DICOM Data Image Sharing

The Adolescent Brain Cognitive Development study (ABCD) is sharing raw DICOM images in a BIDS like file directory format. This script will anonymize DICOM files, add meta-data and create a TGZ file suitable for sharing on a platform such as NDA (National Data Archive).

The Brain Imaging Data Structure (http://bids.neuroimaging.io) describes a file format for sharing of nifti-format encoded volumetric data. ABCD is using its directory structure to share DICOM data. In a trivial process (dcm2nii) the data generated by this tool can be converted to a fully BIDS compatible structure. As is BIDS tools might not be able to process the provided data.

Input Data

Data exists on the local file system packaged as TGZ encoded directory that contain all DICOM files of a single image series. Together with the DICOM files the TGZ also contains a single json file (extension .json) that describes the data. This file is created by the processSingleFile python script available on the FIONASITE repository (FIONASITE/server/bin/processSingleFile.py). Here an example for the content of the generated file:

{
    "AccessionNumber": "",
    "AcquisitionLength": "TA 06:00",
    "AcquisitionMatrix": "[90, 0, 0, 90]",
    "ActiveCoils": "HEA;HEP",
    "Anonymized": "fast track v1.0, @DAIC April 2017",
    "ClassifyType": [
        "SIEMENS",
        "mosaic",
        "original",
        "ABCD-SST-fMRI"
    ],
    "DateOfBirth": "20020215",
    "EchoTime": "30",
    "ImageType": "['ORIGINAL', 'PRIMARY', 'M', 'ND', 'MOSAIC']",
    "IncomingConnection": "removed",
    "InstanceNumber": "445",
    "Manufacturer": "SIEMENS",
    "ManufacturerModelName": "Prisma_fit",
    "Modality": "MR",
    "NumFiles": "removed",
    "PatientID": "NDAR_INVXXXXXXXX",
    "PatientsAge": "116",
    "PatientsSex": "M",
    "PhaseEncodingDirectionPositive": "0",
    "RepetitionTime": "800",
    "SOPClassUID": "MR Image Storage",
    "ScanningSequence": "EP",
    "SequenceName": "epfSM2d1_90",
    "SequenceVariant": "SK",
    "SeriesDescription": "SIEMENS, mosaic, original, ABCD-SST-fMRI",
    "SeriesInstanceUID": "XXX",
    "SeriesNumber": "28",
    "SeriesTime": "122945.581000",
    "SliceLocation": "-59.027603148797",
    "SliceSpacing": "2.3999999741376",
    "SliceThickness": "2.4000000953674",
    "StudyDate": "20170122",
    "StudyDescription": "Adolescent Brain Cognitive Development Study",
    "StudyInstanceUID": "XXX",
    "StudyTime": "112752.432000",
    "siemensDiffusionInformation": [],
    "siemensUUID": "X-X-X"
}

Output

The script will create an output TGZ file that contains a directory structure that follows BIDS closely.

sub-NDARINVXXXXXXXX/
  ses-baselineYear1Arm1/
    func/
      ABCD-SST-FMRI_run-20000000/
      ABCD-SST-fMRI_run-20000000-EventRelatedInformation.txt
      ABCD-SST-fMRI_run-20000000.json

Workflow

Together with the output TGZ the script will store NDA's image03 information in a sqlite database. After anonymizing a sufficient number of image series a call of anonymizer.sh with the option '-e' will export this information as a csv file suitable for the NDA upload tool.