My individual anatomy in MNI coordinates

Discussions
1

Dear all,
I’m loading the anatomy folder (done with brainsuite) inside my protocol. Than I need to put it in MNI coordinates, could you tell me the most precise way to do it?
Moreover I find hard to follow how to apply the following script for translate single MRI coordinates in MNI coordinates; could you help me with same example?

MRI (meters) to MNI (meters)

MRI (meters) => MRI (millimteres) => MNI (millimeters) => MNI (meters)

P_mni = cs_mri2mni(sMri, P_mri' .* 1000)' ./ 1000;

Thank you very much
Luca

2

Hi Luca,

If you are using the individual MRI of your subjects, you are not going to have access to accurate “MNI coordinates”.
This term refers to the MRI coordinates transformed in the MNI stereotaxic space using the MINC tools.
This transformation is not available in Brainstorm yet, and will not be in the near future.

To identify the anatomical regions on the cortex, it will be more precise to use the anatomical atlases that are available in the scouts tab (one atlas available with BrainSuite, five with FreeSurfer).

Francois

3

P_mni = cs_mri2mni(sMri, P_mri’ .* 1000)’ ./ 1000;

Dear all,
I get a strange results from this code, could you help me?

  • I extracted sMri array in the workspace;
  • than I create a variable with MRI x y z (in millimeters) value to translate called P_mri.
  • than I run the following script…P_mni = cs_mri2mni(sMri, P_mri’ .* 1000)’ ./ 1000;
  • Results is an “empty” (no values inside) variable called P_mni;
    Whats wrong?

Thank you very much
Luca

4

Hi Luca,

If your coordinates are in millimeters, and if you want an output in millimeters, then the call should be:
P_mni = cs_mri2mni(sMri, P_mri’)’;
With the size of P_mri being [Npoints x 3].

There are additional constrains to use this function. As the interface tells you, there are no MNI coordinates available except for volumes normalized to the MNI space (Colin27, ICBM152).
When the AC/PC/IH points are defined and this is a normalized MRI, the fields sMri.NCS.R and sMri.NCS.T should be defined and you should see the MNI coordinates in MRI viewer.
If you don’t see the MNI coordinates when you open the MRI in the MRI viewer, this function will not work.

If you want to use manually your own functions that returns “Talairach” or “MNI” coordinates for individual brains, you can, some are available online. But I didn’t come across anything that is correct enough to make it available in a generic platform such as Brainstorm.

Francois

[Please keep on posting follow-up questions on the same thread]

5

Dear Francois,
I get some new relevant informations, thanks. I have R and T value in sMri.SCS, but not in sMri.NCS. Reason why I get any results from code probably. How I can defined sMri.NCS.R and sMri.NCS.T?
Luca

6

If this is an individual MRI, you can’t, Brainstorm doesn’t offer this option.
If this is a MNI template (colin27 or ICBM152), it will be automatically defined when the AC/PC/IH points are set.

7

thanks Francois, now I understand. In a previous post you said “If you want to use manually your own functions that returns “Talairach” or “MNI” coordinates for individual brains, you can, some are available online”, could you indicate me where I can find this functions on the web, since I don’t find out on the web this informations. Before to develope my own functions I would like to check the others results and out to get it. Thank you very much again,
Luca

8

Hi Luca,
I’m sorry, I cannot help you more with this topic. I haven’t find an easy way to do it myself, this is why it is not in Brainstorm.
If you can find a reliable method that you would like to share with other Brainstorm users, please let me know, I could help you with the final integration.
Cheers,
Francois

9

Hi Francois,
I think that develope an automatic way following what the manually alligniment with an MNI template do could resolve the problems. Obviously everything depend by the use the you need to do with the conversion. Thank you very much again,
Luca

10

Hi Luca,

If you use FreeSurfer for the MRI segmentation, one way to get around this limitation is to project your sources on the Colin27 brain:

  1. run the source estimation on the individual anatomy
  2. right-click on the source file > Project sources > Colin27 > cortex_15000V
  3. the projected source file will be on an MNI template, so you would have access to the MNI coordinates.
    http://neuroimage.usc.edu/brainstorm/Tutorials/CoregisterSubjects

Cheers,
Francois

11

FYI, there is now the possibility to compute MNI coordinates for individual subjects:
http://neuroimage.usc.edu/brainstorm/News#MNI_coordinates