Add coil element layout across coils

jcohenadad commented 3 months ago

Given that this paper is about comparing coils, an obvious question when interpreting B1+ and B1- results is: how are the Tx and Rx coils distributed for each coil? We should gather this information from each site (whenever possible, given that some are proprietary), and create a figure with this. Example of layout figure:

We could use the same color coding to display coil arrangement. For example, Tx elements in red, Rx in blue and Tx/Rx in purple.

Ideally, we would display the STL file in the notebook for interactive 3D view of each coil layout.

List of sites to generate a model from:

MGH/MNI: https://pubmed.ncbi.nlm.nih.gov/37439129/
CRMBM: 8ch Tx/Rx (Rapid Biomedical)
MPI: 24ch Tx/Rx (MRItools)
MSSM: 4ch Tx 22ch Rx + 4ch Tx 18ch Rx (Custom)
NTNU: Same as MPI
UCL: Same as CRMBM

👉 Centralized document with added information

honzink commented 2 months ago

@jcohenadad Just to double-check. This is the manuscript?

So I can mention title and co-authors, when I reach out to manufacturers. I will start when I come back from vacation end of July.

evaalonsoortiz commented 2 months ago

Yes, it is.

jcohenadad commented 2 months ago

So I can mention title and co-authors, when I reach out to manufacturers

it is better if I'm the point of contact with manufacturers-- these are very sensitive information that they shared and I prefer that they are not under the impression that I distributed it widely

honzink commented 1 month ago

Before I continue, this is for the coil in line # 2 and 4. Does that serve the purpose? They are CAD models so we can turn and shift and modify later.

It takes me some time to create them though, but it's the only way that I can think of.

jcohenadad commented 1 month ago

Before I continue, this is for the coil in line # 2 and 4. Does that serve the purpose?

Yes, that’s what I had in mind. We should sit down together so I can explain the fine details of how to display the elements.

One strange thing however, is that line #2, which is CRMBM, is an 8ch TxRx, which looks different than the what you showed

honzink commented 1 month ago

@jcohenadad I meant line # 2 in the centralized document. Or do we need only models for the ones you mentioned above?

jcohenadad commented 1 month ago

@jcohenadad I meant line # 2 in the centralized document. Or do we need only models for the ones you mentioned above?

Why are you modeling line #2 of the centralized document? It is not needed— see the list:

MGH/MNI: https://pubmed.ncbi.nlm.nih.gov/37439129/ CRMBM: 8ch Tx/Rx (Rapid Biomedical) MPI: 24ch Tx/Rx (MRItools) MSSM: 4ch Tx 22ch Rx + 4ch Tx 18ch Rx (Custom) NTNU: Same as MPI UCL: Same as CRMBM

honzink commented 1 month ago

CRMBM: 8ch Tx/Rx (Rapid Biomedical)

honzink commented 1 month ago

MGH/MNI. I had to doctor the Rx array to fit the human body model. Coil 3D model from Nibardo.

honzink commented 1 month ago

Changes: -White background -continuous conductors -human body model with higher resolution. I haven't found a fast way to make the human model smoother, it might take a day maybe more. I personally think it's fine, but up to you.

jcohenadad commented 1 month ago

Much better! Yes, I agree the human model looks fine as is. I would just make it lighter so it contrasts better with the coil elements.

Also, would it be possible to have the coil elements appear as 3D wires, as opposed to as a flat surface? They would appear more clearly (eg: we don't see the upper most element very well depending on its orientation)

honzink commented 1 month ago

Grey gives a good contrast, and is skin colour neutral.

3D wires would take me a day. I thickened the surface which serves the purpose, I hope.

jcohenadad commented 1 month ago

Great! Would you be able to make the gray much lighter?

honzink commented 1 month ago

How is this?

honzink commented 1 month ago

This is MSSM:

jcohenadad commented 1 month ago

On the MSSM coil, are the two posterior purple elements Tx/Rx? If it is Tx-only, I would pick another color (eg red)

honzink commented 1 month ago

Yes, that's purple, in their MRM paper they are TR elements.

jcohenadad commented 1 month ago

maybe we should pick another blue or purple, to better distinguish the two (I find them very close in intensity). Maybe a slightly lighter blue?

honzink commented 1 month ago

How is this?

honzink commented 3 weeks ago

Here the paper figure. Let me know if you want any changes.

paper figure.pdf

a) MPI: 24ch Tx/Rx (MRItools), NTNU: Same as MPI b) CRMBM: 8ch Tx/Rx (Rapid Biomedical), UCL: Same as CRMBM c) MSSM: 4ch Tx 22ch Rx + 4ch Tx 18ch Rx (Custom) d) MGH/MNI: https://pubmed.ncbi.nlm.nih.gov/37439129/

red: Tx elements, blue: Rx only elements, purple: TR elements

jcohenadad commented 3 weeks ago

excellent, thank you @honzink.

can you simply output each sub-figure as a separate file (I will create the figure myself in Keynote)
can you increase the brightness of the blue element (to make it easier to distinguish blue from purple)
can you provide a figure of the MNI/MGH coil without the red elements? I'll try to re-create them myself with slightly thinner feature (to not hide the Rx elements)

honzink commented 1 week ago

How is this blue?

jcohenadad commented 1 week ago

How is this blue?

much better, thanks!

honzink commented 1 week ago

Here the four subfigures:

MRI_tools.pdf MSSM_.pdf neuropoly.pdf rapid.pdf

jcohenadad commented 1 hour ago

@honzink for archiving purpose (eg: reproduce/modify figure in the future), I'd like to keep all figure sources inside a centralized location. Can you please point us to the location of the files with which you produced those figures? Thanks

spinal-cord-7t / coil-qc-code

Add coil element layout across coils #77