Skip to main content

Table 5 Neuroimaging pipelines: frameworks, tools, analysis types, output variables

From: Cerebral MRI in a prospective cohort study on depression and atherosclerosis: the BiDirect sample, processing pipelines, and analysis tools

Processing step

Framework

Tool

Input

Type

Variable/biomarker

Standardization: NII and BIDS conversion

R

BIDSconvertR [15]

dicom

dcm2niix [16] BIDS-conversion

json-metadata, id, birthdate, weight

Quality control

Docker

MRIQC [8]

T1w, T2w, bold

MRIQC pipeline

See [8]

Anatomical pipelines

SPM

CAT12 [14]

T1w

VBM

Volume (native/normalized): GM, WM, CSF, WMH + mask

FSL

fsl_anat

T1w, T2w FLAIR

Anatomical pipeline

Volume (native/normalized): GM, WM, CSF

Freesurfer (v6 and v7.1.0)

recon-all (surfer.nmr.mgh.harvard.edu/)

T1w

Cortical thickness

Cortical thickness, ROI-wise

Lesion delineation pipelines

SPM

CAT12 [17]

T1w

Lesion segmentation (intensity-based)

Volume (native/normalized): WMH + mask

FSL

BIANCA [21]

T1w (BET, denoised, FLAIR space)

Lesion segmentation (trained—KNN)

Lesion count + volume (ml) + mask

T2w FLAIR (BET, denoised)

Diffusion-weighted pipelines

FSL

PSMD-Marker [23]

DWI + .bval + .bvec

Diffusion-weighted imaging

PSMD, MSMD, FA/MD (native/normalized and ROI-wise),

TBSS

Functional pipelines

Docker

fMRIprep [24]

T1w + bold

Anatomical and functional preprocessing

Structural and functional derivatives; see [24]

With (disabled Freesurfer processing)

  1. BET Brain-extracted, CSF Cerebrospinal fluid, DWI Diffusion-weighted image, FA Fractional anisotropy, FLAIR Fluid-attenuated inversion recovery, Gm Gray matter, KNN K-nearest neighbors, MD Mean diffusivity, ROI Region of interest, T1w T1-weighted-image, T2w T2-weighted image, TBSS Tract-based spatial statistics, VBM Voxel-based morphometry, Wm White matter