Skip to content

Autism Brain Imaging Data Exchange I (ABIDE I)

From the ABIDE I webpage:

The Autism Brain Imaging Data Exchange I (ABIDE I) represents the first ABIDE initiative. Started as a grass roots effort, ABIDE I involved 17 international sites, sharing previously collected resting state functional magnetic resonance imaging (R-fMRI), anatomical and phenotypic datasets made available for data sharing with the broader scientific community. This effort yielded 1112 dataset, including 539 from individuals with ASD and 573 from typical controls (ages 7-64 years, median 14.7 years across groups). This aggregate was released in August 2012.

Steps already done to produce the preliminary data dictionary saved here

  1. Went to the ABIDE I webpage.
  2. Scrolled down to the ABIDE I Downloads section
  3. Clicked the Phenotypic Data Legend link and downloaded the PDF directly to this ABIDE_I/ subfolder.
  4. Did not rename the PDF file from its originally downloaded name.
  5. Took screenshots of each table within each page of the PDF and named them as ABIDE_LEGEND_V1.02_PAGE_#.jpg

  6. Installed and used the Optical Character Recognition (OCR) software Tesseract to convert JPEG tables to text in TXT files, like the following.

    shell tesseract ABIDE_LEGEND_V1.02_PAGE_1.jpg ABIDE_LEGEND_V1.02_PAGE_1

  7. Renamed each TXT file to a TSV file and "manually" edited, as necessary, to match each PDF table's intent using tabs instead of spaces as field separators.

  8. Concatenated the two table TSV files into a single TSV file and committed that here to Git version control.

Steps to produce this study's data dictionary

  1. Install the required Python 3 library with the following line of code:

    shell python3 -m pip install --user pandas

  2. Run the following line of code within the ABIDE_I/ subfolder:

    shell python3 dictionary.py

Notes about this data dictionary

  1. The SITE_ID field is injected from the ABIDE I webpage.
  2. The BIDS Description is mostly composed here of the comma-separated list of VARIABLE TYPE then MIN then MAX because the unique combination of these three describes the data in that field well.
  3. The entries for FIQ, VIQ, and PIQ were "manually" adjusted to have simplified Description and no more Levels.
  4. The entries for FIQ_TEST_TYPE, VIQ_TEST_TYPE, and PIQ_TEST_TYPE were "manually" adjusted to include better values in each level of the Levels and a shorted Description.
  5. You can review the exact manual fixes toward the bottom of the dictionary.py script.

Steps to convert this study's phenotype data file from CSV to TSV format

  1. Log in on NITRC repository. Register for an account, if you don't already have one.
  2. Download raw phenotype data from ABIDE I page.
  3. Run the following line of code within ABIDE_I/ subfolder with appropriate file paths in place of <INPUT_FILE> and <OUTPUT_DIR>: python data_convert.py -i <INPUT_FILE> -o <OUTPUT_DIR> NOTE should be the path to file downloaded in step 2.

Notes about data conversion script

  1. The ADI_RRB_TOTAL_C field has been renamed to ADI_R_RRB_TOTAL_C to match the one in dictionary.json.
  2. The ADOS_GOTHAM_SOCAFFECT field has been renamed to ADOS_GOTHAM_SOC_AFFECT to match the one in dictionary.json.