EEG whole brain analysis

Discussions
, ,
1

Dear community,

I am interested in the cortico-subcortical connections.

If I wanted to do source analysis in an EEG, shall I do two different processing for same EEG? Meaning cortex and deep cerebral structures by separate?

Do you know if there is any software other than brainstorm that allows to do in just one analysis in case brainstorm doesn't allow?

Thank you for all the help in advance.

Best,
Rene Andrade.

2

It is possible to use mixed head models in Brainstorm, more information can be found in the deep brain activity tutorial. In the same tutorial there are warnings regarding the use these mixed models.

3

Thank you so much for your answer.

If I understand correctly there is a problem with mixed models and it is that you don't have parcellation of the cortex so you either have cortex and all subcortical or cortex detailled with another model but without the subcortical. That's what I mean I believe I need to use Volume models.

I have done MNI atlases AAL and I want to obtain functinal connectome. I have seen this debate in the forum before although I am trying with MEG data from tutorials I aim to do with EEG epilepsy data. AAL3 scouts with MEG.

I have seen this article doi:10.1093/brain/awp028. But they have done this in SPECT not EEG or MEG showing the utility of cerebellar activity in epilepsy. So I am wondering what should I do if it is possible to do with brainstorm in EEG. I have seen this article https://doi.org/10.1016/j.neuroimage.2020.116817. So I just need to know details regarding of brainstorm.

Thank you for all the help in advance.

4

Not really. The inconvenience with mixed models is that are hard to interpret, and their group-level analysis is complex (see Coregistering subjects with mixed models).
@Sylvain, some additional comments?

Regarding parcellations, as the source space is mixed (surface and volume grids, cortex and subcortical structures), it is not possible to have an atlas to contain surface and volume scouts, although this is manageable by using 2 atlases.

1 Like
5

I have seen here Coregistering subjects: group template for mixed models? this comment:

you could indeed run the two procedures (surface and volume template alignments) described in the tutorial then create a mixed surface-volume source model from both, after pruning out the cortical grid points from the volume model, too avoid redundancy with the surface model.

I don't really understand what I could possibly do in this scenario. Correct me please if I am wrong. I really appreciate the help.

I understand I have to do surface also volume template alignment and then mix them in the anatomical model making a mixed surface model pruning out the repeated ROIs.

I don't know how to do this. If you could explain I could try to do it this way.

Anyways I can try with a volume model with AAL atlas. But I don't know how to obtain functional connectivity wiht that either.

I do the the anatomical volume model, then OpenMEEG realistic head model, then source modelling (dSPM) then connectivity coherence 2023 NxN. And then I obtain error saying index must not exceed 10638 bound array.

Am I doing somehting wrong ?

What difference would there be besides simplicity betweeen using mixed model or volume model?

6

I am not sure what your study question is but I recommend you explore your data as follows:

As for connectivity, NxN might be over the top unless you use parcellation to contain N with reasonable computational limits.

7

My study for the moment is limited to BRAPH toolbox for training a DL model to use as diagnostic tool and that only requires ROIs x timeseries, source activity of the brain obtained through dSPM for example for each ROIs of a volume model from an atlas like AAL.

For functional connectivity NxN I use the scouts of my volume model.

Nonetheless, I want to know how to obtain functional connectivity NxN matrices to be able to study cortico-cerebellar connections, in terms of lobes I-X like like in SUITS atlas.

But I intend for epilepsy patients

  1. to see if cerebellum is involved in seizures?

  2. which lobe(s) specifically?

  3. in which frquency range there is a change of functional connectivity?

  4. to which clinical phase of the seizure it affects the most?

  5. it would be very interesting to know what is the role of the cerebellum in epilepsy or viceversa the role of epilepsy for the cerebellum physiopathology?

For now some study has proven with SPECT (CBF) that in post-critical phase of secondarily generalized tonic-clonic seizures (TLE) there is bilateral activity in posterior lobes of cerebellum doi:10.1093/brain/awp028 specially mostly in the vermis intimately related to the limbic role of the cerebellum https://doi.org/10.3389/fnsys.2023.1185752.

Thank you for all the help in advance.