Hi Moran,

Assuming you are testing a simple contrast that tests if a regression coefficient is different than zero (as opposed to differences between regression coefficients), do the following:

1) Use fslmaths to compute the average signal of filtered_func_data,

2) Divide the COPE image by that average

3) Multiply the result by the PPHeight at the top of the design.mat for that EV being tested, and multiply by 100.

That should give the %BOLD change. For more complicated designs need to consider the effective regressor. Jeanette Mumford has a guide on this, please see here: http://mumford.bol.ucla.edu/perchange_guide.pdf

All the best,

Anderson

On 14 March 2018 at 11:12, Moran Artzi <[log in to unmask]> wrote:

Hi all,
Given .feat folder, how can I calculate whole brain percent signal change map?
It seems that .feat contains only the Zmap, and the Featquery FSL tool enables to calculate percent signal changes only for given VOIs
Thanks
Moran