The participants.tsv table has a column for age, but what’s the official way to record ages that differ across sessions? I appreciate that there are ways to do this with the various “additional columns” (e.g., put all age information in the *sessions.tsv files), but I’m wondering about an official (or officially recommended) approach. This question applies to some of the other recommended columns, too.
I consider the sessions.tsv the “official” way to record any information that varies across sessions, even though age isn’t specifically referenced in the spec. That’s also how Nibabies extracts session-wise age information from BIDS datasets.
On a related note, I believe BEP036 proposes a top-level sessions.tsv file that would have one row for each participant-session combination.
1 Like
You can add any columns you want to the participants tsv file (eg age_time1, age_time2), and then just make sure they are described in a participants.json file.
Best,
Steven
I am not sure this is “legit”. I.e. generic Inheritance principle applies to sessions.tsv at the top level and pretty much it just boils down to have a single row per each session_id. left a comment in bep
See and chime into the phenotype related issues:
See in particular: Age at session · Issue #1020 · bids-standard/bids-specification · GitHub
Thanks all! I’m gathering that there isn’t yet an official way, but there are a few places where relevant conversations are ongoing (especially BEP036 and Issue #1020).
1 Like