Siemens RDA file / Realignement/ Segmentation VOI / PRESS

[Sophie Betka]

Dear all,
First I want to apologize as I am sure, my questions had already been asked but I was not able to find a relevant post in the archives.


I measured glutamate (PRESS) in the right mid insula (10x15x25 mm VOI)  on a 1.5T siemens scanner.
I would like to correct for the tissue concentrations of my VOI.
But I have some questions ... your help would be wonderfullll !!

Please find enclosed, examples (from the same subject) of
-rda file;
-VOI position on T1, screenshot;
-comparaison T1/VOI using "PositionVector" coordinates;
-comparaison T1/VOI using "VOIPosition" coordinates

1) RDA file outcomes
-Do you know where I could find the definition of variables measured below (PositionVector etc..)?

PositionVector[0]: -38.331896
PositionVector[1]: -37.468117
PositionVector[2]: 27.262653
RowVector[0]: -0.998851
RowVector[1]: 0.043479
RowVector[2]: 0.020127
ColumnVector[0]: 0.006455
ColumnVector[1]: 0.538372
ColumnVector[2]: -0.842683
VOIPositionSag: -43.277742
VOIPositionCor: -33.212934
VOIPositionTra: 21.043168
VOIThickness: 25.000000
VOIPhaseFOV: 15.000000
VOIReadoutFOV: 10.000000
VOINormalSag: 0.047475
VOINormalCor: 0.841585
VOINormalTra: 0.538034
VOIRotationInPlane: -1.594676

-In orange, it seems to be the coordinates of the centre of my VOI? Can someone confirm that?
However, when I look at these coordinates on my T1  in spm, to have something like the centre of my VOI, I have to take the negative value of z.
is it normal? Is it related to the fact that siemens files are slipped on x-axis?? (i have always been soo rubbish at geometry, and I thought I will be fine doing a phd in neuroscience! Such a mistake haha)

2) PRESS - segmentation using VBM?
So, once  I found the coordinates of each of my VOI corners, I have to create a mask of my VOI.
And then, I was thinking to use VBM to segment my VOI, in spm for example (As I am not using megapress and cant use toolboxes ..).
I would have to segment my T1, using the  inclusive  VOI mask I created.

Do you have any suggestion? advice?comment on this point?
Any toolbox  that I could use?
Any piece of code which could help?

Thank you very much for your help!!
A  slightly desperate phd student called Sophie haha
 

[Martin Wilson]

Hi Sophie,

Yes, RDA is a non-standard format so I expect there are some -ve axes,
I'm afraid trail and error is the only way.

It looks like the rda2nifti scripts might do what you need:

http://lobi.nencki.gov.pl/research/28/

Let us know how you get on...

Martin

> --
> You received this message because you are subscribed to the Google Groups
> "TARQUIN users' group" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to tarquin_users_g...@googlegroups.com.
> To post to this group, send email to tarquin_u...@googlegroups.com.
> Visit this group at https://groups.google.com/group/tarquin_users_group.
> For more options, visit https://groups.google.com/d/optout.

[samira raminfard]

Hi Sophie,

As I know, Position Vector is the spatial coordination of the uppermost- left  point of the voxel.  

[0] =X, [1]= Y and [2]= Z

VOI rotaion in plane =  this is the degrees of voxel rotation in Radians, in axial position too.

I hope this help you.

Bests

Samira, a PhD student like you ;) 

 

[Sophie Betka]

Thanks Samira for the information!!

And thanks a lot Martin!!!!
I used your script and, indeed, it worked super fine (cf enclosed screenshot enclosed P22_VOI_MASK).
I re-sliced the VOI.nii, then my mask has the same dimensions of my T1 (cf enclosed screenshot before reslice and after reslice).

And, I tried to segment my T1, using my VOI mask (using the old segment module in spm12).
But it seemed too easy to be true ....  my segmentation did not worked!
I got this error message from matlab:

Failed  'Old Segment'
NaN's cannot be converted to logicals.
In file "C:\Users\bsms7594\Documents\MATLAB\spm12\spm_preproc.m" (v4916), function "spm_preproc" at line 171.
In file "C:\Users\bsms7594\Documents\MATLAB\spm12\toolbox\OldSeg\spm_run_preproc.m" (v4873), function "spm_run_preproc" at line 20.


Then, I spotted that the datatypes from my mask and my T1 were not the same, so I exported the mask using the same datatype.
I re-sliced. When I loaded my resliced mask, filename.img; it was not NaN anymore.
I segmented the T1, using my VOI as a mask, the segmentation ran well, but without taking in account my mask!! (cf screenshot after segmentation)

Any idea whyy?
Any idea of what I did wrong?
Any idea of how converting the NaN to logicals in my first mask?

Any other way to segment the VOI?

thanks for your help :)

Sophie