AFNI crashing in 3dDeconvolve


I am trying to run a fixation-related (FIRE) analysis, where the event structure for the fMRI is dictated by eye movement data. I created the first script in, and then modified the afni_proc to generate full preprocessing scripts for multiple subjects.

AFNI seems to be crashing at the 3dDeconvolve step. I’m not sure where to look to fix the issue. Any guidance would be greatly appreciated!

Here is the error at the bottom:
3dDeconvolve(27095,0x7fff7d427000) malloc: *** error for object 0x7fc4a2d171e0: incorrect checksum for freed object - object was probably modified after being freed.

Here is my file:

#!/usr/bin/env tcsh

# created by version 1.2 (April 5, 2018)
# creation date: Fri Aug  3 16:59:57 2018

# set subject and group identifiers
set subj  = s$1
set gname = FIRE

# set data directories
set top_dir = /Volumes/EEG/${gname}_fMRI/Data/$1
set anat_dir  = $top_dir/anat
set epi_dir   = $top_dir/fire
set stim_dir  = $top_dir/timings

# run to create a single subject processing script -subj_id $subj                                      \
        -script proc.$subj -scr_overwrite                        \
        -blocks tshift align tlrc volreg blur mask scale regress \
        -copy_anat $anat_dir/${subj}_run-01_T1w.nii              \
        -dsets                                                   \
            $epi_dir/${subj}_run-01_bold.nii                     \
            $epi_dir/${subj}_run-02_bold.nii                     \
            $epi_dir/${subj}_run-03_bold.nii                     \
            $epi_dir/${subj}_run-04_bold.nii                     \
            $epi_dir/${subj}_run-05_bold.nii                     \
        -tcat_remove_first_trs 0                                 \
        -align_opts_aea -cost lpc -giant_move                    \
        -tlrc_base MNI_avg152T1+tlrc                             \
        -volreg_align_to MIN_OUTLIER                             \
        -volreg_align_e2a                                        \
        -volreg_tlrc_warp                                        \
        -blur_size 4.0                                           \
        -regress_stim_times $stim_dir/${subj}allfMRI.1D          \
        -regress_stim_labels                                     \
            ${subj}allfMRI.1D                                      \
        -regress_basis 'dmBLOCK'                                 \
        -regress_stim_types AM2                                  \
        -regress_censor_motion 0.3                               \
        -regress_motion_per_run                                  \
        -regress_make_ideal_sum sum_ideal.1D                     \
        -regress_est_blur_epits                                  \

Then I used this code to run the batch preprocessing.

#!/usr/bin/env tcsh
foreach sub (`cat afniproc_subs.txt`)
	echo $sub
	cd /Volumes/EEG/FIRE_fMRI/Data/${sub}
	mkdir -p ./subject_results/group.FIRE/subj.s${sub}
	cd ./subject_results/group.FIRE/subj.s${sub}
	cp /Volumes/EEG/FIRE_fMRI/Data/Scripts/afni_proc ./
	tcsh afni_proc $sub
	 tcsh -xef proc.s${sub}|& tee output.proc.s${sub}

Nothing stands out there. Are you able to send me
one of the out put text files via email (click on my name)?
Please mail me output.proc.s${sub} for some subject.

While you are at it, please mail me the corresponding
proc script, too (proc.s${sub}).

What OS and version are you running, and how much
RAM is there?


  • rick