DICS, complex filter coeffcients and V irtual Electrodes

[Michael Wibral]

Dear Jan-Mathijs, dear Vladimir,

thanks for your quick replies. I indeed overlooked the cfg.realfilter option. Things seem to work fine now.

I also have some comments on the optimization of DICS for a certain frequency: I think that's just the same in LCMV/SAM, where you also specify the band and time you're looking at and where you shouldn't project unfiltered raw data that contain other bands than the ones used for the computation of the filters. In my opinion you choose the frequency limits explicitly (via preprocessing) in LCMV and implicitly (via the spectral smoothing in the computation of the CSD matrix and the subsequent choice of a centre frequency for beamforming) in DICS. This, of course, also means that you should not compute narrow band beamformer filters (say for higher gamma frequencies) and project unfiltered (broadband) data through your filter. What I wanted to do was to compute narrow band sources (tpsmofrq = 5 or 10, center ferquency of 80Hz), prefilter the rawdata in the appropriate range and then project and interpret them. Should be ok, shouldn't it?

Michael

> -----Ursprüngliche Nachricht-----
> Von: "Vladimir Litvak" <v.litvak at ION.UCL.AC.UK>
> Gesendet: 23.02.09 13:45:11
> An: FIELDTRIP at NIC.SURFNET.NL
> Betreff: Re: [FIELDTRIP] DICS, complex filter coeffcients and Virtual Electrodes


> Hi Michael,
> 
> I'm doing this kind of things and I use LCMV beamformer (where the
> locations of the sources might come from your DICS analysis). This was
> Robert's suggestion. The reason is that DICS is only optimized for the
> particular frequency you are looking at and if you want to extract
> source data that contains other frequencies you need to look at the
> full frequency range. There are options in lcmv (cfg.lcmv.fixedori =
> 'yes') that reduces the two orthogonal sources (in MEG case) to one
> based on the direction of maximal variance. That's something similar
> to what SAM does (although I'm not sure exactly the same).
> 
> Best,
> 
> Vladimir
> 
> 
> On Mon, Feb 23, 2009 at 11:45 AM, Michael Wibral
> <wibral at bic.uni-frankfurt.de> wrote:
> > Dear Fieldtrippers,
> >
> > I was wondering about the correct way to get SAM-like virtual electrodes using DICS. I naively tried to compute some virtual electrodes using the DICS filter coeffcients on the raw timecourses, but that obviously doesn't work because DICS filter coefficients are complex (at least in Fieldtrip 20081208 ?). Next thing I thought about was to FFT the single trial, filter it and then inverse FFT it, because the DICS filters are supposed to work in the frequency domain. But that shouldn't work either, because we have iFFT(filter(FFT(trial))) = iFFT(FFT(filter(trial))) = filter(trial), because all operation are linear.
> >
> > Next thing I am puzzled about is that the Filters (A) times the leadfield matrix (L) should be the identity matrix: AL=1 (see the Gross, 2001, PNAS paper). If A is complex - as it is in Fieldtrip -, then we would also have a complex leadfield, which seems odd to me.
> >
> > Any suggestion would be appreciated,
> > Michael
> >
> >
> > ----------------------------------
> > The aim of this list is to facilitate the discussion between users of the FieldTrip  toolbox, to share experiences and to discuss new ideas for MEG and EEG analysis. See also http://listserv.surfnet.nl/archives/fieldtrip.html and http://www.ru.nl/fcdonders/fieldtrip.
> >
> 
> ----------------------------------
> The aim of this list is to facilitate the discussion between users of the FieldTrip  toolbox, to share experiences and to discuss new ideas for MEG and EEG analysis. See also http://listserv.surfnet.nl/archives/fieldtrip.html and http://www.ru.nl/fcdonders/fieldtrip.
> 



----------------------------------
The aim of this list is to facilitate the discussion between users of the FieldTrip  toolbox, to share experiences and to discuss new ideas for MEG and EEG analysis. See also http://listserv.surfnet.nl/archives/fieldtrip.html and http://www.ru.nl/fcdonders/fieldtrip.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Michael Wibral.vcf
Type: text/x-vcard
Size: 344 bytes
Desc: not available
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20090223/b19eca0c/attachment-0001.vcf>