Dear Anne,
A similar remark was commented upon by Klaas earlier this year:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=SPM;T3yaCg;20090227101708%2B0000
All the best,
Guillaume.
Anne Kathrin Rehme wrote:
> Dear SPM users,
>
> I just want to inform you on our observation that in SPM5 generated
> design files with in SPM5 extracted VOI time series produce different
> DCM results with SPM8 as do SPM8 designs and VOI time series although
> both versions are assumed to provide compatible results.
>
> I set up one SPM5 and one SPM8 design with one data set and the same
> design specifications. Activation maxima slightly differed between both
> SPM-maps. The subsequent DCM model in SPM8 showed the same pattern of
> connections but higher coupling estimates when the design and the VOI
> extraction was performed in SPM8 (i.e., better fit to the anatomical model).
>
>
>
> Do you have an idea what could be the cause for this mismatch?
>
>
>
> Anne
--
Guillaume Flandin, PhD
Wellcome Trust Centre for Neuroimaging
University College London
12 Queen Square
London WC1N 3BG
|