search

arnodelorme / mffmatlabio

MFF Matlab file importer and exporter
GNU General Public License v3.0
11 stars 15 forks source link

Importing 10-20 labels #49

Closed IsaacEGI closed 2 months ago

IsaacEGI commented 8 months ago

Dear Arnaud,

I am reporting this issue more as a feature request. We have some customers who use the DIPFIT plugin in EEGLAB. This plugin requires a co-registration of the electrodes positioning to the 10-20 layout, and as our MFF files use the E1, E2, ... labels, this has to be done automatically.

I was trying different ways to get to use that plugin easier and I tried to import the MFF file to EEGLAB after a montage modification using our Net Station Tools software. This tool, besides being used to re-reference also allows to change the layout to the 10-20 for example. I thought that when importing the MFF in EEGLAB after this modification, the labels would be taken as the 10-20 and not the E1 format.

Is there any way that the mffmatlabio plugin could read the 'sensorLayout.xml' sensor names and introduce them to the electrode locations of EEGLAB? This would allow them to use the automatic selection of the Co-Register function in the DIPFIT.

Thank you,

Isaac

arnodelorme commented 8 months ago

Hi Isaac,

If the channels are in the MFF format, it is not compatible with the 10-20 system (except for vertex electrode Cz). There is nothing that can be done (it is like trying to convert a bolt in imperial unit to metric unit, there is no perfect match). EEGLAB will automatically align the electrodes to a 10-20 montage when you perform source localization.

Arno

IsaacEGI commented 8 months ago

Hi Arnaud,

Thank you for your quick reply and the information shared!

I have not been able to find a way to align them automatically with EEGLAB, how is this done?

While exploring the Co-Register option in DIPFIT, I noticed an EGI option, but in the command window it mentions it is not implemented (please see the screenshot attached). Would this option potentially serve as a way for conducting automatic source localization?

Thank you for your help!

Best wishes,

[Logo Description automatically generated] Isaac Valls-Giménez Product Support Specialist Email: @.**@.> Phone: +31681420058 Connect to my LinkedInhttps://www.linkedin.com/in/isaac-valls-gim%C3%A9nez-297643234/

From: Arnaud Delorme @.> Date: Friday, November 10, 2023 at 5:31 PM To: arnodelorme/mffmatlabio @.> Cc: Isaac Valls-Giménez @.>, Author @.> Subject: Re: [arnodelorme/mffmatlabio] Importing 10-20 labels (Issue #49) This email was sent from outside of Welcony. Please do not click links or open attachments unless you recognise the source of this email and know the content is safe.

Hi Isaac,

If the channels are in the MFF format, it is not compatible with the 10-20 system (except for vertex electrode Cz). There is nothing that can be done (it is like trying to convert a bolt in imperial unit to metric unit, there is no perfect match). EEGLAB will automatically align the electrodes to a 10-20 montage when you perform source localization.

Arno

— Reply to this email directly, view it on GitHubhttps://github.com/arnodelorme/mffmatlabio/issues/49#issuecomment-1806048716, or unsubscribehttps://github.com/notifications/unsubscribe-auth/BAVT4WOMNR34XJFDCL3MXLTYDZJHTAVCNFSM6AAAAAA7GLXU5SVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQMBWGA2DQNZRGY. You are receiving this because you authored the thread.Message ID: @.***>

arnodelorme commented 8 months ago

Can you share a data file? Also feel free to send me direct emails and CC trey.

arnodelorme commented 8 months ago

Closing this report for now.

arnodelorme commented 8 months ago

There is no magic as there is no correspondence between EGI locations and 10-20 (except for Cz). It is just a different system. However, when I imported the files Isaac shared, I have noticed that EEGLAB did not automatically find the transformation to align with the 3-D head model (itself aligned with 10-20). I will try to fix that.