[FieldTrip] Question on tutorial http://www.fieldtriptoolbox.org/tutorial/minimumnormestimate

[parham hashemzadeh]

Hi Jan
  Thank you for your previous response. I guess, what confuses me
  http://www.fieldtriptoolbox.org/tutorial/minimumnormestimate
  is the text  in the section "1. Source model: Volumetric processing in 
FreeSurfer"

  "To create a skullstripped anatomical image, you can do the following. 
We assume that you have executed all steps that are described in the 
section about preprocessing of the anatomical MRI, and that you have a 
file that contains the resliced anatomical image, expressed in the 
MNI-RAS coordinate system. "

It explicitly states that resliced anatomical image is expressed in 
MNI-RAS coordinates.
I guess, the source must be in the MNI coordinates.
such that
  T=transform_vox2ctf * inv(transform_vox2spm)
can be applied to source as in the tutorial
% go to the Subject01/bem directory
sourcespace = ft_read_headshape('Subject01-oct-6-src.fif', 'format', 
'mne_source');
sourcespace = ft_convert_units(sourcespace, 'mm');
sourcespace = ft_transform_geometry(T, sourcespace);

save sourcespace sourcespace

the transformation matrix above multiplies the source points from the 
left.
If we denote the source points by S, then T*S

is that right???

Many thanks
best regards parham hashemzadeh





On 2016-05-08 20:35, Schoffelen, J.M. (Jan Mathijs) wrote:
> Hi Parham,
> 
>> On 08 May 2016, at 19:07, parham hashemzadeh <ph442 at cam.ac.uk> wrote:
>> 
>> Dear all
>> I was going through the tutorial: 
>> http://www.fieldtriptoolbox.org/tutorial/minimumnormestimate
>> My question involves the first realignment and reslicing followed by 
>> second realignment and reslicing. I do not understand the step below:
>> 
>> -> T = transform_vox2ctf/transform_vox2spm= 
>> transform_vox2ctf*inv(transform_vox2spm)
>>   and this matrix is applied to sourcespace.
>> 
>> What confuses me, is that
>> a) we realign the anatomical voxel to spm (MNI) coordinates.
>> b) we reslice and save it as shown in step "3. Preprocessing of the 
>> anatomical MRI: reslicing"
>>                                       && 4. Preprocessing of the 
>> anatomical MRI: save to disk
>> At this stage we get the transformation matrix anatomical voxel to spm 
>> i.e transform_vox2spm
>> Moreover, we reside in the SPM coordinates now.
>> c) We then realign to CTF coordinates, no reslicing is done. we Should
>>  GET transform_spm2ctf. However, the obtained transform is called 
>> transform_vox2ct. Why???
> 
> Because the transformation matrix that is attached to the mri
> structure is always going from voxel space to a human interpretable
> coordinate system.
> 
>> We are using the resliced image mrirs= ft_volumereslice(cfg, mri_spm);
>> c) During the source creation, we are already in the MNI coordinates, 
>> because we are using Subject01.mgz.
> 
> The fact that the file has the extension *.mgz does not mean that it
> is in MNI coordinates. The Subject01.mgz is created in the tutorial to
> have a transformation attached to it, that goes from (resliced) voxels
> to MNI.
> 
>> So, why in "STEP 4. Source model: Co-registration of the source space 
>> to the sensor-based head coordinate system we do we need to "
>>         T = transform_vox2ctf/transform_vox2spm;
>> Should it not be transform_vox2ctf changed to transform_spm2ctf
>>  and T=transform_spm2ctf ???
> 
> No, it shouldn’t.
> 
> Best,
> Jan-Mathijs
> 
> 
>> best regards parham
>> 
>> 
>> 
>> 
>> _______________________________________________
>> fieldtrip mailing list
>> fieldtrip at donders.ru.nl
>> http://mailman.science.ru.nl/mailman/listinfo/fieldtrip
> 
> 
> _______________________________________________
> fieldtrip mailing list
> fieldtrip at donders.ru.nl
> http://mailman.science.ru.nl/mailman/listinfo/fieldtrip

-- 
best regards
Parham Hashemzadeh
Research Associate
Department of Applied Mathematics and Theoretical Physics
University of Cambridge, UK.
email: hashemzadeh at damtp.cam.ac.uk