Importing your dataΒΆ

Suppose you have completed the following steps

  1. You ran QSDR reconstruction with “save mapping” selected
  2. You saved a .trk file from the output from (1)
  3. You have a brain atlas aligned to the B0 volume used to create QSDR. This volume must be interpolated to the SAME grid as the B0 volume

You can launch the data importer from your shell

$ dsi2_import

and will see the following screen

data importer

To access the “Add new item” tab, right click in the white space. A new subject metadata input panel will appear below

label importer

A giant arrow points to a tab where you can add each of the atlases you’ve aligned to the B0 volume.

  • Under Subject Information:
    • Enter your subject’s data including their Scan id, Subject id, Gender, Study, and Group.
    • For “Fib file,” choose the “-.map.fib.gz” file in your output directory.
    • For “Pkl path,” create an output/file/path/name ending in “-.pkl” for your subject
    • For “Pkl trk path,” create an output/file/path/name ending in “-.pkl.trk” for your subject
  • Under “Reconstruction Information on the left-hand side:
    • Enter DSI Studio for “Software”
    • For “Trk file” on the left, choose the “-.trk” file in your output directory (if needed, unzip this file ahead of time):
data importer

Choose the “Label values” tab:

data importer

Add a new atlas by right clicking in the whitespace. Your b0-aligned atlases will be warped to QSDR space using the voxel mapping in the -map.fib.gz file.

Note: The color coding is important in the cells of this sheet. If a cell is red, it means that the current value does not point to an existing file and that this file MUST exist for data to be successfully imported. Blue cells indicate that the file does not yet exist, but will be created during the import process. White cells mean that the file exists on your filesystem.

Under “B0 volume path” :
  • Enter your “-ROIv_scale...thickened...” file
Under “QSDR volume path” and “Numpy path”:
  • Enter your QSDR thickened scale path and Numpy path (“-.npy”)
  • If these do not exist yet, you can type in an output/file/path/name that you would like to save them as and the program will create these for you

Here is an example:

data importer

Once all your cells are white or blue, you can click the Process inputs button to create any files needed for import. Once all the cells are white, you can click the Save button to write out a json file that can be loaded locally OR the Upload to mongodb if you have a local mongodb instance running.