I'm not sure what the specific problem is but in general we found some
issues with DCM-IR when using data different from what it was
originally developed for. We are now working on a more robust version
and will hopefully be able to include in in the next SPM8 update.
2011/11/18 飞鸟 <[log in to unmask]>:
> Dear Vladimir,
> Thank you very much! I have another question about DCM for induced
> responses, I wonder whether you have ever confronted this problem:
> The DCM_IND often seems to stop at one of EM's iterations, I mean that
> this EM step may take several hours. However, if I change the timebin of
> "Tdcm", this problem may diappear, and the time consumption may be several
> seconds. But, this seems not a effective way to solve this problem. Do you
> know the underlying reason causes this problem?
> At 2011-11-18 19:32:53,"Vladimir Litvak" <[log in to unmask]> wrote:
>>There is no formal requirement to do that but it would make sense to
>>use the IMG option when your source locations are based on imaging
>>source reconstruction results.
>>2011/11/18 飞鸟 <[log in to unmask]>:
>>> Dear spm's users,
>>> When we specify the electromagnetic model, we need to choose a
>>> spatial forward(IMG, ECD, LFP) model for DCM. I have a question that whether
>>> this selection should be the same as that in 3D reconstruction>forward step.
>>> Or in other words, if I did 3D reconstruction in IMG forward model, then I
>>> should choose IMG spatial forward model for DCM?
>>> Any help would be greatly appreciated!
>>> Best regards!
>>> Haoran LI (MS)
>>> Brain Imaging Lab,
>>> Research Center for Learning Science,
>>> Southeast University
>>> 2 Si Pai Lou , Nanjing, 210096, P.R.China