Hi Doug,

Thank you so much for clarification. Probably there is an issue with the mailing list, I only received the digest list, and now I saw your reply.

In fact, I found it by myself while looking through fast_selxavg3.m: the baseline being defined by the overall offset. and yes, I was thinking something else.. because contrast=(baseline+contrast)-baseline, so it’s just contrast/offset*100

Thanks! -SG

From: Douglas Greve <greve@nmr.mgh.harvard.edu mailto:greve@nmr.mgh.harvard.edu> Subject: Re: [Freesurfer] [FSFAST] Contrast percent map Date: 1 February 2017 at 17:38:35 GMT+1 To: freesurfer@nmr.mgh.harvard.edu mailto:freesurfer@nmr.mgh.harvard.edu

you have the right idea, but it is

cespct = 100*contrast/baseline

The baseline is the mean offset of the fMRI, the contrast is the deviation from baseline

On 2/1/17 4:49 AM, Seung-Goo KIM wrote:

...

Dear all,

I wish to know how “cespct.nii.gz” is computed in selxavg3-sess. I thought it computes something like a relative change like:

cespct = (contrast-baseline)/baseline*100,

but I cannot figure out what was taken for the baseline. I thought it could be mean beta for constant terms (0-th polynomial) for each run, but my calculation was different from cespct.

I would take average intensity for each voxel I would greatly appreciate a detailed explanation how the program computes it.

Best regards,

Seung-Goo KIM