Dear Jesper,
Thanks for your reply. I have already tried the BET-on and BET-off
settings and unfortunately, the strange masking is done either way.
The following lines in the log file correspond to the masking (and this
is with BET-off):
> /usr/local/fsl-4.1.1-x86_64/fsl/bin/fslstats prefiltered_func_data_mcf -p 2 -p 98
> -0.032415 0.986444
>
> /usr/local/fsl-4.1.1-x86_64/fsl/bin/fslmaths prefiltered_func_data_mcf -thr 0.0694709 -Tmin -bin mask -odt char
>
> /usr/local/fsl-4.1.1-x86_64/fsl/bin/fslstats prefiltered_func_data_mcf -k mask -p 50
> 0.000000
>
> /usr/local/fsl-4.1.1-x86_64/fsl/bin/fslmaths mask -dilF mask
>
> /usr/local/fsl-4.1.1-x86_64/fsl/bin/fslmaths prefiltered_func_data_mcf -mas mask prefiltered_func_data_thresh
After that it goes on with error messages about division by zero. Not
sure if those lines are useful, but here goes:
> divide by zero
> while executing
> "expr $normmean / $median_intensity "
> (procedure "feat5:proc_prestats" line 406)
> invoked from within
> "feat5:proc_prestats $session "
> ("-prestats" arm line 2)
> invoked from within
> "switch -- [ lindex $argv $argindex ] {
>
> -I {
> incr argindex 1
> set session [ lindex $argv $argindex ]
> }
>
> -D {
> incr argindex 1
> set..."
> ("for" body line 2)
> invoked from within
> "for { set argindex 1 } { $argindex < $argc } { incr argindex 1 } {
> switch -- [ lindex $argv $argindex ] {
>
> -I {
> incr argindex 1
> set ses..."
> (file "/usr/local/fsl-4.1.1-x86_64/fsl/bin/feat" line 137)
(I don't know if this comes from the fslmaths command or the one after that)
So bottom line, masking is done in FEAT both with BET and without BET.
It would be nice to be able to switch it off.
bw
Alle Meije
> Dear Alle,
>
>> Is it possible in some way to switch off masking in the FEAT
>> analysis? I have some test data which looks normal, but the masking
>> at the end of preprocessing removes almost the whole image. The
>> stats are now only done on an uninteresting part of the image.
>
> this is rather surprising. The masking is based on running BET on the
> first image of the time series. Might it be that there is a problem
> with that particular volume?
>
> If not, you can just turn off "Bet brain extraction" in the Pre-stats.
>
> Good luck Jesper
|