Closed hjameei closed 7 months ago
Hi @hjameei,
we recently fixed an issue related to ANTs (https://github.com/HALFpipe/HALFpipe/pull/692). Could you check if the container you used was downloaded after February 15th this year?
Best, Lea
Hi @HippocampusGirl
I was using the August 2023 version previously. After trying the new container, the issue is fixed now.
Thanks very much!
What happened? I am running HALFpipe in a HPC on a set of subjects. I am providing T1 and resting state data. The preprocessing I am using contains generating atlas-based connectivity matrices, which runs for most of the subjects. However, on a number of subjects, I encounter an error when N4BiasFieldCorrection is run. Below is an example of the error for one subject (I have attached the err.txt file as well).
Is there an error message?
What is the exact command that you used to run
Halfpipe
? I have attached Here is the script that I submitted on HPC:Which settings did you use? Here is the spec.json file that I used for the preprocessing:
I have tried comparing the header of the fMRI file that resulted in an error to the header of the fMRI that was processed successfully using
fslhd
. However, only the vertex to xyz transformation matrices were different along the two files.Here is the comparison between the two head files (the second one is the one ending in the error):
Here is the difference between header files of the mask and the fMRI file yeilding error:
This is very strange, considering there is not much difference between header files of the fMRI files that were processed successfully and the files that led to an error. Should I try using
N4BiasFieldCorrection
with-d 2
on every slice separately?I appreciate your help in advance. err.txt