standardization methods

Submitted by hoptman on

Hi Chao-Gan,
I am reading your standardization paper and would like to implement it going forward. 

However, as far as I can see, the processing stream calls for slice timing correction, motion correction, and nuisance covariate analyses (using Friston24, CSF, and WM, but without GSR).  Wouldn't this give a mean of zero for the FunImgRC files?  If so, where does the mean come from in the group level analysis?  Do I just take the mean and SD of the FunImgR images and use those?

Best,
Matt

YAN Chao-Gan

Wed, 06/18/2014 - 21:39

Hi Matt,

Yeah, for FunImgRC files, the temporal mean is zero. For the standardization methods in that paper, the global brain mean/SD of fALFF/ReHo/DC was used as standardization parameters. This is different from using the mean of FunImgRC files.

Best,

Chao-Gan