Dear Vivek,

In general, as I mentioned in my previous e-mail you can
generate/modify the batch-based part yourself by first configuring a
batch using GUI (in this case press 'Batch' and go in the menu to
SPM->M/EEG->M/EEG source reconstruction->M/EEG head model
specification). Then you can choose the settings you want, press 'Save
as' and change the default '.mat'  setting to '.m' that will generate
a script that you can look it to see how to configure a particular
setting. More specifically...

On Wed, Oct 12, 2011 at 5:14 PM, Buch, Vivek (NIH/NINDS) [G]
<[log in to unmask]> wrote:
> 1) Is there a way using matlabbatch to replace this single_subj_T1.nii with the coregistered image of the individual subject's T1 to the spm template T1 as you had suggested prior?
>

put

matlabbatch{1}.spm.meeg.source.headmodel.meshing.meshes.mri = {'C:\...
path to your image here...'};

instead of

matlabbatch{1}.spm.meeg.source.headmodel.meshing.meshes.template = 1;

>And will the fiducial coordinates need to be updated as well, or will the template fiducial coordinates still work?

Fiducials are also transformed from the template to individual space.
This might be suboptimal however so you could possibly improve things
by finding the fiducials in your individual images, writing down their
MNI coordinates and then using the 'type' option. In batch it will
look something like:

matlabbatch{1}.spm.meeg.source.headmodel.coregistration.coregspecify.fiducial(1).specification.type
= [0 0 0];

> 2) After this head model is created, how do I import the head model for utilization by DCM? Do I need to incorporate it into the preprocessed spm8_subject.mat file somehow?

It is already in the mat file. You don't need to do anything further.
But you should run this kind of script fore every subject.

Best,

Vladimir

>
> Thank you for the help.
> Vivek
>
> -----Original Message-----
> From: Vladimir Litvak [mailto:[log in to unmask]]
> Sent: Tuesday, October 11, 2011 6:37 PM
> To: Buch, Vivek (NIH/NINDS) [G]
> Cc: [log in to unmask]
> Subject: Re: [SPM] DCM for MEG
>
> Dear Vivek,
>
> It's hard to tell from your description whether you've done everything correctly. I suggest you to use the attached script as a reference. I also added head model specification to it. Note that there is also a tool in MEEGTools toolbox (spm_dcm_estimate_group) that makes it possible to run a set of DCMs defined using GUI on a set of subjects with pre-specified head models.
>
> If you also want to specify your DCM in script there is an example for that in \man\example_scripts\DCM_ERP_subject1.m
>
> Best,
>
> Vladimir
>
> On Tue, Oct 11, 2011 at 10:02 PM, Vivek Buch <[log in to unmask]> wrote:
>> Hi all,
>> I just wanted to clarify what exactly needs to be set up in the .mat file "D" structure before being utilized in the DCM set up.
>>
>> I used fieldtrip to read the MEG CTF raw data file. I then converted into spm, giving me the primary "D" structure in matlab format that spm uses for most scripts. I have added MEG and MRI fiducials coregistered to one another in head coordinate space using the transformation matrix created by spm_eeg_inv_datareg.m followed by ft_transform_headshape. These are stored in D.inv.datareg.fid_meg (MEG fiducials in head coordinate space; this is also in D.fiducials) and D.inv.datareg.fid_mri (MRI fiducials in head coordinate space). The D.sensors which was created by the fieldtrip/spm conversion of raw data gives the location of MEG sensors, which I think is in head coordinate space.
>> The MEG data is stored in D.data.
>>
>> Once I filter, downsample, and remove artifact, will the file be ready for use by the model? Or is there something else I must do first?
>>
>> Particularly, am I supposed to do anything to the actual D.sensors variable as a result of the MRI/MEG coregistration?
>>
>> Thank you for the help.
>> Best,
>> Vivek
>>
>