Skip to content

Latest commit

 

History

History
141 lines (104 loc) · 8.84 KB

README.md

File metadata and controls

141 lines (104 loc) · 8.84 KB
Raw

NDA Study 1887: Structural MRI scans for Autism Subtypes study in BIDS format

This repository provides data curation scripts and a viewable README for the anatomical imaging dataset shared as NDA Study 1887. The NDA Study 1887 is associated with NDA Collection 2368, also known as, the Clinical and Immunological Investigations of Subtypes of Autism.

The study is described in detail in Raznahan et al. 2013 and Smith et al. 2016 among other publications.

We provide a script bidsify_1887.py to convert the data downloaded through NDA into a BIDS format dataset. The script is distributed with the NDA Study 1887 data package under results/ directory and is also available on this repository. Script usage instructions are described below.

NDA Data Download

NOTE: To access the study, you will need permissions to the "NIMH Data Archive" group under Data Permissions --> Active NDA Permissions. If you don't see "NIMH Data Archive" listed under "Active NDA Permissions", then you'll have to request it. More information on access requests can be found here.

To download data from NDA, the user would have to:

  1. Create a data package.
  2. Download the data package.

Creating a Data Package

Steps below demonstrate creating a data package on NDA

  1. Go to NDA study 1887 page and click on the Download button at the bottom of the page.

  2. The cart at top right corner should now have 173 subjects. Give it a few seconds to update. Click on Create Data Package/Add Data to Study button to see the next prompt.

  3. Provide a desired name for the new package. Make sure to check the Include associated data files to download NIfTI images along with the metadata.

  4. Go back to your account dashboard and click on Data Packages. It might take about 15-20 minutes to create the package but once it's ready you should see something like this under data packages list.

NOTE: As of 2023-03-29, these instructions are valid. However, this might not be the case in future. Please report it as an issue on this repo, if the instructions are not valid any more.

Downloading the Data Package

The NDA Study 1887 data package is less than 5 GB in size. It can be downloaded using one of two options:

  1. NDA Tools: Download instructions using the command line utility can be found at https://github.com/NDAR/nda-tools#installing-python (Recommended)

  2. Instructions to download a data package using NDA Download Manager can be found on https://nda.nih.gov/tools/nda-tools.html#download-manager-beta .

Here's a snapshot of the partial directory after the data package has been downloaded:

study1887
├── README.pdf
├── dataset_collection.txt
├── datastructure_manifest.txt
├── fmriresults01
│   ├── manifests
│   ├── sub-NDARXXXXXXXX
...
│   └── sub-NDARXXXXXXXX
├── fmriresults01.txt
├── md5_values.txt
├── package_info.txt
├── results
│   ├── CHANGES
│   ├── README.md
│   ├── bidsify_1887.py
│   ├── dataset_description.json
│   ├── participants.json
│   ├── participants.tsv
│   ├── scans.json
│   └── scans.tsv
└── study_1887.pdf

NDA Data Package to BIDS Directory

Here's an example to copy over the NIfTI and associated metadata files into a new directory.

python3 bidsify_1887.py -i nda-study-1887 -b bids-study-1887 -m copy

The user can also choose other file mapping methods such as softlink and move options. The help prompt for script is as follows:

$ python bidsify_1887.py -h
usage: bidsify_1887.py [-h] -i INPUT_DIR -b BIDS_DIR -m METHOD

This script generates BIDS formatted directory for NDA Study 1887 downloaded from NDA.

optional arguments:
  -h, --help            show this help message and exit
  -i INPUT_DIR, --input INPUT_DIR
                        Path to input directory with data downloaded from NDA Study 1887 containing fmriresults01/ and results/ subfolders.
  -b BIDS_DIR, --bids BIDS_DIR, -d BIDS_DIR, --destination BIDS_DIR
                        Path to output the BIDS formatted directory.
  -m METHOD, --method METHOD
                        Choose a method by which you'd like the files mapped: 
                        copy = Outputs copies to the destination BIDS directory. 
                        move = Moves files (without creating a copy) to the destination BIDS directory. 
                        softlink = Create a softlink (a.k.a. symbolic link or symlink) between NDA files and the destination BIDS directory.

Data Preparation Notes

Anatomical imaging data is shared in a minimally processed, raw format. However, in order to facilitate data analysis, the MRI data are converted to NIfTI and transformed into BIDS format using Dcm2Bids version 2.1.6, which is a wrapper for dcm2niix version 1.0.20211006. The following modality agnostic files have been shared as supporting documentation:

Filename Description
dataset_description.json A JSON file describing the dataset.
README A text file describing the dataset in greater detail.
CHANGES A text file with version history of the dataset (describing changes, updates and corrections).
participants.tsv A tab separated tabular file with additional information like age, sex, and group of each participant.
participants.json A JSON formatted data dictionary describing fields in participants.tsv.
scans.tsv A tab separated tabular file indicating the method used to deface every scan available in the dataset.
scans.json A JSON formatted data dictionary describing fields in scans.tsv.

The structural image types included in the dataset are T1w, T2w, FLAIR, PDw, and MTR. For Magnetization Transfer Ratio ( MTR) images acquired in the presence and absence of an MT pulse have the mt-on and mt-off entities in their filenames, respectively.

To preserve subject privacy, MRI scans are defaced using AFNI Refacer version 2.4. Defaced scans were visually inspected for quality using VisualQC's suite of QC tools. More details on the defacing workflow used can be found here.

8 of the 31 scans that failed first round of QC were manually defaced using FSLeyes image editor and the remaining 23 of 31 were programmatically corrected to ensure defacing quality. Defacing technique used for each scan in the dataset has been documented in the scans.tsv file.

Code availability

Scripts used for DICOM to BIDS format conversion and de-identification of anatomical MRI scans are available on the git repository at https://github.com/nimh-dsst/nda-study-1887.

References

  1. Study design of the Clinical and Immunological Investigations of Subtypes of Autism
  2. Armin Raznahan, et al., Mapping cortical anatomy in preschool aged children with autism using surface-based morphometry, NeuroImage: Clinical, 2013, https://doi.org/10.1016/j.nicl.2012.10.005.
  3. Elizabeth Smith, et al., Cortical thickness change in autism during early childhood, Human Brain Mapping, 2016, https://doi.org/10.1002/hbm.23195 .
  4. Other publications related to the study are listed in the dataset_description.json file.