[L3V8] Pipeline reproduction (SPM, raw)

[mselimata]

Softwares

SPM 12(v6906), Matlab 2017b

Input data

raw data

Additional contex

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 as a template if needed.
• 📥 Create a pull request as soon as you completed the previous task.
• 🧠 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).

---

NARPS team description : L3V8

General

• teamID : L3V8
• NV_collection_link : https://neurovault.org/collections/4888/
• results_comments : We analyzed the data as an confirmative study, no exploration was done.
• preregistered : No
• link_preregistration_form : NA
• regions_definition : In our analysis, the vmPFC was defined as a box ([20mm x 16mm x 16mm]) centered at the mid-centered (x-coordinate set to zero) peak coordinate reported in Tom et al (2007) ([0 39.3 -8.4]). The ventral striatum and the amygdala were defined based on the Oxford-Harvard atlas for subcortical regions, thresholded at 25% tissue probability. The bilateral nucleus accumbens was used as ventral striatum mask.
• softwares : SPM 12(v6906) on matlab 2017b
• general_comments : No

Exclusions

• n_participants : 101
• exclusions_details : The following participants were excluded: sub-016, sub-018, sub-026, sub-030, sub-032, sub-116, and sub-120.
  These participants were excluded based on severe head movement. Participants showing head displacements of >3 mm or >2 degrees were excluded. Head displacements were calculated from the realignment parameters.

Preprocessing

• used_fmriprep_data : No
• preprocessing_order : Realignment, co-registration, segmentation, normalization, smooth (6 mm).
• brain_extraction : Not performed.
• segmentation : Using the default segmentation function implemented in SPM 12.
• slice_time_correction : Not performed.
• motion_correction : Software/method: SPM/Realignment (Est & Res), no fieldmap applied.
  Reference scan: 1st scan.
  Image similarity: mutual information.
  Interpolation type: B-Spline (4th degree). Image transfomrations were combined with normalization.
  No slice-to-volume registration.
• motion :
• gradient_distortion_correction : Not performed.
• intra_subject_coreg : Software/method: SPM 12/Coregister: Estimate. Using the mean image of re-aligned images as reference, T1w image as the source image.
  Type of transofrmation: rigid-body transformation.
  Cost function: Normalised Mutual Information.
  Interpolation method: NA
• distortion_correction : No
• inter_subject_reg : Software/method: SPM/Normalization: Write.
  Volume based registration was used.
  Image type registered: T1.
  Preprocessing to images: Unified segmentation, included the bias field correction.
  Template space: MNI, SPM Tissue Probabiltiy Map (TPM.nii), resolution [1.5 1.5 1.5] mm^3.
  Additional template tranformation for reporting: not used.
  Choice of warp: nonlinear stationary velocity field (deformation field).
  Usef of regularization: yes, default value from SPM ([0 .001 0.5 0.05 0.2]).
• intensity_correction : Yes, built-in unified segmentation.
• intensity_normalization : Yes, we used the default value. i.e., session regressor.
• noise_removal : Not applied.
• volume_censoring : Not applied.
• spatial_smoothing : Software/Method: SPM 12/Smooth,
  Size and type of smoothing kernel: 3D Gaussian kernel, with FWHM [6mm 6mm 6mm].
  Space: MNI volume.
• preprocessing_comments : No

Analysis

• data_submitted_to_model : All time points form 101 participants.
• spatial_region_modeled : Whole-brain.
• independent_vars_first_level : Event-related design was used. Onset of each trial, duration = 0 (impulse response function). Each trial was associated with two parametric modulators: (1) value of gain, (2) value of loss, each modeled as a linear function.
  HRF basis, Canonical only;
  Drift Regressors: SPM built-in cosine functions.
  Movement regressors: None;
  Other nuisance regressors: None
  Orthogonalization of regressors: the parametric modulators (gain, loss) are orthogonalized with respect to the main trial regressor and each other.
  UPDATE: Given the feedback from NARPS team, we found that the default value implicit mask implemented in SPM (spm.stats.fmr_spec.mthresh) resulted a small mask. We changed the value from 0.8 to 0.3. This doesn't change our conclusions.
• RT_modeling : none
• movement_modeling : 0
• independent_vars_higher_level : We used 5 independent 2nd models: (1) equal indifference with gain; (2) equal range with gain; (3) equal indifference with loss; (4) equal range with loss; (5) group effect model with loss. All models were built using SPM's flexible factorial design, with runs as within-subject factor (equal variance), and subject as between-subject factor (equal variance). For model 5, there is an additional factor of the group as the between-subject factor (equal variance).
  For model 1-4, we tested the main effect of runs. For model 5, we tested the interaction between runs and groups.
  No other covariates were included.
• model_type : Mass Univariate.
• model_settings : Random effect model: ordinary least squares in SPM.
  Autocorrelation model: FAST in SPM.
  In our group model (model 5), we assumed equal variance between groups.
• inference_contrast_effect : In our model 1, we used the contrast [1 1 1 1] to get the average positive activation of the parametric modulator (gain value) across four runs.
  In model 2, we used the contrast [1 1 1 1] to get the average positive activation of the parametric modulator (gain value) across four runs.
  In model 3, we used the contrast [-1 -1 -1 -1] to get the average negative activation of the parametric modulator (loss value) across four runs. We used contrast [1 1 1 1] to get the average positive effect of the parametric modulator (loss value) across four runs.
  In model 4, we used the contrast [-1 -1 -1 -1] to get the average negative activation of the parametric modulator (loss value) across four runs. We used contrast [1 1 1 1] to get the average positive effect of the parametric modulator (loss value) across four runs.
  In model 5, we used the contrast[-1 1 -1 1 -1 1 -1 1] to get the group differences between equal indifference group and equal range group on the positive response to the losses.
• search_region : we used the whole-brain cluster-level FWE correction (p < 0.05), based on uncorrected cluster forming threshold (p = 0.001).
  For anatomical labelling, we used Harvard-Oxford Atlas
• statistic_type : we used the whole-brain cluster-level FWE correction (p < 0.05), based on uncorrected cluster forming threshold (p = 0.001).
• pval_computation : We used the standard parametric inference.
• multiple_testing_correction : For whole-brain corrected analysis, we used the whole-brain cluster-level FWE correction based on random field theory.
• comments_analysis : NA

Categorized for analysis

• region_definition_vmpfc : Other
• region_definition_striatum : atlas HOA
• region_definition_amygdala : atlas HOA
• analysis_SW : SPM
• analysis_SW_with_version : SPM12
• smoothing_coef : 6
• testing : parametric
• testing_thresh : p<0.001
• correction_method : GRTFWE cluster
• correction_thresh_ : p<0.05

Derived

• n_participants : 101
• excluded_participants : 016, 018, 026, 030, 032, 116, 120
• func_fwhm : 6
• con_fwhm :

Comments

• excluded_from_narps_analysis : No
• exclusion_comment : Rejected due to large amount of missing brain in center.
• reproducibility : 2
• reproducibility_comment :

[cmaumet]

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