[Homer-users] Deconvolution

[David Boas]

The ability to de-convolve multiple stimuli conditions is a feature request
to hopefully appear in the next release of homer.

However, if you have created an optimal sequence, the de-convolution of a
single stimulus condition should not be altered by the presence of the other
stimulus conditions.  In fact, with an optimal sequence, the de-convolution
can be had by simply doing a block average and thus saving the tremendous
computational effort of a de-convolution. This is true unless your optimal
sequence was designed with blanks.  In that case, you need to subtract the
blank hemodynamic response.





> -----Original Message-----
> From: homer-users-bounces at nmr.mgh.harvard.edu [mailto:homer-users-
> bounces at nmr.mgh.harvard.edu] On Behalf Of Margaret Duff
> Sent: Tuesday, March 01, 2005 5:20 PM
> To: thuppert
> Cc: homer-users at nmr.mgh.harvard.edu
> Subject: RE: [Homer-users] Deconvolution
> 
> hi ted,  thanks for those materials.  as i was going over them, i
> realized a big potential problem in our analysis and i was hoping you(or
> anyone else reading this list) had some thoughts on it.  we have an event
> related auditory task consisting of 3 tones of different frequencies.  i
> used optseq to maximize the design placement of the 3 tones and when we
> have been deconvolving our averages, we would only input the time points
> for a certain frequency.  since homer only has one window to input a
> stimulus train, it did not know that other events were happening within
> the time series that were different frequencies.  since these other events
> could affect the waveforms of our desired frequency, how would we
> deconvolve taking all the different events into account, yet only
> displaying the frequency of interest?  in fMRI analyses, which i suppose
> this deconvolution method is based on, you would input all your stimulus
> types
> and onset times, then chose which you would like to contrast and display.
> does homer do anything similar to this?  thanks for all your help and
> suggestions, margaret
> 
> 
>  On Mon, 28 Feb 2005, thuppert wrote:
> 
> > I'm pretty sure the matlab codes for the deconvolution/averaging are
> > distributed open source wuth the download.  But, admittingly, the
> > documentation there is a bit poor.  The new version, which I hope to put
> the
> > finishing touches on REALLY soon, has more info and the code is more
> > open/documented.
> >
> > Until then, here is a (brief) technical report on deconvolution.
> (attached).
> >
> > I hope that helps ...
> >
> >
> >
> >
> > Ted Huppert, M.Sc.
> >
> > PhD student-Harvard Univ.
> > Dept of Biophysics
> > Photon Migration Imaging lab
> > Mass General Hospital/CNY
> >
> > Tele: (617)726-1223
> > Cell: (617) 869-1205
> >
> > thuppert at nmr.mgh.harvard.edu
> >
> >
> >
> > -----Original Message-----
> > From: homer-users-bounces at nmr.mgh.harvard.edu
> > [mailto:homer-users-bounces at nmr.mgh.harvard.edu]On Behalf Of Margaret
> > Duff
> > Sent: Friday, February 25, 2005 4:13 PM
> > To: homer-users at nmr.mgh.harvard.edu
> > Subject: [Homer-users] Deconvolution
> >
> >
> > Hi Ted,  Do you have any references that apply to the least squares
> > deconvolution method you have employed in HOMer?  We are trying to
> figure
> > out exactly how our data is being treated so any information on how you
> > programmed this would be helpful.  Thanks a lot, Margaret
> >
> > _______________________________________________
> > Homer-users mailing list
> > Homer-users at nmr.mgh.harvard.edu
> > https://mail.nmr.mgh.harvard.edu/mailman/listinfo/homer-users
> >
> 
> _______________________________________________
> Homer-users mailing list
> Homer-users at nmr.mgh.harvard.edu
> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/homer-users