There is no principled way to compare the results between modalities.
Bayesian model comparison will not work here because the data is
different. Also one wouldn't necessarily expect to get the same
results because different modalities might be sensitive to different
cortical sources. So you can either decide that you only use one
modality throughout and ignore the other (as people who do e.g.
beamforming do) or you can look at both results and discuss why they
are similar or different.
On Thu, May 26, 2011 at 11:36 AM, <[log in to unmask]> wrote:
> Dear Vladimir,
> Thanks for your reply! May I also ask a silly question here: in this case,
> for instance, I run VB-ECD for both modalities and get two results.
> After the comparison, I should choose one of them which is most likely
> close to the observation, is it right?
> Many thanks in advance,
>> Dear Jing,
>> At the moment there is no way to combine modalities for VB-ECD and DCM.
>> You can only work with each modality separately and compare the results.
>> Sent using BlackBerry® from Orange
>> -----Original Message-----
>> From: Jing Kan <[log in to unmask]>
>> Sender: "SPM (Statistical Parametric Mapping)" <[log in to unmask]>
>> Date: Thu, 26 May 2011 10:43:35
>> To: <[log in to unmask]>
>> Reply-To: [log in to unmask]
>> Subject: [SPM] Question about the "MEG" and "MEGPLANAR" modalities of
>> Elekta system.
>> Dear SPM experts,
>> Thanks for concern! I am doing the "3D source reconstruction" with the
>> data from Elekta system. As I know from the SPM tutorial, SPM treats the
>> MEG data of the Elekta system as two different modalities: "MEG" and
>> There are two methods used for 3D source reconstruction: "imaging",
>> "VB-ECD" and "DCM". The method "imaging" can let me select both "MEG"
>> "MEGPLANAR" modalities for reconstruction. However, in the methods
>> "VB-ECD" and "DCM", I can only select either "MEG" or "MEGPLANAR" for
>> May I ask how can I deal with this two modalities in "VB-ECD" and "DCM".
>> How this selection will affect my reconstruction result?
>> Many thanks for your help in advance!
>> Kind regards,