help
help > RE: Difficulty with interpreting FC results
Mar 9, 2023 10:03 PM | Alfonso Nieto-Castanon - Boston University
RE: Difficulty with interpreting FC results
Hi Peet,
The interaction stats alone do not provide enough information to properly interpret the result, they just indicate that the post-pre difference in connectivity between those ROIs is "higher" -i.e. more positive- in group2 than in group1, but that could indicate many different things depending on whether connectivity between those two regions is itself positive, or whether the two regions are anticorrelated, and whether the change between pre- and post- is an increase or a decrease in connectivity, etc.
To properly interpret the results you need to know the connectivity between those two ROIs in the four cases that are involved (Group1pre, Group1post, Group2pre, and Group2post). In the results (2nd-level) ROI-to-ROI results tab, after computing the condition-by-group interaction, change the [-1 1] contrasts (both in the between-subjects and the between-condition contrast fields) to [1 0;0 1]. After you do that, select the seed and target ROIs, and click on the button 'display effects' and it should show you a barplot with the average connectivity in those four cases (see example image attached).
Hope this helps
Alfonso
Originally posted by Peet Nowo:
The interaction stats alone do not provide enough information to properly interpret the result, they just indicate that the post-pre difference in connectivity between those ROIs is "higher" -i.e. more positive- in group2 than in group1, but that could indicate many different things depending on whether connectivity between those two regions is itself positive, or whether the two regions are anticorrelated, and whether the change between pre- and post- is an increase or a decrease in connectivity, etc.
To properly interpret the results you need to know the connectivity between those two ROIs in the four cases that are involved (Group1pre, Group1post, Group2pre, and Group2post). In the results (2nd-level) ROI-to-ROI results tab, after computing the condition-by-group interaction, change the [-1 1] contrasts (both in the between-subjects and the between-condition contrast fields) to [1 0;0 1]. After you do that, select the seed and target ROIs, and click on the button 'display effects' and it should show you a barplot with the average connectivity in those four cases (see example image attached).
Hope this helps
Alfonso
Originally posted by Peet Nowo:
Hey guys!
My brain keeps wrapping around the same problem for a while now and I was hoping some of you could help me. I am not able to grasp the meaning of the results I am getting.
Background:
rs-fMRI, functional connectivity, ROI-to-ROI
Two groups, two conditions:
Group 1 & Group 2
Pre & Post
Second-level analysis:
Between-subjects contrast: [-1 1]
Between-conditions contrast: [-1 1]
Results (exemplary):
ROI 1: R_DefaultA_PFCm
ROI 2: R_SalVentAttnA_PrC
beta = 0.19
T(31) = 4.65
p-unc = 0.000017
p-FDR = 0.002062
My interpretation is: Group 2 shows a significantly higher change in functional connectivity between ROI 1 and ROI 2 from condition "Pre" to condition "Post". This change is an increase in functional connectivity.
Is my assumption correct?
Thank you very much!
Peet
My brain keeps wrapping around the same problem for a while now and I was hoping some of you could help me. I am not able to grasp the meaning of the results I am getting.
Background:
rs-fMRI, functional connectivity, ROI-to-ROI
Two groups, two conditions:
Group 1 & Group 2
Pre & Post
Second-level analysis:
Between-subjects contrast: [-1 1]
Between-conditions contrast: [-1 1]
Results (exemplary):
ROI 1: R_DefaultA_PFCm
ROI 2: R_SalVentAttnA_PrC
beta = 0.19
T(31) = 4.65
p-unc = 0.000017
p-FDR = 0.002062
My interpretation is: Group 2 shows a significantly higher change in functional connectivity between ROI 1 and ROI 2 from condition "Pre" to condition "Post". This change is an increase in functional connectivity.
Is my assumption correct?
Thank you very much!
Peet
Threaded View
| Title | Author | Date |
|---|---|---|
| Peet Nowo | Mar 7, 2023 | |
| Peet Nowo | Mar 17, 2023 | |
| Alfonso Nieto-Castanon | Mar 9, 2023 | |
| Peet Nowo | Mar 10, 2023 | |
| Peet Nowo | Mar 14, 2023 | |