Dear Todd,
hope I understand you correctly, but this is what I use for something
similar.
In SPM2, the regression coefficients (B's) per regressor in X, as well
as the error e, are written to your working directory (the current
matlab directory at the time you estimate) per voxel, i.e. beta_000N.img
(N=number of regressor) are the betas and ResMS.img is the residual mean
squared value of the residuals e. In matlab you could then (partially)
reconstruct your signal Y using your design matrix X and the saved
values of (some) Beta and ResMS, depending on what you want to see. Don'
t forget to correct for the scaling factor per time point, saved in
SPM.xGX.gSF.
You do not reconstruct DCT regressors this way, used to model out low
frequency drifts (not included in SPM.xX.X).
At least that is what I use to (partially) reconstruct time series from
the model, hope it is all correct. Heard you could also use MarsBar to
do similar things, but I am not familiar with it.
Not using SPM5 very seriously yet, but I guess things work in a similar
fashion there.
Good luck,
Bas
Todd Penney wrote:
> Hi all,
>
> I was wondering if the predicted/estimated Y values (In Y = XB+e) are
> written out to any matrix or output that you can retrieve in SPM5
> after a parameter estimation has been completed?
>
> Thanks
>
> Todd Penney
--
--------------------------------------------
Dr. S.F.W. Neggers
dept. of Psychonomics,Helmholtz Institute
Utrecht University
Heidelberglaan 2
3584 CS, Utrecht, room 17.09
the Netherlands
Tel: (+31) 30 253 4582 Fax: (+31) 30 2534511
E-mail: [log in to unmask]
Web: http://www.fss.uu.nl/psn/web/people/personal/neggers
--------------------------------------------
|