Category
Trying to run Diffprep on a test subject with multi-shell DWI and I ended up with a segmentation fault after many hours of EPI correction. Script is running on research computing machine with 12 cores and 8gb of ram per core, so memory shouldnt be an issue. Any ideas? Here is my code:
DIFFPREP -i ${basedir}/DWI/${subject}_DWI_coreg_proc/${subject}_DWI_coreg.list \
-s ${basedir}/T1/${subject}_T2_ns_deob_acpc.nii \
--reg_settings registration_settings.dmc \
--will_be_drbuddied 0 \
--do_QC 0 \
--step 0 \
-c quadratic \
-e ITKBSPv4 \
--gibbs_ringing_correction on \
-d for_final \
--is_human_brain 1 \
--upsampling all \
--high_b 0 \
--center_of_mass \
--FoV 210 210 140 \
--keep_intermediate 0 \
--res 1.5 1.5 1.5
Hello Rodolphe,
You are running DIFFPREP with some legacy settings actually. Do you mind running it with the following settings and letting us know?
DIFFPREP -i ${basedir}/DWI/${subject}_DWI_coreg_proc/${subject}_DWI_coreg.list \
-s ${basedir}/T1/${subject}_T2_ns_deob_acpc.nii \
--do_QC 0 \
-e ANTSSyN \
-d for_final \
--res 1.5 1.5 1.5
This is actually near identical to your old settings except the upsampling and EPI distortion correction part. If it crashes again, Would you mind copying/pasting here the output?
Thanks.
@JiayiLiu. Also would you mind copying/pasting the output? and also running it without the "--high_b 1 --MAPMRI_bval 4000" tags, considering your data seems to non human.
Thanks.