Felix, which ‘anatomical data’ do you refer to? As I mentioned, the anatomy data that is associated with the MEG-data is defined in BTI-space. If you are talking about the high resolution anatomical data from the Structural Processing pipeline, these are defined in various coordinate systems, which are either warped, or not (the ’Native’ ones).
If you want to get the low-resolution anatomical data (e.g. the 4k-per-hemisphere cortical sheets, as represented in the MEG anatomy directories gifti-files or mat-files) in the subject’s ’native’ coordinate system, which is RAS, with the origin in the anterior commissure, you need to apply the transformation ‘bti2spm’ to the geometrical objects. In matlab, this can be done with the ft_transform_geometry function. We have been a bit sloppy in our terminology of the coordinate systems, and interchangeably used ‘mni’, ’spm’ and 'ACPC-aligned RAS’, although strictly speaking ‘mni’ could also refer to spatially warped transformations. Spatial transformations that are applied in order to coregister spatially defined objects for MEG source reconstruction, need to be rigid body, i.e. linear without scaling. BTI uses the same convention as CTF, which is ALS with the origin midway between the pre-auricular fiducial locations. The definitions of the landmarks in the ‘vox’ system are useless for you, because we did not release the corresponding anatomical image from which these locations have been derived. Best wishes, Jan-Mathijs J.M.Schoffelen, MD PhD Senior Researcher, VIDI-fellow - PI, language in interaction Telephone: +31-24-3614793 Physical location: room 00.028 Donders Centre for Cognitive Neuroimaging, Nijmegen, The Netherlands > On 10 Jan 2018, at 15:11, Felix Yuan > <[email protected]> wrote: > > Dear Jan-Mathijs, > > Thanks a lot! > Now I understand what your meaning is. > But with other use, if possible could you please tell me which coordinate > system is anatomical data aligned to? Bti or Spm? > > And in the pipeline codes, I found spm coordinate system, bti system, ctf > system. And also I found RAS space, subject headspace, MNI space, BTI space. > Also landmarks is in the vox system. > It seems that bti system is in the BTi space and spm system is in the MNI > space, right? Sorry, I am kind of confused about all the spaces and the > corresponding systems. > > Thanks for your kindly help, > All the best, > Felix > > > 发件人: [email protected] > [mailto:[email protected]]代表 Jan Mathijs Schoffelen > 发送时间: 2018年1月10日 16:21 > 收件人: [email protected] > 主题: [HCP-Users] Fwd: MEG leadfield computation > > Dear Felix, > > In general, the exact coordinate system+units in which the geometric > positions of the data objects are defined is irrelevant, as long as they are > the same across the different pieces of information that are combined to > compute the leadfields. In other words, as long as the sensor positions, > volume conduction model, and source model dipole positions share the same > coordinate system and units, the result will be physically correct (yet the > physical units of the leadfield may vary, which will be due to units, rather > than to the coordinate system). > > Indeed, if you want to compute the leadfields, you should use the anatomy > data that is provided to you by us, which, as I mentioned previously, has > been coregistered to the MEG sensors for you. > As you can see, when you load one of the variables, e.g. the headmodel, the > matlab data structure contains a field ‘coordsys’, and ‘unit’. These specify > the coordinate system and metric units of the coordinates, respectively. You > will see that the coordsys = ‘bti’; If I recall correctly, the units are > in‘mm’. If you need to convert this, you can use the ft_convert_units > function in FieldTrip. > > Best wishes, > > Jan-Mathijs > > J.M.Schoffelen, MD PhD > Senior Researcher, VIDI-fellow - PI, language in interaction > Telephone: +31-24-3614793 > Physical location: room 00.028 > Donders Centre for Cognitive Neuroimaging, Nijmegen, The Netherlands > > > > On 10 Jan 2018, at 05:38, Felix Yuan > <[email protected] > <mailto:[email protected]>> wrote: > > Dear HCP experts, > > Sorry to bother you, I am going to computing the leadfield of resting state > MEG preprocessed data. > > But in the fieldtrip pipeline, I need to give the coordsys as a parameter. > Can I use the anatomy data and MEG resting state data in the preprocessed > folder I downloaded from HCP? > Do I also need to transform the coordinate space of anatomy data to MEG data > space? Could you tell me Whether they have been co-registrated to the same > space or not? > > In this > website,http://www.fieldtriptoolbox.org/faq/how_are_the_different_head_and_mri_coordinate_systems_defined > > <http://www.fieldtriptoolbox.org/faq/how_are_the_different_head_and_mri_coordinate_systems_defined> > I found that if the anatomy data is in the CTF MRI (ALS) space, and MEG data > is in the 4D/BTi space (ALS), the units of them are different, with mm and m, > respectively. > > > Thank you for your attention. > Best wishes, > Felix > _______________________________________________ > HCP-Users mailing list > [email protected] <mailto:[email protected]> > http://lists.humanconnectome.org/mailman/listinfo/hcp-users > <http://lists.humanconnectome.org/mailman/listinfo/hcp-users> > > > > > > _______________________________________________ > HCP-Users mailing list > [email protected] <mailto:[email protected]> > http://lists.humanconnectome.org/mailman/listinfo/hcp-users > <http://lists.humanconnectome.org/mailman/listinfo/hcp-users> _______________________________________________ HCP-Users mailing list [email protected] http://lists.humanconnectome.org/mailman/listinfo/hcp-users
