You can also try the GIFT toolbox:
http://icatb.sourceforge.net
This will also read SPM.mat files from SPM2 or SPM5 so you can easily
compare the ICA timecourses to your model.
BTW, are you stimuli all shifted to the same -albeit unknown- degree? Or is
even event occurance variable (e.g. +/- 15 seconds)? If the former, then
shifting your model as you suggest may be a good way to go as well.
Regards,
VDC
> -----Original Message-----
> From: SPM (Statistical Parametric Mapping)
> [mailto:[log in to unmask]] On Behalf Of Dr. Carsten Konrad
> Sent: Wednesday, April 19, 2006 8:15 AM
> To: [log in to unmask]
> Subject: [SPM] AW: [SPM] Q: Experiment synchronization
>
> Dear Josef,
>
> my suggestion would be to try independent component analysis.
> If you know the the frequency of the stimulus, but not the
> exact synchonization with scan sessions, this might help you
> to detect the relevant components. The MELODIC module of FSL
> is easy to handle, here is the website:
> http://www.fmrib.ox.ac.uk/fsl/melodic2/index.html.
>
> I don't know if there is an individual component analysis for
> SPM, perhaps someone else in the mailing list can give a hint.
>
> Yours
> Carsten
>
> --------------------------------------------------------------
> ----------
> Carsten Konrad, MD.
> Director of the IZKF-Research Group 4 Neuroimaging.
> Specialist in Neurology, Neurological Intensive Care Medicine.
> Department of Psychiatry and Psychotherapy.
> University of Muenster (UKM).
> Albert-Schweitzer-Str. 11
> 48149 Muenster
> Germany
> Tel. ++49-251-8357214
> Fax ++49-251-8356612
> e-mail: [log in to unmask]
> http://psychiatrie-unimuenster.de/mitarbeiter/HP_Konrad.html
> --------------------------------------------------------------
> -----------
>
>
> -----Ursprüngliche Nachricht-----
> Von: SPM (Statistical Parametric Mapping)
> [mailto:[log in to unmask]] Im Auftrag von Jozef Vilcek
> Gesendet: Mittwoch, 19. April 2006 11:13
> An: [log in to unmask]
> Betreff: [SPM] Q: Experiment synchronization
>
> Hi all,
>
> I want to know how you all feeling about this situation...
>
> You have event related esperiment and aquiered fMRI data, but
> because of some problems (not important now) data are not
> synchronous with stimuls...
> so you can´t say that presentation of first stimul is in
> alignment with first fMRI volume... but you know that
> synchronization lay somewhere abou
> +/- 15 seconds.
>
> My question is... due to complexity of fMRI data and nise, is
> it possible to do analysis of such data? Is it possible to
> find synchronization for example by shifting stimulus onsets,
> estimating model and looking for significant activation?
>
> Thans for any advice.
>
>
> Regards,
>
> Jozef
>
|