help
help > RE: Continuous subject-level covariate of interest
Jun 17, 2021 01:06 AM | Alfonso Nieto-Castanon - Boston University
RE: Continuous subject-level covariate of interest
Hi Rany,
Yes, definitely, in CONN this is called a "temporal modulation" analysis: simply enter your SC timeseries (for each subject&session) as a new first-level covariate in CONN, and then in the first-level analysis tab define a new "temporal modulation" analysis and when prompted select your SC covariate as the interaction factor. That will compute for each subject and each connection the temporal association between connectivity strength and skin conductance, and of course then you can enter these measures into second-level analyses as usual. As other analyses in CONN you may use these temporal modulation analyses in the context of seed-to-voxel or ROI-to-ROI connectivity measures.
Hope this helps
Alfonso
Originally posted by Rany Abend:
Yes, definitely, in CONN this is called a "temporal modulation" analysis: simply enter your SC timeseries (for each subject&session) as a new first-level covariate in CONN, and then in the first-level analysis tab define a new "temporal modulation" analysis and when prompted select your SC covariate as the interaction factor. That will compute for each subject and each connection the temporal association between connectivity strength and skin conductance, and of course then you can enter these measures into second-level analyses as usual. As other analyses in CONN you may use these temporal modulation analyses in the context of seed-to-voxel or ROI-to-ROI connectivity measures.
Hope this helps
Alfonso
Originally posted by Rany Abend:
Hi,
I collected skin conductance (SC) data during fMRI, and I would like to use the SC as a regressor of interest (at the individual level) in an analysis that tries to identify edges that correlate with the SC time series (at the group level). To facilitate things, I binned each subject's SC time series into TR-long bins, such that the BOLD and SC signal have the same number of time points.
Is this something that I can do in CONN? Maybe through the Conditions tab? As in, treat the SC time series as a condition in a task?
Thanks!
Rany
I collected skin conductance (SC) data during fMRI, and I would like to use the SC as a regressor of interest (at the individual level) in an analysis that tries to identify edges that correlate with the SC time series (at the group level). To facilitate things, I binned each subject's SC time series into TR-long bins, such that the BOLD and SC signal have the same number of time points.
Is this something that I can do in CONN? Maybe through the Conditions tab? As in, treat the SC time series as a condition in a task?
Thanks!
Rany
Threaded View
| Title | Author | Date |
|---|---|---|
| Rany Abend | Jun 16, 2021 | |
| Alfonso Nieto-Castanon | Jun 17, 2021 | |
| Rany Abend | Jun 17, 2021 | |
| Grace Ringlein | Jul 26, 2021 | |
| Alfonso Nieto-Castanon | Jul 28, 2021 | |
| Grace Ringlein | Aug 4, 2021 | |