Understanding the "-trans.fif" file

In this tutorial, it seems I have to have a “-trans.fif” file to be able to do the computations for source-localized data. However, I cannot find such a file in my freesurfer results. Shall this file be obtained through MRI? I also just have “.con” files as my MEG files.

I will put screenshot of my MRI directory:

Screen Shot 2021-05-29 at 17.13.14
Screen Shot 2021-05-29 at 17.13.36

Which of these files should I use for -trans file? If it is not listed here, how should I obtain it?

Thanks in advance

  • MNE-Python version: 0.23.0
  • operating system: MacOS

My MEG files are in “.con” format…
I could not use the “read_trans” to read these files…

How should I solve the problem?
Screen Shot 2021-05-29 at 18.41.31

Hello Shirin,

the “trans.fif”-file contains a matrix for the transformation between the the “head”-coordinate-frame from your measurement (defined by the digitization of the anatomical landmarks of the participant before the measurement) and the “MRI”-coordinate-frame of the 3D-Brain from the Freesurfer-Reconstruction (your MRI-Directory you showed).
To acquire the “trans.fif”-file you need to align both coordinate-frames manually which is called coregistration. MNE-Python provides a GUI for that.
Here, you can find a tutorial about the coordinate-frames and the usage of the Coregistration-GUI.

Regarding your other question about your “.con”-MEG-files:
You only need mne.read_trans() to read the “.trans”-file you acquired using the Coregistration-GUI from above. The “.con”-files are your Raw-Data and you can read them using mne.io.read_raw_kit().


Thank you very much Martin!!!

If my understanding is correct, the “-trans.fif” file is only based on MRI data. Therefore, I need to calculate it one time even if I have multiple MEG session data.
Is that right?

And is that true also for BEM surfaces?

yes it’s one trans file per acquisition session not per MEG run.