Rename modalities.yaml in schema

[tsalo]

The file modalities.yaml describes the technologies supported in the specification (e.g., MRI, EEG). Given that "modality" has its own meaning, and will soon have a distinct definition in Common principles (see #592), the naming of the file in the schema is misleading. Some viable alternatives might be "technologies.yaml" or "instruments.yaml".

[tsalo]

I'm leaning toward "instruments.yaml" after reading something @robertoostenveld wrote. Additionally, I like the idea of combining the EEG and iEEG data types (i.e., folders) under a shared instrument like "biopotential amplifiers" (again, per @robertoostenveld). What do folks think of this structure?

---
mri:
  name: Magnetic Resonance Imaging
  datatypes:
    - anat
    - dwi
    - fmap
    - func
bioamp:
  name: Biopotential Amplification
  datatypes:
    - eeg
    - ieeg
meg:
  name: Magnetoencephalography
  datatypes:
    - meg
beh:
  name: Behavioral experiments
  datatypes:
    - beh

While this doesn't affect the specification at all at the moment, I think it would eventually impact the organization of the text. Currently, we have an MRI page with subsections for the different data types derived from the same instrument, while EEG and iEEG have separate pages. Ultimately, iEEG and EEG would also fall under one page (or major section) named "Biopotential Amplification".

[yarikoptic]

Although I like instruments, it sounds a bit "too specific". More appropriate term might be "Measurement Technique" or may be "Measurement Method" ( @satra ?) for which there is even an MMO ontology which lists MRI, EEG and may be more.
Filename then indeed might grow in length to measurement_methods.yaml or alike

[robertoostenveld]

@yarikoptic the thing I wrote that @tsalo refers to is this https://docs.google.com/document/d/1cY-iZL8rtgDJ1-YVs1f2h5ugOiuK144zu4ju9UXZEGY/edit#heading=h.r2qz45a4irn1. It is not yet ready for wide dissemination, as I would like to get some more feedback from specific people before opening it up for wider discussion. But then... it clarifies the organization and naming proposed here.

[tsalo]

@robertoostenveld sorry, I knew it wasn't ready for dissemination, so I tried to hint around it, but the proposals there seemed to be directly relevant to this issue.

[robertoostenveld]

Reegarding Instrument versus Measurement Technique or Measurement Method: people working with sEEG will consider their "technique" or "method" to be totally different than ECoG (and rightfully so), but from a data perspective (i.e. someone who wants to (re)use data that already sits on disk) the two are more similar than different, with the most crucial difference being in the placement of the electrodes. Idem for OPM-based versus cryogenic MEG systems. Hence trying to make it specific about the instrument that converts the (complex) measurement to data on disk.

[tsalo]

@robertoostenveld If I'm understanding correctly, you're saying that Instrument is a level that reflects the purpose of the data better than Measurement Technique or Measurement Method?

I should say that this can probably be changed later. Currently, modalities.yaml stands in direct contradiction to the definition of "modality" (esp. once #592 is merged), and the file isn't actually used for anything yet. So as long as the new name doesn't conflict with existing terms in the specification, it should be fine.

@yarikoptic @robertoostenveld What do you both think of the organization I proposed? Basically demoting EEG and iEEG from "instruments" or "measurement techniques" down to just "datatypes" (and of course suffixes) and nesting both under "Biopotential Amplification"? I would end up splitting up the entity tables to match the new convention.