[0ED6] Pipeline reproduction (SPM raw)

[cmaumet]

Softwares

SPM

Input data

raw data

Additional context

see description below

List of tasks

Please tick the boxes below once the corresponding task is finished. 👍

• 👌 A maintainer of the project approved the issue, by assigning a 🏁status: ready for dev label to it.
• 🌳 Create a branch on your fork to start the reproduction.
• 🌅 Create a file team_{team_id}.py inside the narps_open/pipelines/ directory. You can use a file inside narps_open/pipelines/templates file as a template if needed.
• 🧠 Write the code for the pipeline, using Nipype and the file architecture described in docs/pipelines.md.
• 📘 Make sure your code is documented enough.
• 🐍 Make sure your code is explicit and conforms with PEP8.
• 🔬 Create tests for your pipeline. You can use files in tests/pipelines/test_team_* as examples.
• 🔬 Make sure your code passes all the tests you created (see docs/testing.md).
• 📥 create a pull request from your code.

---

NARPS team description : 0ED6

General

• teamID : 0ED6
• NV_collection_link : https://neurovault.org/collections/4994/
• results_comments : NA
• preregistered : No
• link_preregistration_form : NA
• regions_definition : Anatomical definition of ROIs
  vmPFC: combination of Jülich cytoarchitectonic maps from the SPM Anatomy Toolbox (c 3), P >= 0.2: OFC_Fo1, OFC_Fo2, FP2, Cingul_s32 (left and right hemisphere)
  amygdala: as described above including the following maps: SF, MF, IF, LB, CM (left and right)

• ventral striatum: from the striatum atlas included in FSL (5.0.10)

• softwares : SPM12 v6685 www.fil.ion.ucl.ac.uk/spm/ ,
  Matlab R2018a (9.4.0.813654) www.mathworks.com/products/matlab & Insight & Inference for DVARS http://www.nisox.org/Software/DSE/
• general_comments : NA

Exclusions

• n_participants : 108
• exclusions_details : NA

Preprocessing

• used_fmriprep_data : No
• preprocessing_order : 1. motion correction

2. intersubject registration (normalization)
3. spatial smoothing

• brain_extraction : NA
• segmentation : NA
• slice_time_correction : NA
• motion_correction : SPM12, Realign & Unwarp using the Phase map created with SPM12 Fieldmap Toolbox v2.1 (default options): Other than defaults: Estimation: Quality 0.95, Speparation 3, Register to mean, Interpolation 7th Degree B-Spline; Unwarp Reslice: Interpolation 7th Degree B-Spline
• motion :
• gradient_distortion_correction : The single-band reference EPI was also distortion corrected using the SPM12 Fieldmap Toolbox v2.1
• intra_subject_coreg : NA
• distortion_correction : NA
• inter_subject_reg : SPM12: Within each run, the distortion corrected single-band reference EPI was co-registered to the mean EPI from Realignment & Warp using normalised mutual information. Then distortion corrected single-band reference EPI was co-registered to the gray matter probability map in the Old Segmentation toolbox in SPM using normalised mutual information and the distortion corrected EPI time-series as well as the mean EPI remained aligned. The single-band reference EPI was normalized to the SPM MNI152 template space using the classic Unified Segmentation approach in the Old Segment function in SPM, while mitigating overfitting by setting the warp frequency cutoff to 45 limiting the discrete cosine transform (DCT) bases and setting the sampling distance to 2. The resulting deformation field was applied to the distortion corrected EPI time-series, the mean EPI and the single-band reference EPI.
• intensity_correction : NA
• intensity_normalization : NA
• noise_removal : NA
• volume_censoring : To censor time-points significantly influences by noise the DVARS inference approach by Afyouni S. & Nichols T.E, (2017) was applied to each session for all subjects independently. Currupted time-points were identified using the DVARSCalc function.
• spatial_smoothing : SPM12, 5mm smoothing with a fixed kernel in MNI152 space
• preprocessing_comments : NA

Analysis

• data_submitted_to_model : 4 sessions of 449 time points, 108 subjects
• spatial_region_modeled : NA
• independent_vars_first_level : We applied an event-related design with each trial modeled as epochs of 4 sec duration with 3 parametric modulators [gain, loss, reaction time] orthogonalized by demeaning against the task and the respective preceding modulator. The canonical HRF were used for convolution including the temporal derivative. Additionally 6 motion regressors as obtained by realignment were added as regressors of no interest. Also time-points significantly influences by noise as flagged by the DVARS inference approach by Afyouni S. & Nichols T.E, (2017) were censored via an additional regressor. For each participant, all 4 sessions were modeled in one 1st Level design and contrast images for each regressor of interest were computed: [task, gain, loss, raction time].
• RT_modeling : pm
• movement_modeling : 1
• independent_vars_higher_level : We applied one flexible factorial design to examine the effects of the following 4 factors of interest for the two groups, equal Indifference and equal Range: [task, gain, loss, reaction time] resulting in 8 conditions on the 2nd level.
• model_type : Mass Univariate
• model_settings : 1st-level: with autocorrelation model in SPM [AR(1) + w] and a high-pass filter of128 s
  2nd-level: random-effects GLM with weighted least squares in SPM (restricted maximum likelihood estimation) with both between-condition and between-group variances modeled as unequal
• inference_contrast_effect : We estimated linear T-contrasts for the two parametric modulators [gain, loss] in both groups to test for the effects of the 9 hypotheses.
• search_region : NA
• statistic_type : peak-wise
• pval_computation : NA
• multiple_testing_correction : Familywise Error correction via Random Field Theory
• comments_analysis : NA

Categorized for analysis

• region_definition_vmpfc : atlas Jülich cytoarchitectonic
• region_definition_striatum : atlas Jülich cytoarchitectonic
• region_definition_amygdala : atlas Jülich cytoarchitectonic
• analysis_SW : SPM
• analysis_SW_with_version : SPM12
• smoothing_coef : 5
• testing : parametric
• testing_thresh : adaptive
• correction_method : GRTFWE voxelwise
• correction_thresh_ : p<0.05

Derived

• n_participants : 108
• excluded_participants : n/a
• func_fwhm : 5
• con_fwhm :

Comments

• excluded_from_narps_analysis : No
• exclusion_comment : N/A
• reproducibility : 2
• reproducibility_comment :

[bclenet]

Current correlation values with 106 subjects (issues are encountered while processing sub-107 and sub-119) and commit 955dcec :
[0.64, 0.59, 0.64, 0.59, 0.63, -0.30, 0.63, -0.30, 0.53]

For now, the code of #180 is not merged to the project because it contains a dependency to DVARS, whose impact on the project (and CI) should be evaluated first.

[cmaumet]

We look at this code w/ @bclenet as part of the NARPS Open pipelines hackathon Dec 12-13, 2024. Here are a few notes of things that we may want to check:

• Realign is done separatly for each run and not all run as sessions in realignunwarp (this is probably fine)
• Segmentation is done on sbref instead of anat ?

[cmaumet]

Warning: this is using flexible factorial design in SPM that is not yet implemented in nipype (cf. #168 (comment))