Files and folders

Within the bids standard filenames contain information label such as subject name, data type and sessions followed by a suffix containing extra information. Each label if followed by a value \<label\>-\<value\> connected by a "-". Each label is concatenated by a "\_", for example sub-PAT001_ses-001. After the last “_” any number of suffixes can follow. The data type is the first suffix, for example sub-PAT001_ses-001_dwi.nii.gz.

File Names

Next to the folder structure, the file names of a Muscle-BIDS dataset will tell a lot about what kind of file it is and what it contains. A Muscle-BIDS name contains a list of entities followed by a list of suffixes and is ended with an extension. In Muscle-BIDS filenames "-" and "_" have special meanings and can never be used outside their intended use.

• An entity starts with a known key followed by its value (label or index) separated by an "-” and multiple entities are separated by a "_" Example: <key1>-<label>_<key2>-<index>
• A suffix or list of suffixes comes after the entities and each suffix is separated by a "_"
  • The first suffix describes the data type and must be maintained
  • The rest of the suffixes can be ones chosen from the defined suffixes per data type or a custom one Example: <key1>-<label>_<key2>-<index>_<suf1>_<suf2>
• The extension finalizes the file name Example: <key1>-<label>_<key2>-<index>_<suf1>_<suf2>.<ext>

Based on the Muscle-BIDS filename and its location it should be clear what data is contained within the file. Some examples are:

• “sub-test01_ses-001_megre_b0.nii”
• “sub-test01_ses-001_rep-flexion_dwi_dti_md.nii”
• “sub-anonab_ses-002_stk-upper_wt2_t2.nii.gz”
• “sub-anonab_ses-002_stk-upper_wt2_t2.json”

Study folder

A Muscle-BIDS study is one study folder (dir) that holds the file config.json and the following folders (the default names, which can be changed in the folders section of the config). QMRITools creates the ones it needs.

After a run the study folder looks like this (for one subject 001_1 with Dixon data):

<study>/
  config.json
  01_sourcedata/001_1/
  02_rawdata/sub-001/ses-1/
    raw/                    dcm2niix output
    dix/                    sub-001_ses-1_stk-DIXON1_megre_dix_*.nii and .json
  03_derivatives/sub-001/ses-1/dix/    fitted maps and *_check.json files
  04_merged/sub-001/ses-1/
    dix/                    merged maps: sub-001_ses-1_megre_dix_wat.nii
    seg/                    sub-001_ses-1_seg_auto_megre_dix_outph.nii
  05_analysis/
    sub-001/ses-1/          sub-001_ses-1.xlsx
    All_<date>.xlsx         all subjects in one table

More details and a complete step by step example are given in the Muscle-BIDS manual.

Folders

To prevent ambiguity most MRI acquisitions are stored in predefined folders. Muscle-BIDS-compliant software will look in these folders for their needed raw source data. The following folders have been defined in Muscle-BIDS. QMRITools uses the following subfolders to store data in the Muscle-BIDS dataset.

• anat: Structural imaging for anatomical or diagnostic reference (e.g T1 or T2).
• dwi: Diffusion-weighted data (e.g. DTI or IVIM).
• dix: Dixon data.
• quant: Relaxometry data (e.g. T1 or T2 mapping).
• seg: Segmentation labels (e.g. manual segmentations).
• raw: Unspecified non-BIDS files (the dcm2niix output before sorting).

An example of a folder structure is shown below:

• Entities and suffixes
• Files and folders
• Muscle-BIDS
• Muscle-BIDS step by step