[FieldTrip] eLORETA spatial filter issues

Schoffelen, J.M. (Jan Mathijs) jan.schoffelen at donders.ru.nl
Fri Feb 28 13:20:21 CET 2020

Hi Soren,

Just to clarify, if with ‘data dependent’, you mean ‘electrophysiological data dependent’. The reason I mention this, is that clearly one should consider using subject-specific headmodels and leadfields if possible (and with the HCP MEG data this is definitely possible), which will result in different spatial filters for different subjects.

But indeed, it depends on the inverse algorithm whether or not the spatial filter is a function of the leadfields only (e.g. eLORETA), or whether some aspect of the data is taken into account (MNE: noise covariance (optional), beamformers: data covariance).


> My two questions are: 
> 1) Should the filter really be independent of the data like this? Is this just a difference between the two methods (eLORETA and MNE)? 

Yes, and Yes.

> 2) If the independence of the filter is due to a difference between the methods, is eLORETA suitable for the extraction of virtual channels in the same way as described in the tutorial linked above? 

Well, the fact that the eLORETA spatial filter is merely a function of the leadfield (+some additional spatial smoothness constraints), this does not invalidate it to be used as a spatial filter, although I’d rather use a technique (beamformer, or if I have to: MNE) which is a bit more to the spatial point.

Best wishes,

> Any help would be very much appreciated. 
> Best and thanks very much in advance, 
> Soren
> _______________________________________________
> fieldtrip mailing list
> https://mailman.science.ru.nl/mailman/listinfo/fieldtrip
> https://doi.org/10.1371/journal.pcbi.1002202

More information about the fieldtrip mailing list