deconvolved GCA

Submitted by Yoshinari Abe on Sat, 11/23/2013 - 15:32
Dear Zang and REST experts
 
In July, I asked you a question about REST GCA when I was troubled in the case of y2x analysis.
Then, I had acquired reasonable outcomes and thank you for all your help.
 
And now, there are some discussions about deconvolved GCA.
Wu propoed an approach of deconvolution in the article of "A blind deconvolution approach to recover effective connectivity brain networks from resting state fMRI data (Medical Image Analysis 17 (2013) 365–374)", and there are some other approaches.
 
How do you think about deconvolved GCA?
And which approach do you think is the best?
Please, tell me.
 
Thank you in advance.
 
Yours sincerely
 
Yoshinari Abe

Submitted by ZHANG_RESTadmin on Mon, 12/02/2013 - 23:59

to myself, I would trust classical GCA (time-lagged multivariate autocorrelation regression), which is implemented in REST, with small TR (e.g., 400ms).