This is not for DCM but we are trying to find correlation of activation
using one ROI time series as a covariate.
Thanks
Julia Li
Indiana University Medical Center
Psychiatry Department
[log in to unmask]
(317) 278-1771
-----Original Message-----
From: hao zhang [mailto:[log in to unmask]]
Sent: Monday, April 25, 2005 1:45 PM
To: Li, Yu
Cc: [log in to unmask]
Subject: Re: [SPM] Question of connectivity
Dear Li,
I think your steps are not correct for DCM, if that was what you aimed
for.
You get the probability of defined/hypothesized connection and the
driving factor at the end, not T map. You can search on the SPM mailing
list for steps of a DCM study, there are quite questions (including
mine) and answers previously.
Good luck!
Regards,
Zhang Hao
On 4/25/05, Li, Yu <[log in to unmask]> wrote:
>
>
> Dear SPM:
> fMRI study, box-car design, two conditions ABABA. We would like to see
> what areas of the brain are connectivity with a ROI.
>
> Could anyone please let us know if the following steps are correct.
>
> 1. In the step of "specified box car model", the time series extracted
> form a ROI is added as a "user specified regressor".
>
> 2. To see connectivity with this ROI at condition A > B. Set contrast
> 1 (A)
> -1 (B) 1 (ROI time series, the regressor).
>
> We tried this, but t-map is almost the same as we get contrast 1
> (A) -1
> (B) 0 (ROI time series, the regressor). Any suggestion?
>
>
> 3. To see connectivity only in condition A. Set contrast 1 (A) 0 (B)
> 1 (ROI time series, the regressor).
>
> Thanks
>
> Julia Li
> Indiana University Medical Center
> Psychiatry Department
> [log in to unmask]
>
>
|