Data Management

CFMI uses Box as our file sharing platform. We aim to have data on Box within 48 business hours. If after two business days, you notice that your data is not up, please email us. Box should be used for file sharing, not file storage, so please download and inspect your data promptly. Scans 4GB and larger will be deleted from the CFMI console 48 business hours after collection.

Need access to Box?
Because scan data is PHI, access to Box folders must receive PI approval. PIs can reach out directly, or be copied into a request over email. Please use the cfmi-scanning-team@georgetown.edu email address for all matters concerning Box access and data. 

Georgetown users can log in to Box at https://georgetown.app.box.com using their SSO (netID and password). External users (non-Georgetown) will receive an email invitation to Box if an account does not exist. For users from American University, it is important to note that your student.american.edu and american.edu email addresses are not equivalent for signing into Georgetown Box, so take note of which email your account is under. 

• In Box, click into the folder you wish to download. If you have difficulty navigating to this folder, feel free to reach out, and we can provide you with a link to your parent folder. 
• Inside each subject folder, will be a folder for each date the ID was scanned. Click into the date of the study session
• You should see a zip file that matches the ID and date of your study visit in the format ID_YYYY_MM_DD. Click into the .zip file.
• This will pull up a preview. Click download in the top right corner. It will download a .zip file for the entire session. This may take a few minutes. 
• You can now move the folder to its intended directory on your device. Navigate to the folder in your system and fully unzip the folder. You will notice that they have the extension .dcm*, these are your DICOMs (raw image files). 

*data prior to OCT 2022 will have the extension .IMA

The files are in DICOM format, which means you’ll need to download software to convert them to NIfTI and view them. We recommend the use of Chris Rorden’s dcm2niix**. Below I’ve listed both a GUI and a Command Line approach to doing this.

Command Line (Recommended; Faster Conversion)

• Download Chris Rorden’s dcm2niix in the most recent version. Using your command line (bash recommended), open the directory with the DICOM (.ima/.dcm) data. 
• The simplest command for this is dcm2niix /path/to/dicom/folder, which will execute it using default behaviour. See below for a list of default behavior:

• Place uncompressed NIfTI files (.nii) in the same folder the DICOMS are in.
  • If you would like compressed NifTI files (.nii.gz), or save in a custom location, please consult the documentation. 
• Produce a .json text file (BIDS sidecar) that contains important information about the scan in a human (and machine) readable format.
  • This can also be customized, but the default works fine. 
• Names NifTIs with the schema %f_%p_%t_%s.
  • Our strong suggestion is to change this to to use the series description* to name the files in addition to the protocol as it defaults to. You can do this by including %d in the format ( -f ) argument
    • e.g. dcm2niix -f %i_%4s_%p_%d_%t /path/to/dicoms  
    • Refer to this documentation on file naming in dcm2niix for arguments for more information about its naming schemas. 

TL;DR:  You can customize your output to fit your needs, otherwise, we recommend using  dcm2niix -f%i_%4s_%p_%d_%t /path/to/dicoms 

GUI (much slower, but user friendly) 

You’ll then want to open the NifTI files in a viewer such as MRIcroGL, MRIcron, FSLeyes, etc.  See below for instructions on opening and using MRIcroGL.

You can use the most recent release** of MRIcroGL. When you open this, navigate (in top bar) to Import > Convert DICOM to NIfTI. This will open the GUI. 

• You can alter the formatting of the filenames using Output Filename. Our suggestion is to set dcm2niix settings to use the series description (%d)* to name the files, in addition to the protocol (%p) as it defaults to.  
• You can select the format using Output Format: The default for this GUI is Compressed NIfTI (.nii.gz), but you can select Uncompressed NIfTI (.nii) in the dropdown. 
• It will also, by default, produce a text file in BIDS JSON format (.json) that contains important information about the scan in a human readable format.  
• Lastly, either drag your folder into the window, or Select Folder to Convert. 

Once your files are in NifTI format, you can open them with your preferred DICOM viewer. This demonstration is for MRIcroGL. Go to File > Open and select the scan you wish to view.  

The version without _ND will be prettier, and is more suitable for viewing. The version with _ND is more suitable for data analysis.