Submitted by Mamtis13 on Thu, 11/07/2013 - 01:15
Dear experts,
I have conducted a ROI-wise (i.e. 2 ROIs) functional connectivity analysis in REST. To start with, I only ran the FC analysis for each subject, but now I would also like to include a personality score for every person. How can I include the parametric covariate, which is different for every person in the ROI-wise FC analysis in REST?
I know that there is an option to include a text file which contains a covariable. If this is the right way to include the survey score, can you please specify the content of the text file (i.e. all subjects' scores in one column, separate text files per subject, headers?).
Thank you very much!
Kind regards, Mamtis
Submitted by ZHANG_RESTadmin on Fri, 11/08/2013 - 01:03 Permalink
Re: ROI-wise FC analysis in REST: individual covariate scores
you are talking about running ROI-wise FC for only 1 subject, which means this is an individual study. In this case, no need of adding multi-subjects' personality scores as covariate.
In group analysis, e.g., you had done 20 sujbects' FC maps, you can perform correlation analysis to study the relationship between FC and personality. In this case, you can add personality scores as covariate.
Submitted by Mamtis13 on Fri, 11/08/2013 - 04:53 Permalink
Re: ROI-wise FC analysis in REST: individual covariate scores
Thank you very much for the swift response, much appreciated!
I have data from 35 subjects and would like to study the relationship between individuals' FC between 2 seed regions and their personality scores. Can I enter the personality score as a covariate when I specify the ROI-wise FC analysis in REST?
If so, can you tell me how I need to set up the text file through which the covariate gets entered into the ROI-wise FC analysis? For example, do I create 1 text file that contains one column with every subject's personality score? Or will I need to set up the analysis individually for every subject and specify a text file that contains only that particular individual's personality score?
Many thanks again for your help!
Regards, Mamtis
Submitted by ZangYF on Fri, 11/08/2013 - 09:38 Permalink
Re: ROI-wise FC analysis in REST: individual covariate scores
I do not know how to use REST. I'd like to say that a FC between two timecourses will generate a r-map, then transformed into z-map. Then, REST support a group-level correlation between z and personality. In such case, it should not be a covariate.
Submitted by ZHANG_RESTadmin on Tue, 11/12/2013 - 22:36 Permalink
Re: ROI-wise FC analysis in REST: individual covariate scores
I agree.
if you use REST, you need:
1. perform roi-wise FC for all subjects using your two ROI masks.
2. find out the z value of FC for each subject from the REST outputs.
3. correlation between z values and personality scores across subjects using either Matlab or SPSS or Excel.
Submitted by Mamtis13 on Wed, 11/13/2013 - 01:57 Permalink
Re: ROI-wise FC analysis in REST: individual covariate scores
Thank you very much for the massively helpful comments and advice, hugely appreciated!
Regards, Mamtis