Outputs

Brainana writes all results under the output directory you specify (e.g. /output when using Docker). Outputs follow a BIDS derivatives layout.

Output directory layout

The diagram above is the canonical layout.

Subject directory (sub-<id>/)

Session directory (ses-<id>/, optional)

Anatomical outputs (anat/)

Scanner-space synthesis

• <ses_prefix>_space-scanner_T1w.nii.gz — T1w in scanner space (synthesized when multiple runs exist).

• <ses_prefix>_space-T2wScanner_T2w.nii.gz — T2w in native T2w scanner space (synthesized when multiple runs exist).

T2w in T1w scanner space (when T2w data are present)

• <ses_prefix>_space-scanner_T2w.nii.gz — T2w registered to T1w scanner space.

Preprocessed structural images — T1w space

• <ses_prefix>_space-T1w_desc-preproc_T1w.nii.gz — Preprocessed T1w in T1w space.

• <ses_prefix>_space-T1w_desc-preproc_T1w_brain.nii.gz — Skull-stripped T1w in T1w space.

• <ses_prefix>_space-T1w_desc-preproc_T2w.nii.gz — Preprocessed T2w in T1w space.

• <ses_prefix>_space-T1w_desc-preproc_T2w_brain.nii.gz — Skull-stripped T2w in T1w space.

• <ses_prefix>_space-T1w_desc-preproc_T1wT2wCombined.nii.gz — T1w/T2w combined image (T2w-enhanced contrast; when T2w is available).

Segmentation and masks — T1w space

• <ses_prefix>_space-T1w_desc-brain_mask.nii.gz — Binary brain mask.

• <ses_prefix>_space-T1w_desc-brain_hemimask.nii.gz — Hemisphere mask.

• <ses_prefix>_space-T1w_desc-brain_atlasARM2.nii.gz — ARM2 atlas segmentation.

• <ses_prefix>_space-T1w_desc-brain_atlasARM2.tsv — Color LUT for the atlas segmentation.

Template-space structural images

• <ses_prefix>_space-<template>_desc-preproc_T1w.nii.gz — T1w registered to template space.

• <ses_prefix>_space-<template>_desc-preproc_T2w.nii.gz — T2w registered to template space.

• <ses_prefix>_space-<template>_desc-brain_mask.nii.gz — Brain mask in template space.

Transform files

• <ses_prefix>_from-scanner_to-T1w_mode-image_xfm.mat — Scanner-to-T1w conformation (FSL .mat).

• <ses_prefix>_from-T1w_to-scanner_mode-image_xfm.mat — Inverse conformation.

• <ses_prefix>_from-T1w_to-<template>_mode-image_xfm.<ext> — T1w-to-template registration (ANTs .h5 or fireANTS .nii.gz).

• <ses_prefix>_from-<template>_to-T1w_mode-image_xfm.<ext> — Template-to-T1w inverse registration.

• <ses_prefix>_from-T2wScanner_to-scanner_mode-image_xfm.<ext> — T2wScanner-to-scanner(T1w) registration.

• <ses_prefix>_from-scanner_to-T2wScanner_mode-image_xfm.<ext> — Scanner(T1w)-to-T2wScanner registration.

Atlas backprojection (when registration is enabled)

Atlases are backprojected from template space into two subfolders:

• atlas_space-T1w/ — atlas-<name>_space-T1w_<ses_prefix>.nii.gz

• atlas_space-scanner/ — atlas-<name>_space-scanner_<ses_prefix>.nii.gz

Functional outputs (func/)

Session-level files omit task and run entities. Per-run files include them (see <run_prefix> above).

Session-level files are produced only when within-session coregistration is enabled (func.coreg_runs_within_session). When it is disabled, the session-level BOLD reference, brain mask, coreg reference, transforms, and session tSNR below are produced per-run instead (same naming patterns with <run_prefix>).

Session-level images (when coreg enabled)

• <ses_prefix>_space-scanner_desc-coreg_boldref.nii.gz — Session-averaged BOLD reference after within-session coregistration (scanner space).

• <ses_prefix>_space-bold_boldref.nii.gz — BOLD reference in bold (conformed) space.

• <ses_prefix>_space-bold_desc-brain_mask.nii.gz — Brain mask in bold space.

• <ses_prefix>_space-T1w_desc-preproc_stat-tsnr_boldmap.nii.gz — Session-level temporal SNR map (T1w space; when tSNR is enabled).

Session-level transforms (when coreg enabled; per-run when coreg disabled)

• <ses_prefix>_from-bold_to-T1w_mode-image_xfm.<ext> — BOLD-to-T1w coregistration.

• <ses_prefix>_from-T1w_to-bold_mode-image_xfm.<ext> — T1w-to-BOLD transform.

• <ses_prefix>_from-scanner_to-bold_mode-image_xfm.mat — Scanner-to-BOLD conformation.

• <ses_prefix>_from-bold_to-scanner_mode-image_xfm.mat — Inverse conformation.

Per-run images — bold space

• <run_prefix>_space-bold_desc-preproc_bold.nii.gz — Preprocessed BOLD in bold (conformed) space.

• <run_prefix>_space-bold_desc-preproc_boldref.nii.gz — BOLD reference in bold space.

Per-run images — T1w space

• <run_prefix>_space-T1w_desc-preproc_bold.nii.gz — Preprocessed BOLD in T1w space.

• <run_prefix>_space-T1w_desc-preproc_boldref.nii.gz — BOLD reference in T1w space.

• <run_prefix>_space-T1w_desc-brain_mask.nii.gz — Brain mask in T1w space.

• <run_prefix>_space-T1w_desc-preproc_stat-tsnr_boldmap.nii.gz — Run-level temporal SNR map (when tSNR is enabled).

Per-run images — template space

• <run_prefix>_space-<template>_desc-preproc_bold.nii.gz — Preprocessed BOLD in template space.

• <run_prefix>_space-<template>_desc-preproc_boldref.nii.gz — BOLD reference in template space.

• <run_prefix>_space-<template>_desc-brain_mask.nii.gz — Brain mask in template space.

Per-run confounds

Written in BIDS layout; nilearn-compatible for scrubbing via nilearn.interfaces.fmriprep.load_confounds.

• <run_prefix>_desc-confounds_timeseries.tsv — confound regressors, one row per volume. Columns:

  • Motion (24-parameter expansion): trans_x, trans_y, trans_z (mm), rot_x, rot_y, rot_z (radians), each with _derivative1, _power2 and _derivative1_power2 terms.

  • Framewise displacement: framewise_displacement — Power et al. (2012) sum-of-absolute-differences, using the macaque head radius (27 mm) to convert rotations to mm.

  • Relative RMS: rmsd — frame-to-frame RMS head displacement, computed from the motion parameters via the Jenkinson (1999) sphere formula (27 mm radius).

  • DVARS: dvars (non-standardized) and std_dvars (standardized). Require a brain mask; omitted (rather than computed over a whole-FOV mask) when no valid mask is available.

  • Global / tissue signals: global_signal (+ _derivative1/_power2/_derivative1_power2) — also requires a brain mask. csf, white_matter and csf_wm are added only when a T1w anatomical segmentation is available (i.e. skullstripping produced a multi-class segmentation + LUT); they are omitted otherwise. When a mask-based column is skipped, the JSON sidecar records the reason.

  • Outliers (indicator columns): motion_outlier## (FD or std-DVARS threshold crossings) and non_steady_state_outlier## (initial dummy volumes).

  Confounds are regressors only — the BOLD image is never scrubbed by this stage; columns are only indicators for optional user-customized downstream use. The first sample of differenced columns (derivatives, FD, rmsd, DVARS) is n/a.

• <run_prefix>_desc-confounds_timeseries.json — Per-column metadata sidecar (method, FD radius, FD/DVARS thresholds, and whether tissue regressors were produced).

QC figures (figures/)

• sub-<id>/figures/*.png — QC snapshot images for the subject.

Surface reconstruction (fastsurfer/)

When surface reconstruction is enabled, FreeSurfer-compatible outputs are written under fastsurfer/sub-<id>/ (or fastsurfer/sub-<id>_ses-<id>/ when multiple sessions are reconstructed separately). This includes label/, mri/, surf/, and stats/ (meshes, parcellations, and morphometric maps).

Quality control report

• sub-<id>.html — Browsable HTML report at the output directory root (alongside sub-<id>/, not inside it), with summaries, QC snapshots, and methods. View a sample report for sub-example.

The report is always generated on completion, even after a partial failure, and carries a status badge:

• Pass — completed successfully.

• Pass with warnings — completed, but one or more optional steps failed; some outputs may be missing.

• Fail — the run aborted early.

Pipeline reports (nextflow_reports/)

• nextflow_report_*.html — Nextflow execution report.

• nextflow_timeline_*.html — Nextflow timeline.

• nextflow_trace.txt — Task trace log.

• nextflow_dag.svg — Workflow DAG.

• anatomical_jobs.json — Discovered anatomical jobs and metadata.

• functional_jobs.json — Discovered functional jobs and metadata.

• config.yaml — Effective pipeline configuration used for the run.