FSL: FEAT Registration Error

neurobrain · October 23, 2018, 7:14pm

Hello,

I’m running registration on FEAT, and keep getting the same error message:

Feat main script
/bin/cp /tmp/feat_b5ziBY.fsf design.fsf

mkdir .files;cp /usr/local/fsl/doc/fsl.css .files;cp -r /usr/local/fsl/doc/images .files/images

/usr/local/fsl/bin/fsl_sub -T 10 -l logs -N feat0_init /usr/local/fsl/bin/feat /path/to/feat/.feat/design.fsf -D /path/to/feat/.feat -I 1 -init
10261

/usr/local/fsl/bin/fsl_sub -T 68 -l logs -N feat2_pre -j 10261 /usr/local/fsl/bin/feat /path/to/feat/.feat/design.fsf -D /path/to/feat/.feat -I 1 -prestats
child process exited abnormally
while executing
"if { [ catch {

for { set argindex 1 } { $argindex < $argc } { incr argindex 1 } {
switch – [ lindex $argv $argindex ] {

    -I {
        incr arginde..."
(file "/usr/local/fsl/bin/feat" line 312)

FATAL ERROR ENCOUNTERED:
COMMAND:
/usr/local/fsl/bin/fsl_sub -T 68 -l logs -N feat2_pre -j 10261 /usr/local/fsl/bin/feat /path/to/feat/.feat/design.fsf -D //path/to/feat//.feat -I 1 -prestats
ERROR MESSAGE:
child process exited abnormally
END OF ERROR MESSAGE

I set up the registration through the feat gui.

Thank you!

Elvire · October 30, 2018, 2:54am

Hi,

I have met the same problem when running FEAT. Have you figured it out?

Thanks,
Elvire

CaraBohon · July 25, 2019, 12:40am

I had a similar error come up. In my case, I am using func data that was preprocessed with fmriprep, so I turned off the prestats in FEAT and then only did a minimal registration to standard per Jeanette Mumford’s workaround to deal with higher level stats required files… But I can’t get it to run without redoing all the preprocessing…

Here is a full log of the failed FEAT run in case it is helpful.

FSL

Progress Report / Log

Started at Wed Jul 24 17:15:09 PDT 2019

Feat main script

/bin/cp /tmp/feat_uECSDK.fsf design.fsf

/share/software/user/open/fsl/5.0.10/bin/feat_model design

mkdir .files;cp /share/software/user/open/fsl/5.0.10/doc/fsl.css .files;cp -r /share/software/user/open/fsl/5.0.10/doc/images .files/images

/share/software/user/open/fsl/5.0.10/bin/fsl_sub -T 10 -l logs -N feat0_init /share/software/user/open/fsl/5.0.10/bin/feat /home/groups/cbohon/ERstudy/analysis/reappraisal/1stlevel/Run1/sub-105/reappraisal1.feat/design.fsf -D /home/groups/cbohon/ERstudy/analysis/reappraisal/1stlevel/Run1/sub-105/reappraisal1.feat -I 1 -init
21997

/share/software/user/open/fsl/5.0.10/bin/fsl_sub -T 46 -l logs -N feat2_pre -j 21997 /share/software/user/open/fsl/5.0.10/bin/feat /home/groups/cbohon/ERstudy/analysis/reappraisal/1stlevel/Run1/sub-105/reappraisal1.feat/design.fsf -D /home/groups/cbohon/ERstudy/analysis/reappraisal/1stlevel/Run1/sub-105/reappraisal1.feat -I 1 -prestats
child process exited abnormally
while executing
"if { [ catch {

for { set argindex 1 } { $argindex < $argc } { incr argindex 1 } {
switch – [ lindex $argv $argindex ] {

-I {
    incr arginde..."
(file "/share/software/user/open/fsl/5.0.10/bin/feat" line 312)

FATAL ERROR ENCOUNTERED:
COMMAND:
/share/software/user/open/fsl/5.0.10/bin/fsl_sub -T 46 -l logs -N feat2_pre -j 21997 /share/software/user/open/fsl/5.0.10/bin/feat /home/groups/cbohon/ERstudy/analysis/reappraisal/1stlevel/Run1/sub-105/reappraisal1.feat/design.fsf -D /home/groups/cbohon/ERstudy/analysis/reappraisal/1stlevel/Run1/sub-105/reappraisal1.feat -I 1 -prestats
ERROR MESSAGE:
child process exited abnormally
END OF ERROR MESSAGE

Initialisation

/share/software/user/open/fsl/5.0.10/bin/fslmaths /home/groups/cbohon/ERstudy/analysis/fmriprep/fmriprep/sub-105/func/sub-105_task-reappraisal_run-01_space-MNI152NLin2009cAsym_desc-preproc_bold.den prefiltered_func_data -odt float
Total original volumes = 164

/share/software/user/open/fsl/5.0.10/bin/fslroi prefiltered_func_data example_func 82 1

Preprocessing:Stage 1

/share/software/user/open/fsl/5.0.10/bin/mainfeatreg -F 6.00 -d /home/groups/cbohon/ERstudy/analysis/reappraisal/1stlevel/Run1/sub-105/reappraisal1.feat -l /home/groups/cbohon/ERstudy/analysis/reappraisal/1stlevel/Run1/sub-105/reappraisal1.feat/logs/feat2_pre -R /home/groups/cbohon/ERstudy/analysis/reappraisal/1stlevel/Run1/sub-105/reappraisal1.feat/report_unwarp.html -r /home/groups/cbohon/ERstudy/analysis/reappraisal/1stlevel/Run1/sub-105/reappraisal1.feat/report_reg.html -i /home/groups/cbohon/ERstudy/analysis/reappraisal/1stlevel/Run1/sub-105/reappraisal1.feat/example_func.nii.gz -s /share/sw/free/fsl/5.0.6/fsl/data/standard/MNI152_T1_2mm_brain -y 6 -z 90
Option -F ( FEAT version parameter ) selected with argument “6.00”
Option -d ( output directory ) selected with argument “/home/groups/cbohon/ERstudy/analysis/reappraisal/1stlevel/Run1/sub-105/reappraisal1.feat”
Option -l ( logfile )input with argument “/home/groups/cbohon/ERstudy/analysis/reappraisal/1stlevel/Run1/sub-105/reappraisal1.feat/logs/feat2_pre”
Option -R ( html unwarping report ) selected with argument “/home/groups/cbohon/ERstudy/analysis/reappraisal/1stlevel/Run1/sub-105/reappraisal1.feat/report_unwarp.html”
Option -r ( html registration report ) selected with argument “/home/groups/cbohon/ERstudy/analysis/reappraisal/1stlevel/Run1/sub-105/reappraisal1.feat/report_reg.html”
Option -i ( main input ) input with argument “/home/groups/cbohon/ERstudy/analysis/reappraisal/1stlevel/Run1/sub-105/reappraisal1.feat/example_func.nii.gz”
Option -s ( standard image ) selected with argument “/share/sw/free/fsl/5.0.6/fsl/data/standard/MNI152_T1_2mm_brain”
Option -y ( standard dof ) selected with argument “6”
Option -z ( standard search ) selected with argument “90”

Registration

/bin/mkdir -p /home/groups/cbohon/ERstudy/analysis/reappraisal/1stlevel/Run1/sub-105/reappraisal1.feat/reg

Error: you have not selected any images to register to!

FATAL ERROR ENCOUNTERED:
COMMAND:
/share/software/user/open/fsl/5.0.10/bin/mainfeatreg -F 6.00 -d /home/groups/cbohon/ERstudy/analysis/reappraisal/1stlevel/Run1/sub-105/reappraisal1.feat -l /home/groups/cbohon/ERstudy/analysis/reappraisal/1stlevel/Run1/sub-105/reappraisal1.feat/logs/feat2_pre -R /home/groups/cbohon/ERstudy/analysis/reappraisal/1stlevel/Run1/sub-105/reappraisal1.feat/report_unwarp.html -r /home/groups/cbohon/ERstudy/analysis/reappraisal/1stlevel/Run1/sub-105/reappraisal1.feat/report_reg.html -i /home/groups/cbohon/ERstudy/analysis/reappraisal/1stlevel/Run1/sub-105/reappraisal1.feat/example_func.nii.gz -s /share/sw/free/fsl/5.0.6/fsl/data/standard/MNI152_T1_2mm_brain -y 6 -z 90
ERROR MESSAGE:
child process exited abnormally
END OF ERROR MESSAGE
child process exited abnormally
while executing
"if { [ catch {

for { set argindex 1 } { $argindex < $argc } { incr argindex 1 } {
switch – [ lindex $argv $argindex ] {

-I {
    incr arginde..."
(file "/share/software/user/open/fsl/5.0.10/bin/feat" line 312)

Error encountered while running in main feat script, halting.
child process exited abnormally

shaymaa · October 23, 2021, 2:54pm

Hi,
I have the same problem using first level analysis in FSL feat, any solutions?

thanks,
Shaymaa

Chris_Rorden · October 23, 2021, 3:11pm

I would suggest you direct FSL FEAT questions to the FSL JISCMail discussion lists. The developers of FSL monitor that list and are very responsive. Many FSL users subscribe to that list, and are often able to provide insight. I would also try to ensure that your question is sufficiently specified. Specifically, what version of FSL are you using, what errors are you getting.

brainimager · January 25, 2022, 10:48pm

Hello,

I am having a similar issue using the feat registration workaround from:

JISCMail - FSL Archives!

I performed registration for task-fMRI and resting-state fMRI outside feat. I used the workaround but am getting a weird transform for every subject after running “updatefeatreg”:

Steps performed:

Run a normal registration for one subject
Replace .mat files with $FSLDIR/etc/flirtsch/ident.mat
** reg folder now has .mat files as an identity matrix: example_func2highres.mat, example_func2standard.mat, highres2example_func.mat, highres2standard.mat, standard2example_func.mat, standard2highres.mat
Delete warp files
Overwrite the standard.nii.gz image with mean_func.nii.gz (cp mean_func.nii.gz reg/standard.nii.gz)
run “updatefeatreg” in the feat directory
‘The resulting “reg” subdirectory can then be copied into your first level analysis without being changed/resampled’

I am not sure if this is an error with updatefeatreg or if I made a mistake in one of the steps. I
am not sure if any other files in the reg folder need to be deleted.

The higher-level analysis runs, though no results appear compared to a higher-level analysis previously run and before changing the registrations outside of feat.

Not sure where to go from here or where the issue lies. Suggestions welcome!
Many thanks!

dlevitas · January 26, 2022, 2:26am

I haven’t used updatefeatreg, but my understanding of the Mumford video is that the workaround is meant to be applied upon finished level 1 FEAT, but before level 2 FEAT.

The reason is that level 1 requires a flirt registration, but it doesn’t actually get applied until level 2. With the linked workaround, we provide FEAT with an identity matrix, such that when the registration is applied at level 2, it doesn’t do anything (because of the identity matrix). So you’re left with the registration that you performed outside of FEAT>

What this means in the context of this question is that you don’t necessarily need to run updatefeatreg; simply complete level 1, perform the workaround steps in the link, and then proceed to level 2. You can examine the level 2 html reports to assess how the registration went, in addition to step #7 in the link.

brainimager · January 27, 2022, 6:07pm

Thank you for your help! Yes, the identity matrix was applied after Feat 1 but before Feat 2.
It appears the reg folder I made from running a “normal” registration on one person is now not in the same space. I wonder if that is were the weird transformation comes from. Could that be the issue?
For instance, the “mean_func” aligns with MNI152_T1_2mm, though the “example_func” that is within the “reg” folder copied in, is not aligned?

I’m not sure if that has anything to do with it.
I will try copying over the reg folders again without running updatefeatreg and see if that solves this.
Please let me know if you have any other suggestions!

Thanks

dlevitas · January 27, 2022, 8:40pm

Since you’re performing the FEAT workaround steps, I’m assuming that you’ve pre-processed your data with fMRIPrep. Which --output spaces option(s) did you specify in your fMRIPrep command? I ask because the MNI152NLin6Asym template option is the same standard template (I think) to FSL’s MNI152 template, so if you used a different one during pre-processing, that might explain the discrepancy you’ve been seeing.

Regardless, in my experience using this FEAT workaround, I haven’t had registration issues, so long as the workaround steps are correctly followed and you move right to FEAT level 2. Hopefully rerunning without updatefeatreg will solve your issue!

jackpmanning · April 2, 2024, 3:49pm

Hi,

I have fmriprep’d data that I am using in FSL and I had a couple questions about the workaround steps that I am confused about.

Step 1 says to run normal registration for a single subject, but then how does the reg directory get copied to all runs for all subjects? Do I copy the same reg folder into every feat directory I have? I understand that we are using an identity matrix here, but this is my first time using feat, so I am not sure what other files go into a reg folder.

In step 4, what are the paths for the mean_func.nii.gz and the standard.nii.gz (i.e. where are they normally stored or output to)? These seem obvious, but I do not want to assume. But, my assumption is standard.nii.gz is in the reg directory and mean_func.nii.gz is in the feat directory

For step 5, should I be running updatefeatreg or no?

For step 6, this goes along with my initial question and where my confusion stems from: What steps do I do for only one subject, and what steps do I do for all subjects? And, which new files that I create are subjects specific as opposed to subject agnostic? (I am assuming the identity matrix will be subject agnostic, so long as the dimensions are the same).