How to Map MSMALL Surface Vertex in MNI space to T1w Acquisition Space Coordinates

[李佳艺]

Dear all,

Hello, I have a question about surface coordinate mapping in the HCP dataset.

We have identified a vertex of interest in an individual’s
MNINonLinear/fsaverage/_midthickness_MSMALL_32k_fs_LR.surf.gii
surface file. We would like to convert the coordinates of this vertex into the original T1-weighted MRI acquisition space for TMS targeting.

Based on the reference papers, I understand that this might be possible by applying the inverse MNI nonlinear transformation to map the vertex from MNINonLinear/fsaverage/_midthickness_MSMALL_32k_fs_LR.surf.gii to T1w/fsaverage/_midthickness_MSMALL_32k_fs_LR.surf.gii, and then using linear transformations (such as translation and rotation) to obtain the coordinates in the original acquisition space.

I believe the coordinates in T1w/fsaverage/_midthickness_MSMALL_32k_fs_LR.surf.gii are in the subject’s native space, which is derived from the original acquisition space via nonlinear transformations. If this is incorrect, please let me know.

Is this the correct approach? If you have any suggested tools or scripts for these transformations, I would greatly appreciate your advice.

Thank you very much!

Best regards,
Jiayi Li

[Tim Coalson]

T1w/ space is rigidly aligned to make an approximate ACPC space (after correcting for the distortions the scanner is know to produce - for anatomical images, they are minor, and for the level of accuracy TMS can utilize, you could just ignore them - I don't know whether to expect navigator software to know about these distortions, so ignoring them could actually be more accurate, depending).  You can invert the rigid registration the pipelines did to align a raw image to T1w space (T1w/xfms/acpc.mat if you have only one T1w, if you have multiple it gets a little more involved), apply that to the surfaces with wb_command -surface-apply-affine command (the affines are flirt convention, so you will need to specify the T1w and raw scanner volume files to the -flirt option).  You can (and probably should) start with the T1w/ version of the surface instead of applying the reverse of the MNINonLinear template registration (note also that applying a warpfield to a surface uses the inverse of what you might expect, because warpfield inversion is not simple and volume resampling actually works "backwards").

Yes, if you identified a vertex on a 32k MSMAll surface, you should use the MSMAll version of the surface in whatever other space you are interested in.

Tim

--
You received this message because you are subscribed to the Google Groups "HCP-Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hcp-users+...@humanconnectome.org.
To view this discussion visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/38dfbbd2-f9d5-4ab3-81d9-7dccb6b4dcc3n%40humanconnectome.org.

[Glasser, Matthew]

This is becoming a more common request as the HCP Pipelines are being used in clinical-translational work.  We will work to make this easier for users.

Matt.

To view this discussion visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/CANm3QU4AASLEMBUtsOKLxzj_OO2dxKLWtyWVx9R%2BEu5Zz%2Ba9Sg%40mail.gmail.com.

 

---

The materials in this message are private and may contain Protected Healthcare Information or other information of a sensitive nature. If you are not the intended recipient, be advised that any unauthorized use, disclosure, copying or the taking of any action in reliance on the contents of this information is strictly prohibited. If you have received this email in error, please immediately notify the sender via telephone or return mail.