open-discussion
4 Subscribers 
open-discussion > conn setup for resting data
Showing 1-2 of 2 posts
Jun 3, 2015 07:06 PM | Ben Xu - NIH
conn setup for resting data
Hi,
I'm new with CONN (v15.c) and have some questions for resting state fMRI data analysis. Hope you can help me out.
I have two runs/sessions of resting state fMRI data per subject (i.e., pre and post intervention) and two groups (i.e., exp and control). The preprocessing (i.e., slice time, realignment, normalization, smoothing, and structural image segmentations) has been done with spm12. Here are my questions:
1. If I process the two runs (pre and post) in the same setup, will I be able to view and do analyses with these two sessions as separate conditions? I do not want the two sessions to be concatenated into one.
2. If I already have the segmented and normalized WM and CSF images, I can simply load the template T1 structural image for all subjects. The"preprocessing" step will not resegment the T1 image. Is that right?
3. In the ROI section, there are some default ROIs including BA, rsREL, LLP, MPFC, PCC, and RLP, etc. Some of these ROIs are time series (assume they are from Fox studies as indicated in the user manual). How may I use them? Also, is it better to include as many ROIs (including the BA) as I may need later at this step?
4. For setting up ROIs, is it better to set Dimensions to >1? If Dimension = 1, then I should not check the box for "Regress out covariates," right?
5. In setting "Conditions," it does allow me to set different names, onset, and duration for session 1 and session 2 when I enter both sessions. Is this right for resting data? (This question is related to Question 1 above.)
6. For 1st level resting data, should I set "Time-frequency decomposition" to "fixed band-pass filter" 0.1 - 0.008 at this step?
7. In "Options" for preprocessing, is it better to check all the boxes under "Enable analyses" and "Optional output files" (except, maybe, "create confound effects beta-maps")? Will this give me more flexibility for second level analyses?
8. After I finish the Setup, "Preprocessing," and then, Done, do I go for "Denoising?" Is it a necessary step? What is "Denoising" for, specifically? The user manual(s) doesn't seem to mention that and has a different user interface.
Thank you very much for your help in advance.
Ben
I'm new with CONN (v15.c) and have some questions for resting state fMRI data analysis. Hope you can help me out.
I have two runs/sessions of resting state fMRI data per subject (i.e., pre and post intervention) and two groups (i.e., exp and control). The preprocessing (i.e., slice time, realignment, normalization, smoothing, and structural image segmentations) has been done with spm12. Here are my questions:
1. If I process the two runs (pre and post) in the same setup, will I be able to view and do analyses with these two sessions as separate conditions? I do not want the two sessions to be concatenated into one.
2. If I already have the segmented and normalized WM and CSF images, I can simply load the template T1 structural image for all subjects. The"preprocessing" step will not resegment the T1 image. Is that right?
3. In the ROI section, there are some default ROIs including BA, rsREL, LLP, MPFC, PCC, and RLP, etc. Some of these ROIs are time series (assume they are from Fox studies as indicated in the user manual). How may I use them? Also, is it better to include as many ROIs (including the BA) as I may need later at this step?
4. For setting up ROIs, is it better to set Dimensions to >1? If Dimension = 1, then I should not check the box for "Regress out covariates," right?
5. In setting "Conditions," it does allow me to set different names, onset, and duration for session 1 and session 2 when I enter both sessions. Is this right for resting data? (This question is related to Question 1 above.)
6. For 1st level resting data, should I set "Time-frequency decomposition" to "fixed band-pass filter" 0.1 - 0.008 at this step?
7. In "Options" for preprocessing, is it better to check all the boxes under "Enable analyses" and "Optional output files" (except, maybe, "create confound effects beta-maps")? Will this give me more flexibility for second level analyses?
8. After I finish the Setup, "Preprocessing," and then, Done, do I go for "Denoising?" Is it a necessary step? What is "Denoising" for, specifically? The user manual(s) doesn't seem to mention that and has a different user interface.
Thank you very much for your help in advance.
Ben
Jun 4, 2015 04:06 PM | Pierre Bellec
RE: conn setup for resting data
Dear Ben,
Please post in the CONN forum https://www.nitrc.org/projects/conn/
Best,
Pierre
Please post in the CONN forum https://www.nitrc.org/projects/conn/
Best,
Pierre