Structural & Functional Processing & Derivatives🔗
For structural and functional MRI processing, file selection is based on raw data quality control metrics, including:
- Overall passing QC score (
QC= 1) - Motion score below a defined threshold (
QU_Motion≤ 2 for the current release) - If multiple scans are present for a given modality (T1w/T2w), the scan with the highest QC metrics is used
All processing streams utilized both the T1w and T2w if they were present. Processing was still executed with only a single modality present as well, with certain requirements depending on the surface reconstruction method utilized within Infant fMRIPrep (see details):
- M-CRIB-S (T2w-based): requires T2w
- Infant FreeSurfer (T1w-based): requires both T1w and T2w
Multistep Processing Workflow🔗
HBCD structural and functional MRI data are processed through a sequence of BIDS App pipelines. At a high level, BIBSNet generates brain tissue segmentations and masks for T1w/T2w images. These are fed into Infant-fMRIPrep to generate confound files and motion-corrected data (in MNI space, registered to age-specific volumetric atlases) as well as fs_LR32k surface space. Outputs are then fed into XCP-D to run nuisance regression/denoising, parcellate the fMRI data, and compute summary measures.
BIBSNet🔗
BIBSNet is a deep learning model optimized for infant MRI brain tissue segmentation (Hendrickson et al. 2024). The BIBSNet pipeline generates native-space brain segmentations and brain masks (as well as volumes.tsv files with ROI volume statistics), which are fed into Infant fMRIPrep for use in anatomical preprocessing and surface reconstruction.
hbcd/
└── derivatives/
└── bibsnet/
└── sub-[ID]/
└── ses-[V0X]/
└── anat/
├── sub-[ID]_ses-[V0X]_space-{T1w|T2w}_desc-aseg_dseg.nii.gz (+JSON)
├── sub-[ID]_ses-[V0X]_space-{T1w|T2w}_desc-aseg_volumes.tsv (+JSON)
└── sub-[ID]_ses-[V0X]_space-{T1w|T2w}_desc-aseg_brain-mask.nii.gz (+JSON)
How To Read File Trees →
Infant fMRIPrep🔗
Infant-fMRIPrep (also known as NiBabies) performs minimal structural and functional MRI processing. It is an adapted version of fMRIPrep optimized for infant data processing, using age-appropriate templates and surface reconstruction methods optimized for early development (Goncalves et al., 2025). Pipeline outputs include visual quality assessment reports, preprocessed derivatives, and confounds used for denoising in subsequent processing steps.
Anatomical Preprocessing
T1w and T2w images are denoised, bias-corrected, and normalized to the MNI Infant template (0–4.5 yr), then to MNI152 for compatibility with adult datasets. Surface reconstruction is performed via one of the following methods:
| M-CRIB-S | T2w-based method for neonates (Adamson et al., 2020). Infant fMRIPrep runs a modified MCRIBReconAll workflow that uses the BIBSNet-derived brain segmentation. Optimal age range per Goncalves et al., 2025: ≤ 5 months |
| Infant FreeSurfer | T1w-based method for infants 0-2 years old (Zöllei et al., 2020). Infant fMRIPrep executes infant_recon_all with its default configuration. Optimal age range per Goncalves et al., 2025: ≥ 3 months |
Functional Processing
- Motion and distortion correction using fieldmap-based estimation.
- Alignment of functional to anatomical space via boundary-based registration.
- Confound estimation: framewise displacement (FD) and DVARS for motion, CompCor physiological noise regressors, global signals (mean CSF, white matter, and whole brain), and derived regressors (e.g. motion outlier flags for frames exceeding 0.5 mm FD or 1.5 standardized DVARS thresholds)
- Resampling of BOLD data to subject and fsLR-space surfaces, with grayordinates (91k) for surface-based analyses.
Overview
- JSON files excluded for brevity from file trees below
- See How To Read File Trees for additional guidance
- T1w-related files will only be present in the derivatives if a T1w was acquired
hbcd/
└── derivatives/
└── nibabies-{HASH}/
└── sub-[ID]/
├── figures/
├── ses-[V0X]/
│ ├── anat/
│ ├── fmap/
│ ├── func/
│ └── log/
│
└── sub-[ID]_ses-[V0X]_hash-{HASH}.html
# Label Values Legend
HASH: 0f306a2f , 2afa9081
Anatomical Folder Details
...
└── ses-[V0X]/
└── anat/
# Primary volumetric outputs & segmentations
├── *_desc-preproc_{T1w|T2w}.nii.gz
├── *_space-MNI152NLin6Asym_res-2_desc-preproc_T2w.nii.gz
├── *_space-MNI152NLin6Asym_res-2_desc-brain_mask.nii.gz
├── *_space-T2w_desc-ribbon_mask.nii.gz
├── *_space-{STD_SPACE}_dseg.nii.gz
├── *_space-T2w_desc-{aparcaseg|aseg}_dseg.nii.gz
├── *_space-{STD_SPACE}_label-{CSF|GM|WM}_probseg.nii.gz
# Transforms
├── *_from-{SPACE}_to-T2w_mode-image_xfm.h5
├── *_from-T2w_to-{SPACE}_mode-image_xfm.h5
# Surface & CIFTI outputs
├── *_hemi-{L|R}_desc-cortex_mask.label.gii
├── *_space-fsLR_den-91k_{METRIC}.dscalar.nii
├── *_hemi-{L|R}_{METRIC}.shape.gii
├── *_hemi-{L|R}_{inflated|sphere}.surf.gii
├── *_hemi-{L|R}_{SURF}.surf.gii
├── *_hemi-{L|R}_space-dhcpAsym_den-32k_{SURF}.surf.gii
└── *_hemi-{L|R}_space-{dhcpAsym|fsaverage}_reg_sphere.surf.gii
...
# Label Values Legend
File Prefixes (anat/, fmap/ files): sub-[ID]_ses-[V0X]_hash-{HASH}_run-[X]
METRIC: curv , sulc , thickness
SPACE: fsnative , MNI152NLin6Asym , MNIInfant+1 , T1w
STD_SPACE: MNI152NLin6Asym_res-2 , T2w
SURF: midthickness , pial , white
Fieldmap & Functional Folder Details
...
└── ses-[V0X]/
├── fmap/
│ └── *_fmapid-auto[X]_desc-{coeff|epi|preproc}_fieldmap.nii.gz
│
└── func/
# BOLD & masks
├── *_desc-{preproc_bold|brain_mask}.nii.gz
├── *_space-{STD_SPACE}_boldref.nii.gz
├── *_space-{STD_SPACE}_desc-{preproc_bold|brain_mask}.nii.gz
# Motion-corrected or coregistered outputs
├── *_desc-{hmc|coreg}_boldref.nii.gz
├── *_from-orig_to-boldref_mode-image_desc-hmc_xfm.txt
├── *_from-boldref_to-T2w_mode-image_desc-coreg_xfm.txt
├── *_from-boldref_to-auto*_mode-image_xfm.txt
# Surface & CIFTI outputs
├── *_space-fsLR_den-91k_bold.dtseries.nii
├── *_hemi-{L|R}_space-fsnative_bold.func.gii
# Confounds
└── *_desc-confounds_timeseries.tsv
# Label Values Legend
File Prefixes (func/): sub-[ID]_ses-[V0X]_hash-{HASH}_task-rest_dir-PA_run-[X]
STD_SPACE: MNI152NLin6Asym_res-2 , T2w
M-CRIB-S & FreeSurfer🔗
Infant fMRIPrep and XCP-D derivative folder/filenames include unique hash IDs to indicate distinct processing parameters used for a given pipeline. In the case of HBCD data, the hash IDs correspond to which surface reconstruction method was used for processing within Infant fMRIPrep.
M-CRIB-S & FreeSurfer are alternative surface reconstruction methods supported by Infant fMRIPrep, optimized for different age ranges (see details above).
| Method | Hash ID | Description | Visits (Age Range in Months) |
|---|---|---|---|
| M-CRIB-S | 0f306a2f | T2w-based method for neonates | V02 (0-1 m) |
| Infant FreeSurfer | 2afa9081 | T1w-based method for infants 0-2 years old | V02 (0-1 m), V03 (3-9 m), V04 (9-15 m) |
Downstream XCP-D derivatives include a second hash ID (0ef9c88a) indicating the XCP-D processing configuration. This value is identical for all HBCD data because the XCP-D parameters were fixed. Below we summarize the processing workflows and resulting derivative folder names.
The M-CRIB-S and FreeSurfer derivative folders are generated from the intermediate FreeSurfer-like folders produced by Infant fMRIPrep during surface reconstruction. When M-CRIB-S is used, Infant fMRIPrep still creates a FreeSurfer-structured folder containing the M-CRIB-S results mapped to the standard recon-all layout; these appear in the release under freesurfer-0f306a2f/.
hbcd/
└── derivatives/
└── freesurfer-{HASH}/
└── sub-[ID]_ses-[V0X]/
├── label/
│ ├── {lh|rh}.{ATLAS}.annot
│ ├── {lh|rh}.{ATLAS}.auto.nomask.annot
│ └── {lh|rh}.cortex.label
│
├── mri/
│ ├── T2.mgz
│ ├── {ATLAS}+aseg.mgz
│ ├── aseg*.mgz
│ ├── {brain|brainmask}.mgz
│ ├── {lh|rh}.ribbon.mgz
│ ├── norm.mgz
│ ├── orig.mgz
│ └── ribbon.mgz
│
├── stats/
│ ├── aseg.stats
│ ├── brainvol.stats
│ ├── {lh|rh}.{ATLAS}.stats
│ └── {lh|rh}.curv.stats
│
├── surf/
│ ├── {lh|rh}.{white,pial,midthickness}
│ ├── {lh|rh}.{inflated,sphere}*
│ ├── {lh|rh}.smoothwm*
│ ├── {lh|rh}.{area,area.mid,area.pial}
│ └── {lh|rh}.{curv,sulc,thickness,volume}
│
└── scripts/*
# Label Legend
HASH: 0f306a2f | 2afa9081
HEM: lh | rh
ATLAS: aparc | aparc+DKTatlas
hbcd/
└── derivatives/
└── mcribs-0f306a2f/
└── sub-[ID]_ses-V02/
├── RawT2/sub-[ID]_ses-V02.nii.gz
├── RawT2RadiologicalIsotropic/sub-[ID]_ses-V02.nii.gz_symlink_s3_object
├── SurfReconDeformable/
│ └── sub-[ID]_ses-V02/
│ ├── meshes/
│ │ ├── {internal|pial|white|pial+internal|white+internal}.vtp
│ │ ├── pial-{lh|rh}.vtp
│ │ ├── pial-{lh|rh}-reordered.vtp
│ │ └── white-{lh|rh}.{CortexMask.curv|Normals.surf|RegionId.curv|vtp}
│ ├── recon/
│ │ ├── cortical-hull-dmap.nii.gz
│ │ └── regions.nii.gz
│ └── temp/
│ ├── brain-mask.nii.gz
│ ├── ventricles-dmap.nii.gz
│ ├── t2w-image.nii.gz_symlink_s3_object
│ ├── cerebrum-{lh|rh}-dmap.nii.gz
│ ├── cerebrum-{lh|rh}-hull-[X].vtp
│ ├── cerebrum-{lh|rh}-iso.vtp
│ ├── {STRUCT}-mask-[X].nii.gz
│ ├── {cerebrum-lh|cerebrum-rh|pial|white}-[X].vtp
│ ├── {cerebrum-lh|cerebrum-rh|pial|white}-[X]-output_[X].vtp
│ │
│ └── {pial|white}-foreground.nii.gz
├── TissueSeg/
│ ├── sub-[ID]_ses-V02_all_labels.nii.gz
│ ├── sub-[ID]_ses-V02_all_labels_manedit.nii.gz_symlink_s3_object
│ ├── sub-[ID]_ses-V02_brain_mask.nii.gz
│ └── sub-[ID]_ses-V02_t2w_restore.nii.gz_symlink_s3_object
├── TissueSegDrawEM/sub-[ID]_ses-V02/N4/sub-[ID]_ses-V02.nii.gz_symlink_s3_object
├── freesurfer/ # M-CRIB-S–specific outputs
│ └── sub-[ID]_ses-V02/
│ └── mri/
│ └── {brain|orig}.mgz_symlink_s3_object
├── logs/sub-[ID]_ses-V02.log
└── command.txt
# Label Values Legend
STRUCT: brain | cerebrum-{lh/rh} | corpus-callosum | cortex | {deep-gray|gray|white}-matter | ventricles
When downloaded, the symlink files present within the M-CRIB-S derivatives (mcribs-0f306a2f/), appended with *_symlink_s3_object, appear as text files that contain the S3 object path instead of the actual file content. If needed, you may restore these files as symlinks via the following terminal command, which restores all symlink files within your locally downloaded directory and renames them without *_symlink_s3_object to match the original sourcedata filenames:
find . -type f -name "*_symlink_s3_object" -print | while read path ; do
symval=$(cat "$path")
symdir=$(dirname "$path")
symbase=$(basename "$path" _symlink_s3_object)
ln -s "$symval" "$symdir/$symbase" && rm -f "$path" || break
XCP-D🔗
XCP-D performs functional MRI post-processing and noise regression from Infant-fMRIPrep derivatives, producing cleaned and parcellated data (see parcellation atlases) ready for analysis.
Anatomical Processing
Native-space T2w images are transformed into standard MNI152NLin6Asym space (1 mm³ resolution).
Morphometric surfaces (fsLR-space) from Infant fMRIPrep are copied to the XCP-D derivatives. HCP-style midthickness, inflated, and very-inflated surfaces are generated from the white-matter and pial surface meshes and mapped to fsLR space.
Functional Processing
For each BOLD run, XCP-D performs a series of cleanup and quality-control steps:
- First 4 volumes (dummy scans) are removed.
- Motion correction: Framewise displacement (FD) is calculated per Power et al. (2014); volumes with FD > 0.3 mm flagged as high-motion outliers.
- Nuisance regression: 36 confound regressors (motion, tissue, and global signals plus derivatives) regressed out following the 36P strategy.
- Despiking and filtering: Data despiked, temporally filtered (0.01–0.08 Hz), and smoothed (6 mm FWHM).
- Censoring: High-motion volumes are interpolated and later censored to minimize motion artifacts.
- Amplitude of Low-Frequency Fluctuations (ALFF) and Regional Homogeneity (ReHo) metrics computed from cleaned data.
- Parcellated time series are extracted for each atlas and pairwise functional connectivity is calculated as the Pearson correlation between regional time series.
- Postprocessed derivatives are concatenated across runs.
# JSON files excluded for brevity How To Read File Trees → hbcd/ └── derivatives/ └── xcp_d-{HASH}/ └── sub-[ID]/ └── ses-[V0X]/ ├── anat/ │ │ # File Prefix: sub-[ID]_ses-[V0X]_hash-{HASH}_run-[X] │ ├── *_space-MNI152NLin6Asym_desc-preproc_T2w.nii.gz # Preprocessed T2w in MNI standard space │ ├── *_hemi-{L|R}_space-fsLR_den-32k_{SURF}.surf.gii # fsLR 32k cortical surfaces (L/R) │ ├── *_space-fsLR_den-91k_{METRIC}.dscalar.nii # Dense scalar maps (fsLR 91k grayordinates) │ └── *_space-fsLR_seg-{PARC}_stat-mean_desc-{METRIC}_morph.tsv # Atlas-based summary statistics │ ├── func/ │ │ # Run-specific files ('run-[X]') are omitted for brevity if concatenated files are present │ │ # File Prefix: sub-[ID]_ses-[V0X]_hash-{HASH}_task-rest │ │ │ # Primary denoised BOLD outputs in fsLR grayordinate space + confound files │ ├── *_space-fsLR_den-91k_desc-{denoised|denoisedSmoothed}_bold.dtseries.nii │ ├── *_{motion|outliers}.tsv │ ├── *_dir-PA_run-[X]_design.tsv │ │ # Dense scalar maps │ ├── *_dir-PA_run-[X]_space-fsLR_den-91k_stat-alff_desc-smooth_boldmap.dscalar.nii │ ├── *_dir-PA_run-[X]_space-fsLR_den-91k_stat-{alff|reho}_boldmap.dscalar.nii │ │ # Parcellated outputs │ ├── *_space-fsLR_seg-{PARC}_den-91k_stat-mean_timeseries.ptseries.nii │ ├── *_space-fsLR_seg-{PARC}_stat-mean_timeseries.tsv │ ├── *_space-fsLR_seg-{PARC}_stat-pearsoncorrelation_relmat.tsv │ ├── *_dir-PA_run-[X]_space-fsLR_seg-{PARC}_den-91k_stat-pearsoncorrelation_boldmap.pconn.nii │ ├── *_dir-PA_run-[X]_space-fsLR_seg-{PARC}_den-91k_stat-coverage_boldmap.pscalar.nii │ ├── *_dir-PA_run-[X]_space-fsLR_seg-{PARC}_stat-coverage_bold.tsv │ ├── *_dir-PA_run-[X]_space-fsLR_seg-{PARC}_stat-{alff|reho}_bold.tsv │ │ # Quality control metrics │ ├── *_dir-PA_run-[X]_space-fsLR_den-91k_desc-linc_qc.tsv │ └── *_desc-abcc_qc.hdf5 │ ├── figures/* ├── sub-[ID]_ses-[V0X]_hash-{HASH}_executive_summary.html └── sub-[ID].html # ── Label Legend ───────────────────────────────────────────── HASH : 0f306a2f+0ef9c88a , 2afa9081+0ef9c88a METRIC : curv , sulc , thickness PARC : 4S-{156|256|...|1056}Parcels , Glasser , Gordon , MIDB , MyersLabonte , HCP (func/ only) , Tian (func/ only) SURF : midthickness , pial , white , inflated , vinflated
QC Pipelines: MRIQC & BME-X🔗
MRIQC extracts image quality metrics (IQMs) for each T1w/T2w and functional BOLD run and generates visual .html reports. The BME-X pipeline performs motion correction, resolution enhancement, denoising, and harmonization of MR images.
hbcd/
└── derivatives/
├── mriqc/
│ ├── sub-[ID]/
│ │ └── ses-[V0X]/
│ │ ├── anat/
│ │ │ └── sub-[ID]_ses-[V0X]_run-[X]_{T1w|T2w}.json
│ │ └── func/
│ │ └── sub-[ID]_ses-[V0X]_run-[X]_{T1w|T2w}.json
│ └── sub-[ID]_ses-[V0X]_run-[X]_{T1w|T2w}.html
│
└── bme-x/
└── sub-[ID]/
└── ses-[V0X]/
└── anat/
|__ sub-[ID]_ses-[V0X]_run-[X]_desc-{enhanced|preproc}_{T1w|T2w}.nii.gz (+JSON)
|__ sub-[ID]_ses-[V0X]_run-[X]_space-{T1w|T2w}_desc-brain_mask.nii.gz (+JSON)
|__ sub-[ID]_ses-[V0X]_run-[X]_{T1w|T2w}.nii.gz (+JSON)
How To Read File Trees →
MRI Derivatives Quick Start Guide🔗
Below is a summary of key MRI derivatives used for structural morphology and resting-state functional MRI (rsfMRI) functional connectivity analyses. Key derivatives, produced by the XCP-D pipeline, include volumetric and surface-based time series for each participant. The data release also includes dense and parcellated time series with at least 2.5 minutes of low-motion data (FD>0.3), functional connectivity matrices, regional homogeneity values, and amplitude of low-frequency fluctuation values.
Curvature, Sulcal Depth, & Cortical Thickness
File: anat/*_space-fsLR_den-91k_{curv|sulc|thickness}.dscalar.nii
Recommended for: Vertex-wise cortical morphology analyses (e.g., folding, curvature, thickness comparisons).
These CIFTI scalar files contain surface-based structural metrics derived from reconstructed L/R cortical surfaces, aligned to the fsLR template (~64k vertices per hemisphere).
- Curvature: Characterizes cortical folding and morphology; often used as a covariate in morphometric analyses.
- Sulcal depth: Complements curvature to describe cortical shape and folding complexity.
- Cortical thickness: Distance between pial and white matter surfaces (mm); typically averaged within ROIs or compared across participants to study development, aging, or group effects.
Parcellated Structural Measures
File: anat/*_space-fsLR_seg-{PARC}_stat-mean_desc-{curv|sulc|thickness}_morph.tsv
Recommended for: Region-based (ROI-level) analyses such as group comparisons or developmental modeling.
Tabulated summaries of cortical metrics (curvature, sulcal depth, thickness) within anatomical regions defined by parcellation atlases. These files provide regional averages for statistical modeling or visualization.
Midthickness, Pial, and White Matter Surfaces
File: anat/*_hemi-{L|R}_space-fsLR_den-32k_{midthickness|pial|white}.surf.gii
Recommended for: Visualizing cortical anatomy or mapping functional data to anatomical space.
3D surface models representing the midthickness, gray–white matter boundary, and pial surfaces for each hemisphere. Useful for rendering structural data, computing surface-based metrics, or visualizing functional overlays.
Dense Timeseries
File: func/*_task-rest_space-fsLR_den-91k_desc-{denoised|denoisedSmoothed}_bold.dtseries.nii
Recommended for: Voxelwise or seed-based FC analyses, timeseries analysis via sliding windows or markov chains, etc.
CIFTI dense time series containing fully preprocessed, temporally filtered, and nuisance-regressed BOLD data. These files combine the left and right surfaces, aligned to the standard fsLR surface template, with the subcortical volume annotated by subcortical structure. Each greyordinate (~96k total) represents a vertex or voxel with pre-processed resting-state functional MRI time-series.
Parcellated Timeseries
File: func/*_task-rest_space-fsLR_seg-{PARC}_stat-mean_timeseries.tsv
Recommended for: ROI-to-ROI connectivity or network analyses using mean BOLD signals per region.
Tabulated mean BOLD time series for each region in the
parcellation atlases.
Also available as CIFTI .ptseries.nii files, where columns = regions and rows = timepoints.
Connectivity Matrices
File: func/*_task-rest_space-fsLR_seg-{PARC}_stat-pearsoncorrelation_relmat.tsv
Recommended for: Quick inspection, validation, or as input to network and graph analyses.
Tab-delimited matrices of pairwise Pearson correlations between atlas regions, computed from parcellated time series using all available low-motion data (motion censored with a framewise displacement threshold of 0.3 mm). These matrices form the foundation for ROI-to-ROI connectivity analyses.
Motion Detection and Confound Files
File: func/*_task-rest_dir-PA_run-{X}_{design|motion|outliers}.tsv
Required for: Motion assessment and filtering low-quality data prior to group analyses.
Includes framewise displacement values and nuisance regressor design files. Design files contain one column per regressor (e.g., motion parameters, high-motion outlier volume indicators). See the XCP-D documentation for details.
See Parcellations & Atlases in the XCP-D documentation for more details.
| Atlas | Description |
|---|---|
Glasser |
Multimodal anatomical atlas (population-level) Derived from multimodal MRI data (Glasser et al., 2016) Surface-based morphology, population-level structure |
Gordon |
Functional atlas (333 ROIs) rs-fMRI boundary detection (120 young adults, ~14 min per subject; Gordon et al., 2016) Functional network mapping, group-level FC analyses |
HCP |
Multimodal cortical atlas (360 ROIs) Combined task, resting-state, and diffusion MRI (210 young adults; Glasser et al., 2013) Cross-modal structural–functional alignment |
MIDB |
Precision functional atlas (individualized) Derived from ABCD data using a 75% probability threshold (Hermosillo et al., 2024) Individualized functional network mapping |
Myers-Labonte |
Infant probabilistic functional atlas 50% probability threshold; infant population (Myers et al., 2023) Infant functional network mapping |
Tian |
Subcortical parcellation atlas High-resolution subcortical segmentation (Tian et al., 2020) Subcortical connectivity analyses |
4S{X}56Parcels |
Multimodal atlas (multi-resolution) Schaefer cortical parcellations (100–1000 parcels) supplemented with subcortical and cerebellar regions (AtlasPack) Cross-modality alignment across XCP-D, QSIPrep, and ASLPrep |
Adamson, C. L., Alexander, B., Ball, G., Beare, R., Cheong, J. L. Y., Spittle, A. J., Doyle, L. W., Anderson, P. J., Seal, M. L., & Thompson, D. K. (2020). Parcellation of the neonatal cortex using Surface-based Melbourne Children’s Regional Infant Brain atlases (M-CRIB-S). Scientific Reports, 10(1), 4359. https://doi.org/10.1038/s41598-020-61326-2
Dean III, D. C., Tisdall, M. D., Wisnowski, J. L., Feczko, E., Gagoski, B., Alexander, A. L., ... & HBCD MRI Working Group. (2024). Quantifying brain development in the HEALthy Brain and Child Development (HBCD) Study: The magnetic resonance imaging and spectroscopy protocol. Developmental Cognitive Neuroscience, 70, 101452. 10.1016/j.dcn.2024.101452
Fair, D. A., Schlaggar, B. L., Cohen, A. L., Miezin, F. M., Dosenbach, N. U. F., Wenger, K. K., Fox, M. D., Snyder, A. Z., Raichle, M. E., & Petersen, S. E. (2007). A method for using blocked and event-related fMRI data to study “resting state” functional connectivity. NeuroImage, 35(1), 396–405. doi: 10.1016/j.neuroimage.2006.11.051
Goncalves, M., Moser, J., Madison, T. J., McCollum, R., Lundquist, J. T., Fayzullobekova, B., Hadera, L., Pham, H. H. N., Moore, L. A., Houghton, A., Conan, G., Styner, M. A., Alexopoulos, D., Smyser, C. D., Stoyell, S. M., Koirala, S., Nelson, S. M., Weldon, K. B., Lee, E., … Fair, D. A. (2025). FMRIPrep Lifespan: Extending A robust pipeline for functional MRI preprocessing to developmental neuroimaging. In bioRxivorg. https://doi.org/10.1101/2025.05.14.654069
Power, J. D., Barnes, K. A., Snyder, A. Z., Schlaggar, B. L., & Petersen, S. E. (2012). Spurious but systematic correlations in functional connectivity MRI networks arise from subject motion. NeuroImage, 59(3), 2142–2154. doi: 10.1016/j.neuroimage.2011.10.018
Power, J. D., Mitra, A., Laumann, T. O., Snyder, A. Z., Schlaggar, B. L., & Petersen, S. E. (2014). Methods to detect, characterize, and remove motion artifact in resting state fMRI. NeuroImage, 84, 320–341. doi: 10.1016/j.neuroimage.2013.08.048
Power, J. D., Schlaggar, B. L., & Petersen, S. E. (2015). Recent progress and outstanding issues in motion correction in resting state fMRI. NeuroImage, 105, 536–551. doi: 10.1016/j.neuroimage.2014.10.044
Zöllei, L., Iglesias, J. E., Ou, Y., Grant, P. E., & Fischl, B. (2020). Infant FreeSurfer: An automated segmentation and surface extraction pipeline for T1-weighted neuroimaging data of infants 0-2 years. NeuroImage, 218(116946), 116946. https://doi.org/10.1016/j.neuroimage.2020.116946