Hello everyone,

So I'm fairly new to the afni world and still getting accustomed with all the technical terms and all that. However, I started experimenting with something on one of the datasets with which I am currently working and I ended up interested in running a Functional Connectivity analysis for some preliminary hypothesis testing.

The design is event related and in one particular case I wanted to regress out some events (fairly few, compared to the remaining ones). Now, after some thought on how this could be achieved, I ended up running 3dDeconvolve with the ideal 1D (previously computed in yet another 3dDec) stimuli of interest as stim_base, and then using the X.nocensor matrix with 3dTproject on the catenated all_runs file.

After all this, I actually ended up documenting a bit on how to run the FC per-se and then I encountered 3dSynthesize. After reading the manual for both 3dSynth and 3dTproject I found that there isn't much difference between them. Or so I think. So my question to you is: could there be any considerable differences between the two approaches regarding this particular case in which task is regressed? How about when no task is regressed and only motion parameteres are used to clean the data?

Thanx in advance!
Ș



Edited 1 time(s). Last edit at 04/28/2015 03:46AM by Şerban Condrea.