BIDS Format

BIDS (Brain Imaging Data Structure) is the main data structure we support on XNAT. It provides a standardized way to organize neuroimaging data that unlocks access to numerous analysis pipelines.

What is BIDS?

BIDS is a community-driven standard for organizing brain imaging data. When you organize your data in BIDS format via dcm2bids, it opens your data up to many available pipelines such as MRIQC, fMRIPrep, TractoFlow, and many others.

Key Benefits:

Pipeline Compatibility: Works with modern analysis tools (fMRIPrep, MRIQC, etc.)
Standardized Structure: Consistent organization across research projects
Metadata Integration: JSON sidecar files preserve acquisition parameters
Data Sharing: Facilitates collaboration and reproducible research

For comprehensive information about BIDS, visit the BIDS Starter Kit.

BIDS on XNAT

How to Get BIDS Data:

Upload your DICOM data to XNAT
Run the dcm2bids pipeline
Your data will be converted to BIDS format in the rawdata directory

Resulting example structure:

Resources/
└── rawdata/
    ├── dataset_description.json
    ├── participants.tsv
    └── sub-<subject>/
        └── ses-<session>/
            ├── anat/
            │   ├── sub-<subject>_ses-<session>_T1w.nii.gz
            │   └── sub-<subject>_ses-<session>_T1w.json
            ├── func/
            │   ├── sub-<subject>_ses-<session>_task-<task>_bold.nii.gz
            │   └── sub-<subject>_ses-<session>_task-<task>_bold.json
            └── fmap/
                ├── sub-<subject>_ses-<session>_dir-<dir>_epi.nii.gz
                └── sub-<subject>_ses-<session>_dir-<dir>_epi.json

anat: Structural images (T1w, T2w, FLAIR)
func: Functional images (bold, sbref)
fmap: Fieldmap images for distortion correction
dwi: Diffusion-weighted images

File Naming and Organization

BIDS uses systematic file naming with key-value pairs separated by underscores:

Basic Pattern: sub-<subject>_[ses-<session>_][task-<task>_]<suffix>.nii.gz

Each NIfTI file has a corresponding JSON file with the same name containing acquisition metadata.

For detailed naming conventions, see the BIDS specification on files and folders.

BIDS Validation

dcm2bids pipeline includes built-in BIDS validation
Use pipelines such as the ari-validator for project-specific validation (if available for your project)
External validation: BIDS Validator
Complete specification: BIDS Specification

Available BIDS Pipelines

Once your data is in BIDS format, you can run these pipelines:

MRIQC - Automated quality metrics and reports
fMRIPrep - Robust fMRI preprocessing
TractoFlow - Diffusion MRI processing
Additional format conversions available through BIDS-compatible tools

Next Steps

Convert your data: See dcm2bids to get started
Run quality control: Use MRIQC to assess data quality
Preprocess your data: Choose appropriate preprocessing pipelines
Learn more: Visit the BIDS Starter Kit for comprehensive tutorials