open-discussion
open-discussion > RE: Resting-state seed connectivity analysis
Oct 30, 2015 09:10 AM | ladan shahshahani
RE: Resting-state seed connectivity analysis
Dear Andrew
have you tried using independent component analysis?
As you say there is no reference function for HDR in resting state. So data-driven methods are better
Originally posted by Andrew Song:
have you tried using independent component analysis?
As you say there is no reference function for HDR in resting state. So data-driven methods are better
Originally posted by Andrew Song:
Dear experts,
I am doing Resting state functional connectivity using Seed based correlation method. As there is no clear 'reference' function (from event-stimulated hemodynamic function), I was wondering what the best way to do the analysis is.
From the entire raw data, I first regress out WM/CSF regressor and MC parameters using fsl_regfilt. Assuming that this procedure will have cleared out most noise, I then extract mean time signal from my seed ROI. After this, I use GLM on denoised data with the extracted time signal as the 'reference'.
Do you think this approach is sound?
In some of the literature I've read, they seem to extract mean time signal from the raw ROI and use this as the reference in GLM along with WM/CSF regressors. I am a bit confused, as I think the extracted time signal already contains non-neuronal noise, which cannot be accounted for by the external WM/CSF regressors.
Any insights would be helpful!
I am doing Resting state functional connectivity using Seed based correlation method. As there is no clear 'reference' function (from event-stimulated hemodynamic function), I was wondering what the best way to do the analysis is.
From the entire raw data, I first regress out WM/CSF regressor and MC parameters using fsl_regfilt. Assuming that this procedure will have cleared out most noise, I then extract mean time signal from my seed ROI. After this, I use GLM on denoised data with the extracted time signal as the 'reference'.
Do you think this approach is sound?
In some of the literature I've read, they seem to extract mean time signal from the raw ROI and use this as the reference in GLM along with WM/CSF regressors. I am a bit confused, as I think the extracted time signal already contains non-neuronal noise, which cannot be accounted for by the external WM/CSF regressors.
Any insights would be helpful!
Threaded View
| Title | Author | Date |
|---|---|---|
| Andrew Song | Aug 28, 2015 | |
| ladan shahshahani | Oct 30, 2015 | |
| Jean-Baptiste Poline | Sep 11, 2015 | |
| Alfonso Nieto-Castanon | Sep 11, 2015 | |