Hi,
I have just compared the differences in FEAT processing in FSL3.3 and FSL4.1
on the same dataset with the same design.fsf settings. Referring back to
previous posts:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind0711&L=FSL&P=R44759
I can see why I get pretty much the same result within the brain tissue, but
do get marked differences around the edges of the brain tissue (i.e. there
is more brain tissue remaining after BET in the FSL3.3 analysis than in FSL
4.1). My understanding is that this is due to improvements in the
application of spatial filtering to 4D datasets (including a move away from
betfunc to running bet on the mean_func data instead).
My questions are thus:
(1) Is my understanding of these differences correct (so that I can explain
them to our users)?
(2) Is there a simple recipe to reproduce the default settings in FSL3.3
with 4.1? – if not .. could we possibly get hold of the source for 3.3 so
that we can have a dual installation (there are some users re-visiting older
analyses and would thus like to re-produce past results.
Thanks for your help ...
Andre’ and the team @ YNIC
|