NOTE: This code is now deprecated in favor of pypmi

PPMI to BIDS

The code and software contained in this repository is designed to convert neuroimaging data from the Parkinon Progression Markers Initiative (PPMI) into BIDS format.

Table of contents

• Description
• Requirements
• Instructions

Description

Requirements

• Python 3.5 (or higher) with the pandas library installed
• Singularity v2.4 or higher (N.B. this has only been tested with Singularity 2.4 and 2.5)

Instructions

1. Download raw DICOM data from the PPMI website and unzip it into the PPMI folder. The organization of the PPMI directory should look something like:

   └── PPMI/
       ├── 3000/
       |   ├── scan_type_1/                        (e.g., MPRAGE_GRAPPA)
       |   |   ├── 2014-03-10_10_25_03.0/
       |   |   |   └── SXXXXXX/
       |   |   |       └── *dcm
       |   |   └── 2015-04-04_09_58_25.0/
       |   ├── scan_type_2/                        (e.g., AX_FLAIR)
       |   |   ├── 2014-03-10_10_34_42.0/
       |   |   └── 2015-04-04_10_05_31.0/
       |   ├── .../
       |   └── scan_type_N/
       ├── 3001/
       ├── .../
       └── NNNN/

   That is, subjects should each get their own directory (3000, 3001, ..., NNNN). Within each subject directory there should be a unique directory for each scan type and, within those, a directory for each visit. This is the default layout that occurs if you download DICOM data from the PPMI website, so you shouldn't have to do any re-organizing!

2. Download this Singularity container with the correct versions of the conversion software and place it in the code directory.

3. Run bash code/run_conversion.sh and go to lunch. Check out the docstring at the top of that script for more info on what it does. Biefly, it uses code/ppmi_prep_heudiconv.py to do some light reorganizing of the downloaded PPMI data and then runs heudiconv to convert the DICOM data to BIDS format, using code/ppmi_heuristic.py to map the raw DICOM series to BIDS-compatible naming conventions. The output BIDS data will be in the data directory.

4. Use the organized PPMI BIDS dataset located in data for analyses!