Formalize participants' age to clarify the reference point

[yarikoptic]

A satellite issue to

• https://github.com/bids-standard/bids-specification/issues/1633 and could be of interest to @bids-standard/bep032 .

ATM age column is assumed to be "since birth" but in some experiments, animal or human (fetal MRI) or some organoids, it would not really be "since birth". To acknowledge the possibility of such different reference points e.g. in dandischema we defined AgeReferenceType and ATM we just use BirthReference but allow for GestationalReference happen we hit such a dataset.

I think it would be valuable for us to think in BIDS about such annotations and how we could handle them. Since somewhat of a "niche" issue really, a simple workaround solution could potentially be to rely on TermURL side car field to point to specific ontology defining gestational age, e.g. http://www.ebi.ac.uk/efo/EFO_0005112 (attn @satra -- should we use this or some other ontology right away in dandischema?). That would allow to disambiguate.

[satra]

the animal/cell world is more complicated including notions of stages and not age, and often characterized in relation to species. here is an ontology of age that defines postgestation and postnatal as terms: https://bioportal.bioontology.org/ontologies/FMA/?p=classes&conceptid=http://purl.org/sig/ont/fma/fma312872

[satra]

it may also be helpful to separate out age concepts versus age related data elements. for example age_at_death from EFO can be useful for ex vivo datasets of structure and cannot be captured by age as a keyword alone.