search

openMetadataInitiative / bids2openminds

A tool to generate openMINDS metadata from BIDS datasets
MIT License
1 stars 3 forks source link

New techniques: inplane T1, inplane T2 #46

Closed spieschnik closed 3 months ago

spieschnik commented 3 months ago

How should inplane T1 and inplane T2 be mapped in openMINDS? Suggestions: 1. Mapping to MRIPulseSequence https://openminds.ebrains.eu/instances/MRIPulseSequence/T1 PulseSequence AND technique https://openminds.ebrains.eu/instances/technique/magneticResonanceImaging Indicating that inplane is a setting/parameter change of the suggested mapping and no new openMINDS instance needs to be created

2. Mapping to technique https://openminds.ebrains.eu/instances/technique/magneticResonanceImaging AND new openMINDS MRIPulseSequence for inplane Indicating that inplane is a its own MRI pulse sequence or even technique? A new openMINDS instance needs to be created. This 2. options makes more sense to me!

definition: T1-weighted anatomical image matched to functional acquisition (https://tu-dresden.de/bereichsuebergreifendes/nic/ressourcen/dateien/nic-data-management/series_description_naming_convention_20200618.pdf?lang=en)

definition: T2-weighted anatomical image matched to functional acquisition (https://tu-dresden.de/bereichsuebergreifendes/nic/ressourcen/dateien/nic-data-management/series_description_naming_convention_20200618.pdf?lang=en)

lzehl commented 3 months ago

@spieschnik @tgbugs and I had a discussion about this in the last ontology meeting. We can safely say the following:

"inplane T1" includes:

"inplane T2" includes:

We are still unsure how the "in-plane" vs "through-plane" has to be captured. we might just want to add them as additional techniques "in-plane imaging" vs "through-plane imaging".

tgbugs commented 3 months ago

@yarikoptic do you know a good source that explains what in-plane vs through-plane means in the context of BIDS or neuroimaging in general and might help understand whether inplane T1 is different than regular T1 imaging?

yarikoptic commented 3 months ago

Uff, you know how to ask simple albeit tricky questions @tgbugs ! ;)

I never used them so had to figure it out myself and somewhat mistrusted their description in https://bids-specification.readthedocs.io/en/stable/glossary.html#inplanet1-suffixes as "matched to a functional (task) image". So those are destined to be collected in the same orientation/planes as bold likely to make it easier (no registration "needed" presumably) to overlay functional results over to anatomy. As I reported their use seems to be going down -- typically older openneuro datasets have them.

This exploration lead me to discovery of a number of defects in our tools:

and raising an issue with bids-specification

lzehl commented 3 months ago

@yarikoptic thanks for helping us with interpreting these suffixes. My follow up question would be the following: is the difference between an "inplane T1" and "T1w" just the intend of matching it to a functional image series or are these technically differently created images?

yarikoptic commented 3 months ago

TL;DR: AFAIK overall just "intent",

Longer version: depending on what you mean by "technically different" though. As in "quantity depicted"/technology -- "the same" since AFAIK should be the same scanning parameters etc. As in "human actions/procedure to acquire" -- "different" since likely the scanning field of view is retained between bold and "inplane T1" so the slices do correspond, whenever for T1w they might not care and acquire larger field of view for anatomical than for bold/functional.

UlrikeS91 commented 3 months ago

@yarikoptic, thank you for the feedback! Then we will not add a new MRISequence for the inplane-T1 since it is the same as T1w.

@Peyman-N and @lzehl FYI

yarikoptic commented 3 months ago

FWIW, I recalled about NIDM-Terms for BIDS, but seems tobe quite not useful here: https://nidm-terms.github.io/

image

lzehl commented 3 months ago

@yarikoptic thanks. We can try discussing this at the INCF Assembly a bit more (I hope you are joining as well). There seems a bit of a mix in BIDS of techniques, intended use, and data types. We'll try to map as best as we can and then discuss the rest.

lzehl commented 3 months ago

incorporated in #9