[Eeglablist] Regarding Simultaneous fMRI-EEG Data
Makoto Miyakoshi
mataothefifth at yahoo.co.jp
Thu Sep 3 20:28:29 PDT 2009
Dear Gaurav,
> I have got similar 3D rendering and SPMmip images but
> your design matrix with contrast, as given in the 'pdf file' you sent
> to me, does not match when giving the contrast values here; can you
> please tell me that how you are giving the no. of volumes to use
> after stabilization of fMRI scanner and contrast parameter.
I discarded first 10 scans (=30seconds duration) to secure data quality.
There are therefore 110 scans. I found it takes while for EEG data to be
stabilized. There is actually seven seconds of preparatory scan that is not
counted as data, but it seems it is not sufficient. Contrast parameter is
alpha-power time series extracted from one of ICs that represent occipital
alpha. The following is the matlab script prototype I used. Be sure that
you launched EEGLAB and loaded the set file (epoched). I used cycles = 22
at 8-12Hz, which is approximately 3sec. it picks up the highest alpha power
between 8-12Hz by 0.1Hz step. Note also that the hrf is convolved in the
end (so you need SPM installed).
%%%%%%%%%%%%%%%%%%%%
for n=[11 13 14]; % enter IC numbers
[ersp,itc,powbase,times,freqs,erspboot,itcboot] = newtimef(...
EEG.icaact(n,:),... % data
EEG.pnts,... % frames
[EEG.xmin EEG.xmax]*1000,... % tlimits
EEG.srate,... % srate
22,... % cycles
'timesout', 330,...
'freqs', [8 12],...
'nfreqs', 41,...
'plotersp', 'off',...
'plotitc', 'off',...
'baseline', NaN);
[power freqn] = max(mean(ersp,2));
disp([num2str(freqs(freqn)) 'Hz selected'])
regressor = ersp(freqn, :);
regressor = resample(regressor, 110, length(regressor));
regressor = conv(regressor, spm_hrf(3));
regressor = regressor(:, [1:110]);
save(['IC' num2str(n)], 'regressor')
end
%%%%%%%%%%%%%%%
> I tried to analyse the eegdata in EEGLab and removed gradient and
> pulse artefacts using the given tools in EEGLab, also tried to
> localize the source using BESA electrode placement design; but I am
> not very much sure about results I am getting, because of two
> reasons, one is the all the parameters I am to give in the software
> are not available to me and I am putting them as given in manual of
> the EEGLab, and the second reason, as being new to this field,
> relating them with biological aspects and brain anatomy.
I have never used BESA plugin, so I don't know anything about it. One of
MEG persons came to me and said the dipole, which was located at left
occipito-temporal area, looks absurd. I don't doubt he is true, because my
data have only 32 channels. My understanding of significance of source
estimation is that it provides spatial information as one of constraint
conditions in interpreting the data, which is not necessarily
physiologically valid given that the current recording condition is far
from ideal. You'll need 128 channels to achieve the spatial Nyquist of a
human head (Srinivasan, 1995).
> I am studying the paper you had sent to me and I will let you know if
> I will be able to implement the idea given in paper.
Great! Please tell me if it works. I also wondered if it is possible to
create a file format converter between EEGLAB and FAST (or SPM8).
> If you also tried for fusion or integration of both simultaneously
> recorded EEG and fMRI data earlier or you know any tool available for
> this purpose, then please let me know.
I also use VisonAnalyzer that came with BrainAmp. I found it convenient to
identify ECG peak manually with it because it has intuitive interactive
interface. I hope the similar interactive interface is implemented in
EEGLAB.
Makoto
--- Gaurav Garg <Garg-G at email.ulster.ac.uk> wrote:
> Hello Makoto,
>
> I am doing well here, How is everything with you?
> Its really nice to see your mail. Thankyou very much for sharing your
> analysis results with me and I have also regenerated your results
> here using SPM. I have got similar 3D rendering and SPMmip images but
> your design matrix with contrast, as given in the 'pdf file' you sent
> to me, does not match when giving the contrast values here; can you
> please tell me that how you are giving the no. of volumes to use
> after stabilization of fMRI scanner and contrast parameter.
>
> I tried to analyse the eegdata in EEGLab and removed gradient and
> pulse artefacts using the given tools in EEGLab, also tried to
> localize the source using BESA electrode placement design; but I am
> not very much sure about results I am getting, because of two
> reasons, one is the all the parameters I am to give in the software
> are not available to me and I am putting them as given in manual of
> the EEGLab, and the second reason, as being new to this field,
> relating them with biological aspects and brain anatomy.
>
> I am studying the paper you had sent to me and I will let you know if
> I will be able to implement the idea given in paper.
> If you also tried for fusion or integration of both simultaneously
> recorded EEG and fMRI data earlier or you know any tool available for
> this purpose, then please let me know.
>
> Thanks,
>
> Best Regards
> Gaurav Garg
> =====================================
> Postgraduate Research Scholar
> Room No. MS125
> School of Computing and Intelligent Systems
> University of Ulster
> Magee Campus
> Northland Road
> Derry
> Northern Ireland BT48 7JL
>
> Tel - +442871371207
> ________________________________________
> From: Makoto Miyakoshi [mataothefifth at yahoo.co.jp]
> Sent: 28 August 2009 10:15
> To: Gaurav Garg
> Subject: RE: Regarding Simultaneous fMRI-EEG Data
>
> Hello Gaurav,
>
> How are you doing?
> I recently had chance to analyze the data I shared with you.
> The attached PDF is from one of the best results.
> I used alpha power fluctuation from haku2close as a regressor and
> asked
> negative correlation using SPM5.
> Isn't it great?
>
> About other subjects, I had difficulty in identifying qrs peaks and
> could
> not subtract BCG well. If you have reliable pqr peaks, then you can
> perform
> good BCG remove algorithm such as constraintICA (Leclercq et al.,
> 2009
> Neuroimage) but with no good EKG, I cannot help it.
>
> Makoto
>
>
>
> --- Gaurav Garg <Garg-G at email.ulster.ac.uk> wrote:
>
> > Dear Makoto,
> >
> > Thanks for this information about subjects; also it would really be
> > great idea to share your analysis results and findings to me to get
> > good understanding of the facts. I will also let you know if I will
> > find some remarkable things.
> >
> > Thanks,
> >
> > Gaurav
> >
> > ==================================================
> > Gaurav Garg
> >
> > Postgraduate Research Scholar
> > Room No. MS125
> > School of Computing and Intelligent Systems
> > University of Ulster
> > Magee Campus
> > Northland Road
> > Derry
> > Northern Ireland BT48 7JL
> >
> > Tel - +442871371207
> > ________________________________________
> > From: Makoto Miyakoshi [mataothefifth at yahoo.co.jp]
> > Sent: 17 July 2009 03:41
> > To: Gaurav Garg
> > Subject: RE: Regarding Simultaneous fMRI-EEG Data
> >
> > Dear Gaurav,
> >
> > They are healthy young subjects, no history of psycopathology.
> > Haku is a 29 year-old male, right handed.
> > Ojima is a 25 year-old female, right handed.
> > Sawada is a 26 year-old female, right handed.
> >
> > I'm also mining these datasets right now.
> > I'll tell you if I find something interesting.
> >
> > Makoto
> >
> > --- Gaurav Garg <Garg-G at email.ulster.ac.uk> wrote:
> >
> > > Dear Makoto,
> > >
> > > I am going through the data, you had, provided to me; and till
> > > everything is going fine.
> > > I have only one query at the time and It would be great, if you
> > could
> > > tell me the condition of test subjects, like this data is
> recorded
> > > from the healthy subjects or for some specific disease related
> > > diagnosis/confirmation purpose.
> > >
> > > Thanks,
> > > Stay in touch,
> > >
> > > Gaurav
> > >
> > > ==================================================
> > > Gaurav Garg
> > >
> > > Postgraduate Research Scholar
> > > Room No. MS125
> > > School of Computing and Intelligent Systems
> > > University of Ulster
> > > Magee Campus
> > > Northland Road
> > > Derry
> > > Northern Ireland BT48 7JL
> > > ________________________________________
> > > From: Makoto Miyakoshi [mataothefifth at yahoo.co.jp]
> > > Sent: 13 July 2009 06:15
> > > To: Gaurav Garg
> > > Subject: RE: Regarding Simultaneous fMRI-EEG Data
> > >
> > > Dear Gaurav,
> > >
> > > Sorry for the slow response.
> > > I've been to Switzerland to visit CERN.
> > >
> > > I'm glad you received the DVD.
> > > If you have any question don't hesitate to ask.
> > >
> > > Makoto
> > >
> > >
> > > --- Gaurav Garg <Garg-G at email.ulster.ac.uk> wrote:
> > >
> > > > Dear Makoto,
> > > >
> > > >
> > > >
> > > > Thank you very much!
> > > >
> > > > Today, I have received the EEG-fMRI data DVD, which you have
> sent
> > > to
> > > > me.
> > > >
> > > > I am grateful to you for your efforts, which you had put for
> > > > continuous assistance.
> > > >
> > > >
> > > >
> > > > Please, stay in touch.
> > > >
> > > >
> > > >
> > > > Thanks and Regards
> > > >
> > > > Gaurav Garg
> > > >
> > > > Postgraduate Research Scholar
> > > > Room No. MS125-9
> > > > School of Computing and Intelligent Systems
> > > > University of Ulster
> > > > Magee Campus
> > > > Northland Road
> > > > Derry
> > > > Northern Ireland BT48 7JL
> > > >
> > > >
> > > >
> > > > ________________________________
> > > > From: Gaurav Garg
> > > > Sent: 16 June 2009 12:44
> > > > To: eeglablist at sccn.ucsd.edu
> > > > Subject: Regarding Simultaneous fMRI-EEG Data
> > > >
> > > > Hi,
> > > >
> > > > I am working for some mental degeneration models using
> > Simultaneous
> > > > fMRI-EEG Data Recordings, but I don't have the simultaneously
> > > > captured data available. so, please guide me in this matter,
> that
> > > > where can I get the data.
> > > >
> > > > looking forward for a favourable reply.
> > > >
> > > > Thanks and Regards
> > > > Gaurav Garg
> > > >
> > > >
> > >
> >
>
More information about the eeglablist
mailing list