[FieldTrip] AAL atlas source parcellation

[Schoffelen, J.M. (Jan Mathijs)]

I'm doing source parcellation with MEG data by using AAL atlas. At some point in the documentations I read the following :
*atlas.pos = source_conn.pos; % otherwise the parcellation won't work*
and after that has used ft_sourceparcellate function. I have some questions about it.

1- I want to use ft_volumelookup for my parcellation. Do I need to replace my atlas pos field with source_model pos field too (with ft_volumelookup)?

I don’t know. As far as I am aware, the tutorial that you refer to does not use ft_volumelookup.

2- why the commented part says that "otherwise the parcellation won't work". I cannot understand it!

The parcellation won’t work, because the function requires a one-to-one mapping of source positions in the functional data, and locations in the atlas.

3- My atlas dimensions are not equal with source_model dimensions (so the above replacement doesn't sense I think). Does this really matters?

Yes, the parcellation function requires your atlas to be sampled in the same locations as your functional data. This means that you either need to interpolate your source data onto the atlas positions, or (better, because less computationally demanding) the other way around.

4- when I load AAL atlas it doesn't have any pos field! why? Do I need to replace pos field for this atlas too?

The AAL atlas does not have a ‘pos’ field, because it’s a volumetrically defined data structure (a so called volume structure, where the mapping between ‘voxel’ positions in the 3D block of data and world coordinates is given by the transform field. Your functional data is a so-called ’source structure’, where the world coordinates of the individual data elements are directly represented in the pos field. If you look around a bit on the website (start with searching for ft_datatype_source ft_datatype_volume) things ware explained.

Jan-Mathijs



_______________________________________________
fieldtrip mailing list
https://mailman.science.ru.nl/mailman/listinfo/fieldtrip
https://doi.org/10.1371/journal.pcbi.1002202

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20190718/f60c9a8e/attachment-0002.html>
-------------- next part --------------
_______________________________________________
fieldtrip mailing list
https://mailman.science.ru.nl/mailman/listinfo/fieldtrip
https://doi.org/10.1371/journal.pcbi.1002202