14.2.2. Beynel et al. (2024). Lessons learned from an fMRI-guided rTMS study …

Introduction

Here we present commands used in the following paper:

• Beynel L, Gura H, Rezaee Z, Ekpo EC, Deng ZD, Joseph JO, Taylor P, Luber B, Lisanby SH (2024). Lessons learned from an fMRI-guided rTMS study on performance in a numerical Stroop task. PLoS One 19(5):e0302660. doi: 10.1371/journal.pone.0302660.

  https://pubmed.ncbi.nlm.nih.gov/38709724/

Abstract: The Stroop task is a well-established tool to investigate the influence of competing visual categories on decision making. Neuroimaging as well as rTMS studies have demonstrated the involvement of parietal structures, particularly the intraparietal sulcus (IPS), in this task. Given its reliability, the numerical Stroop task was used to compare the effects of different TMS targeting approaches by Sack and colleagues (Sack AT 2009), who elegantly demonstrated the superiority of individualized fMRI targeting. We performed the present study to test whether fMRI-guided rTMS effects on numerical Stroop task performance could still be observed while using more advanced techniques that have emerged in the last decade (e.g., electrical sham, robotic coil holder system, etc.). To do so we used a traditional reaction time analysis and we performed, post-hoc, a more advanced comprehensive drift diffusion modeling approach. Fifteen participants performed the numerical Stroop task while active or sham 10 Hz rTMS was applied over the region of the right intraparietal sulcus (IPS) showing the strongest functional activation in the Incongruent > Congruent contrast. This target was determined based on individualized fMRI data collected during a separate session. Contrary to our assumption, the classical reaction time analysis did not show any superiority of active rTMS over sham, probably due to confounds such as potential cumulative rTMS effects, and the effect of practice. However, the modeling approach revealed a robust effect of rTMS on the drift rate variable, suggesting differential processing of congruent and incongruent properties in perceptual decision-making, and more generally, illustrating that more advanced computational analysis of performance can elucidate the effects of rTMS on the brain where simpler methods may not.

Study keywords: task-based FMRI, EPI, MPRAGE, human, adult, rTMS

Main programs: afni_proc.py, @SSwarper

Github page:

See these authors’ github page for descriptions and downloads of codes and supplementary text files:

https://github.com/afni/apaper_rtms_fmri_stroop

Download scripts

To download, either:

• ... click the link(s) in the following table (perhaps Rightclick -> “Save Link As…”):

  do_13_ssw.bash	run @SSwarper for nonlinear alignment to a template, and skullstripping of the subject’s T1w anatomical volume
  do_20_ap_targeting.bash	run afni_proc.py for task-based FMRI analysis (Stroop task); this uses nonlinear warps estimated with @SSwarper

• ... or copy+paste into a terminal:

  curl -O https://afni.nimh.nih.gov/pub/dist/doc/htmldoc/codex/fmri/media/2024_BeynelEtal/do_13_ssw.bash
  curl -O https://afni.nimh.nih.gov/pub/dist/doc/htmldoc/codex/fmri/media/2024_BeynelEtal/do_20_ap_targeting.bash
  

View scripts

do_13_ssw.bash

#!/bin/bash

# ===========================================================================
# Name:       do_13_ssw.bash
# Author:     LB
# Date:       7/12/22

# Syntax:     bash do_13_ssw.bash SUBJ
# Arguments:  SUBJ: subject ID

# Desc:       This script will run all chosen scans through the 
#             @SSwarper program for skullstripping (SS) and nonlinear
#             alignment (warping) of T1w anatomical to a template
# Req:        1) AFNI
# Notes:      Run this before afni_proc.py, and provide its outputs there.
# ===========================================================================

# Biowulf loading
module load afni

# ------------------------------ get inputs ---------------------------------

if [ "$#" -eq 1 ]; then
   subj=$1
else
   echo "Specify participant ID Please"
   exit 1
fi

# ------------------------------ define paths ---------------------------------

pwd_dir=`pwd`
proj_dir=${pwd_dir%/*}

anatdir=${proj_dir}/Anat/${subj}           # dir with T1w anatomical
funcdir=${proj_dir}/Func/${subj}           # dir with FMRI/EPI
stimdir=${proj_dir}/Behav/${subj}          # dir with stim timing files
analysisdir=${proj_dir}/Analysis/${subj}   # dir with SSW results; output here

MNItemplate=${pwd_dir}/MNI152_2009_template_SSW.nii.gz
echo "++ Template: ${MNItemplate}"

# --------------------------- check input dsets ------------------------------

# anat
if [ -f "${anatdir}/${subj}_T1.nii" ] ; then
cat <<EOF
++ Found T1 of ${subj}... 
EOF

else
cat <<EOF
++ ** ERROR: This participant does not have T1 scan .. EXITING ++
EOF
   exit 1
fi

# --------------------------- run SSW program ------------------------------

cd ${analysisdir}

@SSwarper                                                                    \
    -input  ${anatdir}/${subj}_T1.nii                                        \
    -base   ${MNItemplate}                                                   \
    -subid  ${subj}                                                          \
    -odir   ${analysisdir}

exit 0

do_20_ap_targeting.bash

#!/bin/bash

# ===========================================================================
# Name:       do_20_ap_targeting.bash
# Author:     LB
# Date:       7/12/22

# Syntax:     bash do_20_ap_targeting.bash SUBJ
# Arguments:  SUBJ: subject ID

# Desc:       This script will run all chosen scans through the 
#             afni_proc.py pipelining program
# Req:        1) AFNI
# Notes:       --
# ===========================================================================

# Biowulf loading
module load afni python/3.9

# ------------------------------ get inputs ---------------------------------

if [ "$#" -eq 1 ]; then
   subj=$1
else
   echo "** Please specify participant ID **"
   exit 1
fi

# ------------------------------ define paths ---------------------------------

pwd_dir=`pwd`
proj_dir=${pwd_dir%/*}

anatdir=${proj_dir}/Anat/${subj}           # dir with T1w anatomical
funcdir=${proj_dir}/Func/${subj}           # dir with FMRI/EPI
stimdir=${proj_dir}/Behav/${subj}          # dir with stim timing files
analysisdir=${proj_dir}/Analysis/${subj}   # dir with SSW results; output here

MNItemplate=${pwd_dir}/MNI152_2009_template_SSW.nii.gz
echo "++ Template: ${MNItemplate}"

cd ${analysisdir}

# --------------------------- check input dsets ------------------------------

# anat
if [ -f "${anatdir}/${subj}_T1.nii" ] ; then
cat <<EOF
++ Found T1w dset of ${subj} ++
EOF

else
cat <<EOF
** ERROR: This participant does not have T1w dset ... EXITING **
EOF
   exit 1
fi

# epi
if [ -f "${funcdir}/${subj}_Block1.nii" ] ; then
cat <<EOF
++ Found functional dset for participant ${subj}, moving to next step ++
EOF

else
cat <<EOF
** ERROR: Subject ${subj} does not have functional dset(s).
   Checked: ${funcdir}/${subj}_Block?.nii
   ... EXITING **
EOF
   exit 1
fi

# --------------------------- build pipeline ------------------------------

### Set up FMRI pipeline with AFNI's afni_proc.py
### (need to add fixation and ITI and incorrect as stim)

afni_proc.py                                                                 \
    -subj_id                  "${subj}"                                      \
    -blocks                   tshift align tlrc volreg blur mask scale       \
                              regress                                        \
    -copy_anat                "${analysisdir}/anatSS.${subj}.nii"            \
    -anat_has_skull           no                                             \
    -dsets                    "${funcdir}/${subj}"_Block?.nii                \
    -tcat_remove_first_trs    2                                              \
    -align_opts_aea           -cost lpc+ZZ                                   \
                              -giant_move                                    \
                              -check_flip                                    \
    -anat_uniform_method      unifize                                        \
    -align_unifize_epi        local                                          \
    -volreg_align_to          MIN_OUTLIER                                    \
    -volreg_align_e2a                                                        \
    -volreg_tlrc_warp                                                        \
    -tlrc_base                "${MNItemplate}"                               \
    -tlrc_NL_warp                                                            \
    -tlrc_NL_warped_dsets     "${analysisdir}/anatQQ.${subj}.nii"            \
                              "${analysisdir}/anatQQ.${subj}.aff12.1D"       \
                              "${analysisdir}/anatQQ.${subj}_WARP.nii"       \
    -volreg_compute_tsnr      yes                                            \
    -mask_epi_anat            yes                                            \
    -radial_correlate_blocks  tcat volreg                                    \
    -regress_est_blur_errts                                                  \
    -blur_size                4.0                                            \
    -regress_motion_per_run                                                  \
    -regress_censor_motion    0.3                                            \
    -regress_censor_outliers  0.05                                           \
    -regress_reml_exec                                                       \
    -regress_3dD_stop                                                        \
    -regress_stim_types       AM1                                            \
    -regress_stim_times       "${stimdir}/Congruent_Corr_${subj}.txt"        \
                              "${stimdir}/Incongruent_Corr_${subj}.txt"      \
                              "${stimdir}/Incorrect_${subj}.txt"             \
    -regress_stim_labels      Congruent Incongruent Incorrect                \
    -regress_basis            'dmBLOCK'                                      \
    -regress_opts_3dD         -jobs 16                                       \
                              -gltsym 'SYM: +Incongruent -Congruent'         \
                              -glt_label 1 Incongruent-Congruent             \
    -html_review_style        pythonic                                       \
    -execute


exit 0