fiuneuro / fiu-cis-phantoms Goto Github PK

@mriedel56 has experience setting up automated download workflows for new data, so we can probably adopt that approach for conversion and pushing to GIN.

We want to version control with datalad and GIN. I haven't done this before, but the datalad handbook has info here.

It would be preferable to incorporate the version control into the conversion step itself.

Given that we don't need to worry about anonymizing phantom data, here's what I'm thinking:

1. Use phantom-specific unique identifiers for sub, if we have that info. We'll need to figure out what the situation there is.
2. Use the scan date for ses. YYYYMMDD of course.
3. Include pictures for the phantoms somewhere in the dataset. Just for reference.
4. Use acq field for different protocols.