Optogenetics/implant BEP

[TheChymera]

As part of BEP032 the issue of documenting implants popped up.

I've dealt with this in the past as part of opto-fMRI in animals, but originally decided against including that in BIDS since it seemed like burdening the standard with too many ancillary details. However, since it's now being done for e-phys anyway, maybe a dedicated BEP to make sure implant support is provided either for both modalities in their own ways or in a modality-agnostic fashion would be a good idea.

The details I required for my experiments were:

• Implant insertion site:
  • dorsoventral coordinate (float, mm)
  • leftright coordinate (float, mm)
  • rostrocaudal coordinate (float, mm)
  • anatomical reference (string, “bregma” or “lambda”)
• Implant insertion:
  • pitch (float, deg)
  • yaw (float, deg)
  • roll (float, deg — usually zero since optogenetic implants have radial symmetry, might be different if this is to cover e-phys as well)
  • depth (float, mm)

The reason for not just specifying the target is that the target is the intention of the operation and neither the description of the operation nor necessarily the result of the operation. The info above documents the operation, and determining where the target ended up being is an analysis aspect. This has the advantage of providing a description of the intervention, and not a summary which may be contingent on thresholding/etc as part of an analysis process applied to MRI data. I imagine determining the exact actual target is even more unreliable in e-phys (as I understand BEP032 the proposed method is operator image evaluation).

This is in some contrast to BEP032 which specifies:

• Probe:
  • x
  • y
  • z
  • width
  • height (this is — unless I misunderstand the text — a probe parameter but it is also considered to be the insertion depth)
  • depth (this is the probe depth, not the insertion depth, i.e. the thickness which is orthogonal to both height and width)
  • dimension_unit
  • alpha_rotation
  • beta_rotation
  • gamma_rotation
  • hemisphere
  • associated_brain_region(_id)
  • reference_atlas (for associated brain region disambiguation)

The list above excludes parameters which are relevant for the respective sorts of implants (e.g. numerical aperture and transmittance for optic fibers) which — if we should want to track them — would necessitate additional metadata space for each modality.

I'm curious whether you think that:

1. adding optogenetics implant support would be relevant for BIDS.
2. If so whether this should be done in a dedicated optogenetics PR or merged with the implant support of BEP032.
3. If merged with the implant support of BEP032, whether this should be a part of BEP032 or be extracted into a dedicated all-implant BEP, focusing on the parameters relevant for any sort of implant (and excluding e.g. optical or e-phys specific parameters)

@chrisgorgo @yarikoptic @bendichter

[bendichter]

@TheChymera I mostly agree, but I would rather not use aircraft rotational coordinates (yaw, pitch, and roll), since they only have meaning if you know what the front and top of the probe are. Euler angles (alpha, beta, gamma) have a defined relationship to the x,y,z axes, so if one knows what the positive x,y and z directions are on the probe you can determine the rotational coordinates in Euler angles. If we do use aircraft coordinates, we'll need some way of determining what is the front and what is the top.

[yarikoptic]

attn @effigies @bids-standard/bep032