Hi,

Not sure about the first part, sorry, you are right that including/excluding the effect of rest should produce only a very minor difference (and that all preprocessing, denoising, and analysis steps are replicable so they should produce exactly the same results when run in the same from twice on the same data), have you cheked whether anything else in the project (e.g. quality control plots) looks different from what you recall, which may give some indication as to what differs?

And regarding the second part, yes, if you are entering already preprocessed data in CONN you would need to manually create and enter the "realignment" and "scrubbing" covariates in the Setup.Covariates (1st-level) tab (entering the corresponding rp_#.txt and art_#.mat files for each subject there), in addition to entering the corresponding gray/white/CSF masks for each subject (i.e. wc1, wc2, and wc3#.nii files) in the Setup.ROIs tab. 

Hope this helps

Alfonso

Originally posted by lidizpe:

Hi everyone!

I am seeking some help and clarification since I am kind of confused right now. I had some significant results in a seed-voxel analysis, and I found a specific pattern of activation in the ICC analysis. After a meeting, I was told that in my case, the effect of rest was not necessary, so I did the denoising without this confound. I found that the significant effect was gone, which I thought was a little weird given what I understood about the effect of rest. I decided to try denoising with the effect of rest again to check, and again, there were no significant results. Then I decided to check the ICC analysis and found that the pattern of activation was completely different, which doesn’t make sense to me.

I have to emphasize that I did the whole preprocessing and denoising twice because I had a problem with a backup. I had the same results both times, so I am very confident in these results with that pipeline.

While I am running the whole pipeline again, which is going to take some time...

I checked the files in the setup because I saw in the forum that it might be caused by double preprocessing, but the files are good, not "swauswau.." for example.

The other thing I wanted to ask is regarding the realignment and the scrubbing. I decided to try a new project with the preprocessed structural and functional files, but when the preprocessing step is done, I only have the confounds (White matter, CSF, Effect of rest). I don't have the effects of realignment and scrubbing. Do I need to upload another file with that separately?

Thanks in advance for your help!