help
help > RE: Seed driven rs-fcMRI
Oct 28, 2015 06:10 PM | Alfonso Nieto-Castanon - Boston University
RE: Seed driven rs-fcMRI
Hi Denise,
Regarding (1) it is perfectly fine to use already-preprocessed data. In your case I would probably suggest to still apply the ART-based outlier detection preprocessing step to your data (so you can later apply scrubbing during the Denoising step), and of course to still perform the Denoising step (which applies aCompCor, motion regression, band-pass filtering, etc.)
Regarding (2) both methods are perfectly valid. Typically subject-specific ROIs can more accurately capture the specificities of each subject anatomical and/or functional boundaries, but fixed MNI-space ROIs are considerably easier to define, create, and replicate. Personally I take fixed MNI-space ROIs as a perfectly valid starting point. As knowledge about individual areas grows then using more accurate definitions and/or parcellations of these individual areas is often advantageous.
Hope this helps
Alfonso
Originally posted by Denise Pergolizzi:
Regarding (1) it is perfectly fine to use already-preprocessed data. In your case I would probably suggest to still apply the ART-based outlier detection preprocessing step to your data (so you can later apply scrubbing during the Denoising step), and of course to still perform the Denoising step (which applies aCompCor, motion regression, band-pass filtering, etc.)
Regarding (2) both methods are perfectly valid. Typically subject-specific ROIs can more accurately capture the specificities of each subject anatomical and/or functional boundaries, but fixed MNI-space ROIs are considerably easier to define, create, and replicate. Personally I take fixed MNI-space ROIs as a perfectly valid starting point. As knowledge about individual areas grows then using more accurate definitions and/or parcellations of these individual areas is often advantageous.
Hope this helps
Alfonso
Originally posted by Denise Pergolizzi:
Hi all-
I am learning fMRI analyses, starting with a resting state data set. We have patients and controls and would like to characterize differences in connectivity between a seed in bilateral MFG using the conn toolbox 15.f. The data has already been preprocessed - slice time corrected, realigned, coregistered, segmented, normalized and smoothed.
My initial questions are:
1- Can we skip preprocessing in this version or would you recommend preprocessing within conn?
2- How do I define the MFG seed - simply select bilateral MFG and delete other ROIs from the selected atlasROI.nii or manually define MFG across participants?
Thanks so much for any feedback,
Best,
-Denise
I am learning fMRI analyses, starting with a resting state data set. We have patients and controls and would like to characterize differences in connectivity between a seed in bilateral MFG using the conn toolbox 15.f. The data has already been preprocessed - slice time corrected, realigned, coregistered, segmented, normalized and smoothed.
My initial questions are:
1- Can we skip preprocessing in this version or would you recommend preprocessing within conn?
2- How do I define the MFG seed - simply select bilateral MFG and delete other ROIs from the selected atlasROI.nii or manually define MFG across participants?
Thanks so much for any feedback,
Best,
-Denise
Threaded View
| Title | Author | Date |
|---|---|---|
| Denise Pergolizzi | Oct 28, 2015 | |
| Alfonso Nieto-Castanon | Oct 28, 2015 | |
| Denise Pergolizzi | Nov 5, 2015 | |