[Eeglablist] Combine EEG and MRI data for identification of channels of interest/connectivity analysis

[Makoto Miyakoshi]

Dear Fabio,

> - is it technically possible to use MRI scans that have already been
pre-processed into MNI space?

Yes, why not. If necessary, you can back-transform to individual space. SPM
generates these forward-backward transform matrix.

> - if so, what is the best way to achieve the EEG-MRI coupling?

If you mean channel locationis coregistration, you still need to 'digitize'
i.e., measure spatial locations of the scalp electrode using a special
devise; see also my colleague's recent work to do it using a ipad with a
lazer scanner, https://sccn.ucsd.edu/wiki/Get_chanlocs

> - does all of the above seem reasonable, or given my situation (find the
electrodes above the area of interest and in parallel start do connectivity
analysis in pathological recordings) it would be better to use another
approach?

Unfortunately I have rather a pessimistic opinion on EEG's spatial
locations even with these inputs for the following reasons.

   1. Head's electric forward model is still very simple. For example,
   standard models do not have blood vessels yet (See Fiedeler et al., 2016
   NeuroImage)
   2. Skull's electric conductivity has unknown individual differences, and
   it could vary up to three times (See Akalin Acar et al., 2015 or around...
   so called 'SCALE' paper)
   3. The results heavily depends on what algorithm you use.

After all, EEG is not MRI, in the sense that with MRI you can perform MR
microscopy, but even with 10,000 EEG channels you cannot perform EEG
microscopy. You should think about this fundamental difference carefully if
you haven't done so.

Makoto


On Thu, Jul 19, 2018 at 3:25 AM Giatsidis, Fabio <fabio_giatsidis at brown.edu>
wrote:

> Hello EEGLAB list,
>
> I have been using EEGLAB in the past few months to do some spectral
> analysis on pathological EEGs, but I now need to couple these with the MRIs
> of the respective patients, in order to both pinpoint/identify only the
> electrodes above the areas of interest and start performing some
> connectivity analysis.
> Additional notes:
> 1) the EEGs were recorded using EGI 128-channel HydroCel caps, and are
> resting states only,
> 2) the MRI scans were already converted to MNI space (and lesions marked)
> before the EEGs were recorded,
> 3) I cannot go back to the original images and redo the above due to
> technical and time limitations.
>
> I have read a bit about NFT, Fieldtrip and SPM12, but all of these seem to
> assume that one has to start from the raw images (which I currently cannot
> do, see point 3).
>
> My questions would then be:
> - is it technically possible to use MRI scans that have already been
> pre-processed into MNI space?
> - if so, what is the best way to achieve the EEG-MRI coupling?
> - does all of the above seem reasonable, or given my situation (find the
> electrodes above the area of interest and in parallel start do connectivity
> analysis in pathological recordings) it would be better to use another
> approach?
>
> Thank you very much for all your help!
> Best,
> -Fabio
>
> --------------------------------
> *Fabio Giatsidis, M.D.*
> Resident in Neurology - University of Rome "Tor Vergata" - Rome, Italy
> Post-doctoral research fellow - Brown University - Providence, RI, USA
> _______________________________________________
> Eeglablist page: http://sccn.ucsd.edu/eeglab/eeglabmail.html
> To unsubscribe, send an empty email to
> eeglablist-unsubscribe at sccn.ucsd.edu
> For digest mode, send an email with the subject "set digest mime" to
> eeglablist-request at sccn.ucsd.edu



-- 
Makoto Miyakoshi
Swartz Center for Computational Neuroscience
Institute for Neural Computation, University of California San Diego
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20180725/f75d0899/attachment.html>