search

martahedl / SC-construction-using-MSMT-CSD

0 stars 0 forks source link

Example data #2

Closed Lestropie closed 1 month ago

Lestropie commented 1 month ago

There are several outstanding questions regarding what to do with the example data in light of construction of this repository.

  1. Location: Currently the Dockerfile is pulling the data from this OSF repository. That is broadcast as being relevant to the BATMAN tutorial, which is intended to be superseded by the publication associated with this repository.

  2. Arrangement: OSF works best when there is just one or a small number of files to be downloaded, as direct download links are easily generated. That OSF repository currently exposes a large number of .dcm files, which is the converse to the ideal scenario. I've found a way to get the Dockerfile to download all files, but I don't know what the long-term reliability of that link will be. Also, the DICOM data seems to me to be duplicated in two different locations? Eg. BATMAN/dicoms/T1w/ and BATMAN/T1/?

  3. Data: As noted previously, the data originally acquired for this tutorial are not ideal. In particular, the diffusion sensitisation directions are not balanced for eddy currents, and this leads to a warning being issued by dwifslpreproc when the example data are processed. If we want to create a new repository to hold sample data that is more specifically intended for this publication, there is the opportunity to use a different acquisition session.

  4. Script interface: I think that it would be good to both have the pre-computed derivative data embedded within the repository, and be able to, with a single command-line invocation, re-generate those derivatives in an alternative location. The execution script will need to be configured to permit this.

martahedl commented 1 month ago

HI Rob,

1: Happy to upload the data onto github, if that makes things easier?

2: The data at "BATMAN/dicoms" is the correct data for this tutorial. The data under "BATMAN/T1" and "BATMAN/DWI" is outdated. I kept them since they are in line with an older version of the tutorial. But it's probably best to remove them now to avoid any confusion.

3: Related to the above: The dMRI data under BATMAN/dicoms should be balanced. The old version was not.

  1. Happy to run the code locally and upload the derived files onto github. Will provide that data asap.
martahedl commented 1 month ago

Created a new OSF repository as discussed. All raw T1w / dMRI data as well as all derivatives (expect for the tracks_10m.tck - too large) are uploaded. You'll find the OSF repository here: https://osf.io/tm5x8/. I invited you as a contributor so you should be able to read and write. The project itself is still private. Since the example code is also uploaded onto github, I'll close this issue.