On 3/21/2021 3:33 PM, Adam Li wrote:

        External Email - Use Caution        

Hi,

I am a bit noob on FreeSurfer and coordinate systems and I had a few questions. I am wondering if I have the T1.mgz file of FreeSurfer after running recon-all and this is the 256^3 image re-scaled and re-oriented from your orig.mgz file, then how does one define the fsnative (volume RAS) vs surface RAS (tkras in FreeSurfer)? Assuming there was no preprocessing done on the original scanner nifti, I run the following commands:

(base) workstation$ mri_info ~/bids/derivatives/freesurfer/la02/mri/T1.mgz --vox2ras
  -1.00000    0.00000    0.00000  126.81799
   0.00000   -0.00000    1.00000 -121.57332
   0.00000   -1.00000    0.00000  159.64563
   0.00000    0.00000    0.00000    1.00000 

(base) workstation$ mri_info ~/bids/derivatives/freesurfer/la02/mri/T1.mgz --vox2ras-tkr
  -1.00000    0.00000    0.00000  128.00000
   0.00000    0.00000    1.00000 -128.00000
   0.00000   -1.00000    0.00000  128.00000
   0.00000    0.00000    0.00000    1.00000

Questions:

1. Applying the `vox2ras` to any voxel, I would get RAS xyz coordinates right?

Yes, you get Scanner RAS

2. If I apply `vox2ras-tkr` to any voxel, I would also get RAS xyz coordinates?

You would get "TKR" coordinates. TKR stands for tkregister, one of our old registration GUIs.

3. So what is the difference exactly between these two transforms? Is there anything special / consistent in terms of the definition of the origin and orientation? E.g. Orientation is always RAS, and origin is always... estimated center of the patient's brain? Or is the origin somewhere special like the AC?

The surface is in TKR coords (eg, if you take a surface vertex  XYZ and apply the inverse vox2ras-tkr, you will get the voxel col, row, slice where the vertex lives). The center is at the center of the volume (almost).

3. When would I use one transform over the other?

I've searched online and read through the following, but I feel like there is not explicit discussion of the above questions:

- MailScanner has detected a possible fraud attempt from "secure-web.cisco.com" claiming to be https://surfer.nmr.mgh.harvard.edu/fswiki/CoordinateSystems

- MailScanner has detected a possible fraud attempt from "secure-web.cisco.com" claiming to be http://www.grahamwideman.com/gw/brain/fs/coords/fscoords.htm


--
Best Regards,

Adam Li (he/him)
PhD Student in Biomedical Engineering 
Johns Hopkins University & School of Medicine

_______________________________________________
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer