Covariates

Submitted by bharton on Thu, 10/04/2012 - 04:06

Hi all,

I am adding motion (3 translational and 3 rotational) as well as CSF and white matter signals as covariates in my functional connectivity analyses. Before I run functional connectivity, I have been regressing out these covariates from my data files using the Regress Out Covariates in the Utilities GUI. Then, when I run functional connectivity, I have not been including the text file of covariates in the Functional Connectivity GUI since I already regressed them out. Is this correct? Should I regress out the covariates using the Utilities GUI AND enter those same covariates into the Functional Connectivity GUI? Or do I enter in the covariates only once and if only once, where do I enter them in?

Thank you for your guidance,

Brett

Submitted by Myselves on Mon, 10/08/2012 - 20:02

"Should I regress out the covariates using the Utilities GUI AND enter those same covariates into the Functional Connectivity GUI? Or do I enter in the covariates only once and if only once, where do I enter them in?"

In REST 1.8, the latter is right, two GUI use the same method to regress out covariates. In FC GUI, we regress out covariates first and then calculate FC like DPARSF/DPARSFA.

Note:

This only applies to REST 1.8

Xin-di Wang

Submitted by bharton on Wed, 10/10/2012 - 21:17

Thank you for your response, however, I am using REST 1.6. For this version, where should I regress out the covariates - from the Utilities menu or enter the text file into the Functional Connectivity GUI?

Thank you again,
Brett

Submitted by Myselves on Thu, 10/11/2012 - 14:05

If you are using REST 1.6, you should regress out the covariates from the Utilities menu and calculate FC from FC GUI then.

Xin-di Wang