Zewang,
I wouldn't make the change in that field. Change the field SPM.xBF,
and then go into spm_fmri_design and see where it generates the basis
function : SPM.xBF = spm_get_bf(xBF)
if ~exist('SPM.xBF.bf')
SPM.xBF = spm_get_bf(xBF)
end
that way if it already exists it will just go with it. Then spm will
generate the design matrix for you with your bf. Mind you the exact
phrasing might be different, and I'm pretty sure thats in fmri_design,
but you get the general idea. Why are you using a custom basis
function if you don't mind my asking? What type of bf did you have in
mind? I have been messing around with custom basis functions for a
while now for modeling stimulus absent time periods, i.e., working
memory maintenance, anticipation, etc., so let me know if I can be of
some help, and I'd definitely be interested in hearing what you are
doing in this regard.
Best,
Michael Rubens
--
Research Associate
Gazzaley Lab
Department of Neurology
University of California, San Francisco
On Mon, Apr 28, 2008 at 10:49 AM, zewang <[log in to unmask]> wrote:
> Dear all,
> Could I just do this by replacing the corresponding column of SPM.xX.X
> with my own reference function? Like in my experiment, there are two
> conditions baseline and task, and I modeled them explicitly during setting
> up the design matrix. The reference function reflects the global fluctuation
> time course. Should I just replace the task associated column in SPM.xX.X or
> should I change the model to just explicitly specify the task condition (so
> there will be only one condition modeled in the SPM.mat) and then replace
> the column in SPM.xX.X?
>
> Thanks.
>
>
> ---------------------------------------------
> Ze Wang, Ph. D
> Center for functional neuroimaging,
> Dept of Neurology,
> Medicine school,
> University of Pennsylvania
> 3400 Spruce Stree,
> 3W Gates building,
> Philadelphia, PA 19104
> Tel: 215-662-7341
>
|