CAT12 segmentation order

[AlexBo924]

Hi Edouard,

We spoke last time about the little option in nirstorm for the 5 tissues, I see that in the latest release you have added the option of choosing between the nirstorm format (1) skin 2) skull 3) csf 4) gm 5)wm or the standardized format.

However I see that for the standardized option we have 1 for WM and 2 for GM, while SPM/CAT segmentation give you 1 for GM and 2 for WM.

I also checked the added lines from Francois which also confirms 1 for GM or 2 for WM ??

If possible to double check.

Thanks

[Edouard2laire]

Hi, From what I understand in your picture is 1. White matter, 2. Gray matter, 3. CSF, 4. Skull, and 5.Scalp. This is confirmed here : https://neuroimage.usc.edu/forums/t/cat12-5-tissue-segmentation/19129/15?u=edoudela12 by Francois.

Edit; yes the map that are outputted by SPM are names p2 for GM, and p1 for WM but it is then relabelled so 1 is GM and 2 is WM :)

[AlexBo924]

Hi Edouard,

I understand but according to Francois, and also when I re-checked the individual ".nii" file.

P1.nii is the Gray matter P2.nii is the White matter

In other words the SPM/CAT output file are name p2 for WM and p1 for GM

If you look at them or if you look at the picture taken from Francois post you see that P1.nii is the GM and P2.nii is the White Matter

That is what is confusing me

[AlexBo924]

In other words, I don't know the error comes from Francois code or from the order in nirstorm process

But when I take a close look at the individual .nii file from CAT12 segmentation

We can see that P1.nii is indeed the GM and P2.nii (as well as P0*.nii) is indeed the WM

I think (but I can be wrong) it is mislabed in nirstorm.

I understand that afterwords, you need to relabel them, but the options (in the begening seems to be inverted for WM and GM).

I don't know that why I wanted to see with you what you though

[Edouard2laire]

Here is the segmentation I got last time on this MRI. WM is well identified as 1. The file name doesn't correspond to the label that is given in brainstorm. My thought is that SPM/CAT uses (1:GM, 2:WM,...) but brainstorm uses (1:WM,2:GM,...) and nistorm uses (5:GM,4:WM,3:CSF,2:skull, 1:scalp)

I can contact Francois if you want

[AlexBo924]

Thank you Edouard for the fast reply, it did confuse me a bit but what you said makes some sense.

One reason why I got worried is that when I processed CAT12 on the ICBM and compared it to yours I see that for the skin, scalp and even the CSF we have the same, but for the WM and GM I can't seem to have it like yours, I played with the colormaps but I cant seem to understand why it does not "color" identify the WM and GM like in your post.

Did you process it directly in brainstorm (CAT12 segmentation) or did you segment it outside of brainstorm using CAT12 and re-import it (I tried both and it gives me the same)

[Edouard2laire]

yes, I don't understand; your figure is weird. Could you post the question on the GitHub of brainstorm. It seems I cannot tag Francois here; he might be able to help

Did you process it directly in brainstorm (CAT12 segmentation) or did you segment it outside of brainstorm using CAT12 and re-import it (I tried both and it gives me the same)

yes I just called cat12 from brainstorm

[AlexBo924]

I'll post it and see what he thinks,

For the figure, honestly I don't know I thought in the beginning it was related to my earlier question.

Thanks again Edouard, I'll look at it a bit more see what I can find.