Hi all,
we are working with MEG-data, recorded from a Yokogawa system, and would like to run tSSS via brainstorm. However, since the ‘maxwell_filter (tSSS)’ process in brainstorm is designed for/ requires Neuromag data (FIFF-format ), it doesn´t work with the Yokogawa .con format file.
I wonder if anyone has had the same problem and already tried to clean data from a KIT / Yokogawa / Ricoh system using tSSS?
Thanks for your help!
What error do you get?
I think Brainstorm does not impose any restriction on the acquisition device to allow running the SSS/tSSS process.
If you have specific questions about this process, you should probably ask them to the MNE-python community instead.
Your Python and/or MNE-Python environment is not properly set up.
See this page: https://neuroimage.usc.edu/brainstorm/MnePython
Hi Francois,
You were right! I have now set up my Python Environment correctly and tried the process again. Unfortunately it still doesn´t work properly. Now I get the following error report:
Error: [process_mne_maxwell] Pre-process > MNE-Python: maxwell_filter (SSS/tSSS)
** Line 133: _decomp_svd>svd
** Python Error: ValueError: illegal value in 4th argument of internal gesdd
**
** Call stack:
** >_decomp_svd.m>svd at 133
** >fixes.m>_safe_svd at 84
** >maxwell.m>_regularize_in at 1947
** >maxwell.m>_regularize at 1066
** ><decorator-gen-460>.m>_regularize at 12
** >maxwell.m>_get_decomp at 1012
** >maxwell.m>_run_maxwell_filter at 555
** >maxwell.m>maxwell_filter at 353
** ><decorator-gen-457>.m>maxwell_filter at 12
** >process_mne_maxwell_py.m at 24
** >process_mne_maxwell.m>Run at 134
** >process_mne_maxwell.m at 27
** >bst_process.m>Run at 203
** >bst_process.m at 38
** >panel_process1.m>RunProcess at 124
** >panel_process1.m at 26
** >gui_brainstorm.m>CreateWindow/ProcessRun_Callback at 776
** >bst_call.m at 28
** >gui_brainstorm.m>@(h,ev)bst_call(@ProcessRun_Callback) at 297
This could be either a bug in the creation of a python object to be sent to MNE-Python, or a bug in MNE-Python.
Can you please share an example data file? (upload the file somewhere and post the download link here)
Please post a screen capture of the options you used for the Maxwell filter process as well.
Thank you for the example dataset.
I could reproduce the error, but I don't know if the problem is coming from Brainstorm or MNE.
I posted the question on the MNE-Python user forum, together with the first 10s of the recordings you shared (the file contains only the signals and sensor positions, no metadata at all - fully anonymized):
https://mne.discourse.group/t/calling-maxwell-filter-on-yokogawa-recordings-from-brainstorm/6452
Thank your for your fast answers and your support!
We are still investigating the problem with the MNE-Python group.
I've just opened a GitHub issue for that: https://github.com/mne-tools/mne-python/issues/11548
They would need the link to download the example dataset that you shared with me.
You can maybe sent it as a private message to @Eric89GXL from this forum, or wait until they contact you by email.
The problem is now solved with this commit: https://github.com/brainstorm-tools/brainstorm3/commit/93f86ffa5847505c1454ae38c98a100f1a2baaa8
The problem was caused by three channels without names and positions in the Yokogawa file. There interpretation in Brainstorm data structures was incorrect, causing this problem in MNE-Python. I modified the Brainstorm reader for Yokogawa files to ignore completely these three channels, and it seems to work.
Update Brainstorm, delete the "Link to raw file", link the file again to the database, and try again.