related to each single task?
Thanks in advance.
--
Iole Indovina.
Laboratorio di Neuroimmagini funzionali
Fondazione Santa Lucia
Via Ardeatina 306
Roma 00179 (Italy)
Tel (+39) 0651501376
Fax (+39) 0651501213
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|