Hi again Stephen and listserv,
Thanks for your help so far.
On Apr 2, 2014, at 10:06 AM, Stephen Smith <[log in to unmask]> wrote:
>> If you look at the bottom left corner of the rendered zstat map, you’ll see some activation outside of the brain, which traces the outline of the brain. The images in in fslview were simply example_func with the unthresholded zstat map as the overlay. I’m attaching another screenshot from fslview, see below. It just seems like the zstat images are in a different space than example_func.
>
> that's pretty weird - I can't see how FEAT could have generated that - what do you get if you do fslinfo on example_func and zstat1 from the same FEAT output folder? Have you generated the zstat/renderedzstat in a different space than native?
Here is the direct comparison of the zstat images produced by FEAT with
a) two-step registration (slab EPI -> whole-brain EPI -> MPRAGE): https://www.sas.upenn.edu/~mspits/feat_sample/rendered_thresh_zstat1.png
b) direct registration (slab EPI -> MPRAGE, i.e. with ‘Expanded structural image’ turned off): https://www.sas.upenn.edu/~mspits/feat_sample_no_initialhighres/rendered_thresh_zstat1.png
The dimensions of both sets of volumes (example_func and zstat1) are the same in both cases. Apart from turning off the registration ‘Expanded structural image’, nothing in the design.fsf has changed, so I am not sure why the zstat1 end up in a different space.
I’ve looked into the log files (http://www.sas.upenn.edu/~mspits/feat_sample/report_log.html and http://www.sas.upenn.edu/~mspits/feat_sample_no_initialhighres/report_log.html), and apart from the obvious differences, found some discrepancies:
a)
echo 143554 > thresh_zstat1.vol
zstat1: DLH=4.50981 VOLUME=143554 RESELS=1.02607
b)
echo 12413 > thresh_zstat1.vol
zstat1: DLH=7.58642 VOLUME=12413 RESELS=0.609955
Why the values for VOLUME would be different in the two cases is a mystery to me.
Do you have any suggestions you have as to debug this issue?
Thanks in advance,
Manuel
|