Francois (or others),
Is it possible to re-reference an EEG time series to the laplacian in Brainstorm?
- Brian
Drexel U., Kounios CRLab
Francois (or others),
Is it possible to re-reference an EEG time series to the laplacian in Brainstorm?
Hi Brian,
No, there is no way to easily change the reference of the EEG in Brainstorm so far. The only possible operation is to convert to the average reference.
The development of tools for handling better the EEG recordings is planned for later this year. This will include the possibility to define custom montages/references.
For now, you have to use your acquisition software for re-referencing the recordings, or do it manually in Matlab (possible with the export/import features of Brainstorm, but requires some manipulation and coding).
Cheers,
Francois
Indeed, if you have your own Matlab script to compute the Laplacian transform, another possibility would be that you copy the data file in the database (right click>File>copy), export its contents to Matlab (right click > Export to Matlab), run the Laplacian transform on Data.F (which contains the EEG or MEG data traces) and save it back into the new file (right-click > Import form Matlab). You can also develop your own plug-in for as a Brainstorm process. Let us know what you think and we could help a bit further in either direction, if necessary.
Cheers,
Great, I’ve been working through the Brainstorm tutorials and now have a handle on how to do that. The MANSCAN importer is working very well gathering all the event data.
Coherence and Granger causality are on the list for what’s next in Brainstorm. Do you have a general estimate of when those features might become available, for planning purposes? We’re moving to currently running a project using these methods, and would like to use Brainstorm if possible.
Thanks for all your work and quick responses.
Most of the connectivity tools should be finalized and documented over the summer, and ready for september/october (approximately).
Hey,
I would like to know if there is a new solution to re-reference EEG to laplacian, please.
Have a nice day,
Emmanuelle
No, no new developments in this direction, still on our to-do list, sorry.
I guess the idea is to use create automatically a montage (https://neuroimage.usc.edu/brainstorm/Tutorials/MontageEditor) that computes weights in similar manner to function ft_scalpcurrentdensity in FieldTrip (http://www.fieldtriptoolbox.org/reference/ft_scalpcurrentdensity)
If you are interested in help us with this task, you are more than welcome to contribute to the project. Let me me know if you are interested, I would give you additional pointers on where to look for in the code.
Hi!
Is there any new development regarding a feature like this?
Thank you all!
Leonardo
Yes, scalp current density has been implemented as Montage for EEG:
https://neuroimage.usc.edu/brainstorm/Tutorials/MontageEditor#Scalp_current_density
A wrapper to the FieldTrip function is also available as a process: Standardize > FieldTrip: ft_scalpcurrentdensity
Hello!
I am currently working with the ft_scalpcurrentdensity function in FieldTrip and encountered an error while attempting to use the Standardize > FieldTrip:ft_scalpcurrentdensity
option. The error message I received is as follows:
"Line 541: ft_checkdata
This function requires raw data as input, see ft_datatype_raw."
The input files are EEG epoch files processed entirely in BRAINSTORM .
When I stop the code at Line 541, I observe that the variable data
is structured as a struct containing three fields: trial
, label
, and elec
. The trial field contains a 1x1 cell containing an array of doubles, label holds electrode names, and elec has channel file information. This is the structure used by the ft_datatype_raw
as provided in the fieldtrip github page.
Could you please assist me in resolving this issue? Any guidance on how to properly format the input data for the ft_scalpcurrentdensity
function would be greatly appreciated.
If this is a fieldtrip error, please refer to their manual. It looks like you are trying to feed epoched data to a function that expects raw, ongoing recordings as input though.