Home

1. Data Preparation

• Anatomical Image: /EYEMEM###_t1.nii.gz
• Functional Images: /EYEMEM###/mri/ EYEMEM###_.nii.gz
• Preprocessing Images: /EYEMEM###/preproc2//EYEMEM###_.nii.gz
• Configuration Script: preproc2_config.sh

Important: When reviewing the IC's it was noticed that several participants had large spikes in the beginning of the timecourse, which led us to consider deleting volumes. We chose to delete the first 4 volumes for resting state images and the first 12 volumes for task images.

2. BET

Brain extraction is done using ANTs (antsBrainExtraction)

• Script: A_preproc2_BET_ANTs.sh
• Input: EYEMEM###/mri/t1/EYEMEM###_t1.nii.gz
• Output: EYEMEM###/mri/t1/EYEMEM###_t1_brain.nii.gz

Parameters

 ImageDimension: 3
 TemplateImage: /ANTS/MICCAI2012-Multi-Atlas-Challenge-Data/T_template0.nii.gz
 ProbabilityImage: T_template0_BrainCerebellumProbabilityMask.nii.gz
 RegistrationMask: T_template0_BrainCerebellumRegistrationMask.nii.gz
 KeepTemporaryFiles: 0

Bet Script: https://github.com/ANTsX/ANTs/blob/master/Scripts/antsBrainExtraction.sh

3. Prepare Fieldmap

• Script: B_preproc_prepareFM.sh
• Inputfiles: 1. <subject>_phasediff.nii.gz 2. <subject>_magnitude[1,2].nii.gz
• Outputfiles: 1. <subject>_fmap_MeanMagnitude_brain_mask.nii.gz; 2. <subject>_fmap_rads

Documentation: https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/FUGUE/Guide (Siemens Data)

3. FEAT

• Script: C_preproc2_FEAT.sh 
• Input: /EYEMEM###/mri//EYEMEM###_.nii.gz
• Output: EYEMEM###/preproc2//FEAT.feat

Documentation: https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/FEAT

Standard Parameters:

 Voxelsize: 3 
 TR: 1.0 
 ToggleMCFLIRT= YES
 Volumes: Resting State: 600; Test Conditions: 474
 Deleted Volumes: 12 (task) or 4 (rest)
 Highpass Filter: NO
 Smoothing Kernel: 7
 RegisterStructDOF: BBR
 Bandpass: OFF
 BET Input: YES
 MNI Image: MNI152_T1_3mm_brain.nii.gz

Secondary Parameters

NonLinearReg: NO 
NonLinearWarp: 10 (default)
Intensity Normalization: 0
SliceTiming Correction: 0 

FieldMap Parameters

Unwarping: 1
UnwarpDir: -y
EpiSpacing: 0.28499967 # in ms
EpiTE: 30 # in ms
SignalLossThresh: 10

4. Motion Outliers

• Script: D_preproc_motion_outliers.sh
• Inputfiles: FunctionalImage
• Outputfiles: /motionout/<subject>_motionout.txt

Documentation: https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/FSLMotionOutliers

5. Detrend and Highpass Filter

Script: E_preproc_detrend_filter_call.sh

5.1 Detrending

• Input: ./FEAT.feat/filtered_func_data.nii.gz
• Output: ./EYEMEM###_<condition>_feat_detrended.nii.gz

Parameters

 Polynomical Order:3 

5.2 Highpass Filter

• Input:  ./EYEMEM###_<condition>_feat_detrended.nii.gz
• Output: ./EYEMEM###_<condition>_feat_detrended_highpassed.nii.gz

Parameters

TR: 1.0
LowCutoff: 0.01
No High Cut Off
Filter Order: 8 

Detrending and filtering is done in Matlab using the preproc_detrend.m and preproc_filter.m functions which can be found in E_matlab_steps. Refer to the ReadMe file for more information

6. ICA

• Script: F_preproc_ICA.sh   
• Input: ./EYEMEM###__feat_detrended_highpassed.nii.gz
• Output: ./FEAT.feat/filtered_func_data.ica

Parameters

Dimest: mdl 
Bgthreshold: 3 
Report: ./FEAT.feat/filtered_func_data.ica/report.html 
TR: 1.0 
Mmthresh: 0.5 
Background: ./anat2func.nii.gz 

• Quality Check Script: ICA_melodic_fsleyes.sh (loops through subjects and condtions of the preprocessed fMRI dataset, opening FSL MELODICS IC threshold maps with an T1 underlay)

Documentation: https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/MELODIC

7. Denoising

• Script: G_denoise_regfilt.sh   
• Inputfiles: 1. Detrended/Filtered functional image 2. FEAT.feat/filtered_func_data.ica/melodic_mix 3. Textfile containing rejected components
• Output: ./_task-_run-_bold_feat_detrended_highpassed_denoised.nii.gz

Documentation: https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/MELODIC#fsl_regfilt_command-line_program

8. Normalization

• Script: H_preproc_normalize_lin.sh
• Input: Denoised functional image
• Output: ./_task-_run-_bold_feat_detrended_highpassed_denoised_MNI.nii.gz