how to convert a label/metric from freesurfer space to a native 32k space?

[lei wei]

Dear HCP expert

I followed the HCP pipeline and then used the ciftify tools to process my T1w images. I tried to extract diffusion metrics from the low-resolution 32k native space.

I found that the Ciftify pipeline resamples the original ~164k native surface to the 32k native surface by simply using the 'wb_command -surface-resample' with a standard 32k sphere. I followed the above command but replaced 'surface-resample' with 'label-resample' and tried to resample my native high-resolution atlas to the 32k native space.

I used the native 164k atlas and the rh.sphere as input, and the standard 32k sphere in the zz_template that created by Ciftify as target sphere.

 To visualize the file in wb_view, I also convert the atlas to dlabel.nii format. However, the output results look weird，I attached the photo below. How can I resample a native surface annotation to a low-resolution 32k space?

Really appreciate for any help !

Thanks !

[Glasser, Matthew]

Use the 32k and 164k FS_LR spheres.

 

Matt.

 

Really appreciate for any help !

 

Thanks !

 

--
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 on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/9eecd7c9-2ea6-4c6c-bcf3-7762ab80426fn%40humanconnectome.org.

 

---

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.

[Glasser, Matthew]

Actually, you need to use the FS_LR registered native sphere and the 32k FS_LR sphere.  Perhaps you can paste in what spheres CIFTIFY makes available to you as I don’t recall those details now.

Matt.

To view this discussion on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/E676C6EA-0986-4D3B-841B-FF726AD889A4%40wustl.edu.

[lei wei]

Thank you for your reply! 

I tried using the sphere file with MSMSulc.native.surf.gii as the source sphere and sphere.32k_fs_LR.surf.gii as the target sphere. 

The problem seems to be solved, and the transformed atlas looks good. 

If all surfaces were resampled to the 32k surface using a standard 32k template sphere, does it mean all the native 32k surfaces have the same correspondence to the vertices in the atlas space? 

Or can I directly use the atlas in the standard fs_LR32 space on these native 32k surfaces? 

 Thanks!  

[Glasser, Matthew]

That should be fine. That is the MSMSulc registration to FS_LR, which is the one we currently recommend if you have only folding available.

 

I don’t know what you mean by native 32k.  The native mesh is the original FreeSurfer mesh.  The 32k mesh is always FS_LR.  Any 32k mesh is interchangeable with any other assuming they have the same topology.

 

Really appreciate for any help !

 

Thanks !

 

--
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 on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/9eecd7c9-2ea6-4c6c-bcf3-7762ab80426fn%40humanconnectome.org.

 

 

---

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.

--
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 on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/E676C6EA-0986-4D3B-841B-FF726AD889A4%40wustl.edu.

 

---

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.

--
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 on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/f24c424c-0c0e-4b2d-8b60-a5f5486436c8n%40humanconnectome.org.

[lei wei]

Thanks for your help ! 

I understand that HCP-style data includes two spaces: MNI and T1w. Additionally, the surface data comes in two resolutions: native FreeSurfer (~164k) and 32k. The file I used is in T1w space with 32k vertices. If I understand correctly, the vertices on this surface are "warped" from the MNI 32k space. Therefore, the vertices on the T1w 32k surface correspond to the same vertices on the MNI 32k surface.

I tried to compute some dMRI metrics SNR across 100 subjects. I projected the dMRI data onto the 32k surface in T1w space, concatenated the data for 100 subjects and 32k vertices into a matrix [100 x 32492], and computed the group mean and standard deviation to estimate the SNR. Is this approach correct ? 

The native 32k surface was generated using the '--resample-to-T1w32k' option in ciftify.  According to the documentation, this option can be used to output an additional T1w/fsaverage_LR32k folder as seen in HCP. In the HCP pipelines, I found it seems implemented by resample the warpfield from native to Freesurfer standard space  'sphere.reg'  to the 32k space, does it correct ?  I apologize if my question looks stupid.

Thanks again for your reply

[Glasser, Matthew]

We would say that the 32k mesh in the T1w folder is in the subject’s physical space.  What you did sounds fine.

Matt.

Error! Filename not specified.

 

Really appreciate for any help !

 

Thanks !

 

--
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 on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/9eecd7c9-2ea6-4c6c-bcf3-7762ab80426fn%40humanconnectome.org.

 

 

---

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.

--
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 on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/E676C6EA-0986-4D3B-841B-FF726AD889A4%40wustl.edu.

 

---

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.

--
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 on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/f24c424c-0c0e-4b2d-8b60-a5f5486436c8n%40humanconnectome.org.

 

---

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.

[Tim Coalson]

There is more than one "fs_LR" mesh, and 164k is one of the standard ones (and full-resolution fsaverage is also 164k - technically, both of these are 163,842 vertices, but we shorthand them as 164k).  The native mesh is different per subject, but can be 150k vertices or similar.  Saying specifically "~164k" is confusing if you aren't talking about the 164k fsaverage or 164k fs_LR spheres.

The surfaces for any subject are initially generated with an arbitrary mesh, typically in a rigid-aligned space (not MNI).  They are then surface-registered to some surface template (fsaverage or fs_LR) to make resampled versions on standard meshes, and versions of the surfaces are often also generated with an MNI volume registration applied to them. 

Tim

To view this discussion on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/1e214fe5-3b85-4670-822c-c509d194a3aan%40humanconnectome.org.

[lei wei]

Thanks for your correction !

Tim Coalson <tim.c...@gmail.com> 于2024年7月3日周三 04:13写道：

To view this discussion on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/CANm3QU6d%3DDeBLboxULiv%3DGZKtA5iOhQcVL8AQ2R3zrKJn_Qhvw%40mail.gmail.com.