Yokogawa coordinate system not implemented in MNE?

💬 Support & Discussions
1

MNE Version: 3.8.10
operating system: Windows 10

Hello all,
we are working with MEG data recorded by a Yokogawa system. I wonder if the Yokogawa coordinate system is not implemented in MNE?
Using Maxwell_Filter an ValueError message appears saying Only 3 head digitization points of the specified kinds (“eeg”, “extra”,), at least 4 required
It seems that MNE does not recognize the head model that is needed for processing? Does anyone have experience with this?
Thanks a lot!

1 Like
2

that error message says that “we need extra points (besides just the cardinal landmarks nasion, LPA, and RPA) in order to accurately estimate head position, but we’re not finding those digitized points in the file”. Are you able to share the file so we can check whether the information is truly missing (indicating a problem with the file) or if the info is there but MNE is not finding it (indicating a problem with MNE)?

cc @larsoner who is more familiar with KIT / Yokogawa than I am.

3

This is when you use origin='auto'. I suggest using mne.bem.make_sphere_model with some suitable choices of r0 and radius and plotting these with plot_alignment, and checking visually the result based on how you acquired the data. If it looks correct, the r0 is the origin you want for maxwell_filter

4

Hi @drammock
thank you for your input!
Unfortunately, I am not able to share the entire file with you, as we are dealing with patient data.
However, I have created a link that contains the position data from one patient. The .mrk-file is the marker file measured right before the actual MEG-measurement and the .sfp-files contain the zebris-position data.
Does this help you already?

Thank you so much!
Hannah

5

Hi @larsoner

Thank you so much! What do you think would be suitable choices of R0 to try?

6

Usually something close to (0., 0., 0.04) works but I’d let plot_alignment visually guide the decision