Dear Pascal,
is there a possibility to assess the percent BOLD signal change over the
>activated region in FEAT?
I've previously posted a script that will basically do what you want
(http://www.jiscmail.ac.uk/cgi-bin/wa.exe?A2=ind0307&L=fsl&P=R4134&I=-1).
By setting the mask to the active cluster you're interested in and
specifying either the conditions of interest (in pe files) or contrasts of
interest (cope files), you can extract the mean percent BOLD signal change
per subject per file. The values are all relative to a baseline of
approximately 10000 so a signal change of 110 is a 1.1% signal change.
To get a slightly more accurate baseline, you can apply the mask
(registered appropirately) to the individuals' filtered_func_data and
compute the mean over the time series. It'll be very close to 10000, though.
Joe
|