Display sensors (cortex) shows this. Why is an extra bar added (circled)?
I guess your channels are not properly configured.
Right-click on your channel file > Edit channel file, make sure all the contacts are set to SEEG and each electrode was correctly assigned a different group (eg. for contacts A1,A2…A8, Group=“A”)
Indeed that fixed it. Earlier I had the electrodes assigned to one group. The reason for this is because I wanted to re-reference them to a global common average and not the default local common average. With what I am doing I can have huge amplitude signals in say 2 of 10 contacts on an electrode, and with a local common average there is too much reference contamination of remaining contacts.
Is there anyway to have the electrodes grouped but still apply a global common average?
Indeed, there is no option for global average reference in SEEG in the Brainstorm file viewer.
But I had the impression that using a global average reference was never indicated in SEEG. Options I heard about are: bipolar montages, monopolar recordings / common reference, closest white matter (not straightforward to implement), and local average reference.
For a full review of referencing issues in SEEG:
http://www.sciencedirect.com/science/article/pii/S1053811915001329
http://www.sciencedirect.com/science/article/pii/S1053811916304232
With the way the code is organized, it is not straightforward to offer both a local and a global average reference.
Do you have any clear evidence it might be an interesting tool for processing or displaying SEEG recordings?
Thanks for the papers. Lado looked commonly used reference methods, including global average. This is quite widely used with SEEG studies though they point out the problems, especially with respect to correlations.
I am intrigued by the local average reference method but for my purposes it doesn’t work that well. I’m looking at baseline shifts (no high pass filter) so the amplitudes are often quite large in 1 or 2 contacts on an electrode. If there are only 10-12 contacts on an electrode the influence in the remaining contacts is too great (2/10 of the reference signal is of high amplitude) and spurious baseline shifts are created. A global average provides a better reference of these purposes.
I can simply assign the electrode positions after re-referencing the data.
Does this improve significantly the display as compared with a single common reference?
(you can create a re-referencing montage with the montage editor: http://neuroimage.usc.edu/brainstorm/Tutorials/MontageEditor)
Could you please post two screen captures illustrating why you prefer a global average reference as opposed to a common reference?
Thanks
Francois
This shows what I mean. The first figure compares the original data (common reference - 2 white matter contacts) to the local common average. See how the baseline shifts in the lower contact number cause spurious shifts in the higher contacts numbers
The second figure shows it use the 'global' common average. Having more contacts in the reference helps avoid this problem.
Yes, in this case I could probably well use the original (common reference) data but in other cases the common reference is not very good. Referential recording work better than bipolar for my purposes and using a global common average avoid me having to hunt for a 'good reference'.
A highly individualized problem.
Thank you for the explanations.
I will work on adding a global average reference option, but I cannot give you a precise timeline for the implementation.
FYI: We improved significantly the tools available in Brainstorm for processing and visualizing SEEG and ECOG data, including new options for volume coregistration. They are now documented in a new tutorial:
http://neuroimage.usc.edu/brainstorm/Tutorials/Epileptogenicity
(Global average reference is still on the to-do list)
Thank you - these are extremely nice new features.
Hello,
I changed the logic for the average reference: now the menu “Average reference” always uses a global average reference, and I added a new “local average ref” option in the “SEEG” submenu.
Francois