Doubt about MRI transformation at importing anatomy, and source modeling

Good morning!
for some time now I have a question that I am embarrassed to ask, maybe because it is very basic...

Sometimes, when importing an MRI, a message appears that says:
"There is a transformation available in the MRI file..."
I have never known what I should press, nor what changes this option generates, if I should apply that tranformation or not, I have been randomly accepting or rejecting that option for years...
What is the best?

:face_with_peeking_eye: :flushed_face:

Thanks in advance!

1 Like

That is a great question (that i never dare asking). I tend to always accept but don't know why.

Replying to get notified when the answer get poster :slight_smile:

1 Like

any thoughts on this @Raymundo.Cassani @Sylvain ?? :folded_hands:

From the tutorial on epileptogenicity, it looks like the embedded transformation transforms the MRI to the RAS coordinate system such that the saggital, coronal and axial views are correctly oriented. (https://neuroimage.usc.edu/brainstorm/Tutorials/Epileptogenicity?highlight=(apply)|(transformation))

1 Like

Just to add more details on @SBeumer answer:

There are two transformations:

  1. vox2ras, this transformation that can be included in the MRI data (for MGH or NIfTI formats), for transforming from Voxel (integer units) to Right-Anterior-Superior (RAS) orientation (millimeters).

  2. tReorient: this is not explicitly included in the MRI data, but computed from the vox2ras. This second transformation is used to reorient the Voxel volume into the Brainstorm expected orientation. The idea with this is to have a MRI in voxels always oriented in the same way, regardless of the MRI device voxel orientation.

You can check the details in cs_nii2bst.m:

2 Likes

What they said + whenever you get this message and reply yes/no, it is strongly recommended that you verify the alignment between the anatomy and sensor locations: you can do so by visualizing the 3D head surface along with sensor locations.

2 Likes

thank you very much for everything! @Sylvain @Raymundo.Cassani @SBeumer @edelaire
Now I think it's clearer for me :slight_smile: