Does fslorient resample the image to have the data on disk reoriented? If not, as Bruce and Eugenio commented, it is reading the same data with same orientation.

I think you can use ‘mri_diff --notallow-geo f1 f2’ to check if f1 and f2 have the same data orientation on disk.

You can use ‘mri_vol2vol --mov <movimage> --targ <targimage> --regheader --o <outimage> --interp nearest’ to reorient the data.

Best,

Yujing

From: freesurfer-bounces@nmr.mgh.harvard.edu <freesurfer-bounces@nmr.mgh.harvard.edu> On Behalf Of Fotiadis, Panagiotis
Sent: Tuesday, January 23, 2024 12:11 PM
To: Freesurfer support list <freesurfer@nmr.mgh.harvard.edu>
Subject: Re: [Freesurfer] MRIread and RAS orientation

External Email - Use Caution

Thanks, Bruce! Just to reiterate my previous message below, would you happen to know why the MRI.vol matrices of both files come out looking identical?

-----------

… I took a nifti file that was in RAS orientation (file1), flipped it into LAS (using fslorient -swaporient file2), and then opened both files in MATLAB with MRIread. The resulting .vol matrices for both files were, however, identical (i.e., voxel at location <x,y,z> in file1.vol had the exact same value as voxel in location <x,y,z> in file2.vol). Would you happen to know if the indexing of the vol structure is expected to come out the same with MRIread, regardless of the orientation?

-----------

Thank you all again!

Best,

Panos

From: freesurfer-bounces@nmr.mgh.harvard.edu <freesurfer-bounces@nmr.mgh.harvard.edu> on behalf of Fischl, Bruce R.,PHD <BFISCHL@mgh.harvard.edu>
Date: Tuesday, January 23, 2024 at 11:06 AM
To: Freesurfer support list <freesurfer@nmr.mgh.harvard.edu>
Subject: Re: [Freesurfer] MRIread and RAS orientation

I agree with Eugenio – it doesn’t do any resampling. Just reads in the data as is

Cheers

Bruce

From: freesurfer-bounces@nmr.mgh.harvard.edu <freesurfer-bounces@nmr.mgh.harvard.edu> On Behalf Of Huang, Yujing
Sent: Tuesday, January 23, 2024 8:49 AM
To: Freesurfer support list <freesurfer@nmr.mgh.harvard.edu>
Subject: Re: [Freesurfer] MRIread and RAS orientation

I’m not sure about the MATLAB codes. When the C codes read in images, no automatically transformation is done.

I think you can verify it by checking several [c, r, s] locations of both images loaded with MRIread(). I think c and r are reversed in MATLAB. And the index starts from 1.

Best,

Yujing

From: freesurfer-bounces@nmr.mgh.harvard.edu <freesurfer-bounces@nmr.mgh.harvard.edu> On Behalf Of Fotiadis, Panagiotis
Sent: Tuesday, January 23, 2024 1:54 AM
To: freesurfer@nmr.mgh.harvard.edu
Subject: [Freesurfer] MRIread and RAS orientation

External Email - Use Caution

Hello,

I have two nifti files that are both in the same MNI space, the former of which is saved in LAS orientation while the other in RAS orientation (as showcased by the “Orientation” field of mri_info). I just wanted to verify: When loading both on MATLAB using MRIread, both of their volumes are automatically loaded (or transformed in the case of the first file) in the same RAS orientation, right (using the vox2ras transformation)?

Thanks in advance!

Best,

Panos

**********************************************************
Electronic Mail is not secure, may not be read every day, and should not be used for urgent or sensitive issues