Given that we don't need to worry about anonymizing phantom data, here's what I'm thinking:
- Use phantom-specific unique identifiers for
sub, if we have that info. We'll need to figure out what the situation there is.
- Use the scan date for
ses. YYYYMMDD of course.
- Include pictures for the phantoms somewhere in the dataset. Just for reference.
- Use
acq field for different protocols.