[FieldTrip] Interpolating atlas and comparing ROIs across subjects

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

Hi,

The idea behind using individually back-projected template grids is to facilitate downstream processing (e.g. no need to ‘upsample’ to an artificially high resolution when doing an interpolation/normalisation to template space). When using backprojected templates, the resulting individual grids may differ topologically w.r.t. one another in terms of their ‘inside’ fields. Dipole positions that count as inside for some - but not all - subjects will not count for the average, so it is defendible to just use the intersect of the inside mask across all subjects.

W.r.t. interpolation of the atlas onto the individual participants: I wouldn’t know how to do this, but you’d need to anyhow coregister the individual anatomies to the template space in which the atlas is defined (e.g. by means of ft_volumenormalise) so in the end of the day it boils down to the same thing (at the expense of a lot of extra work). In other words: fine to downsample the atlas and use a fixed parcellation for all subjects. Spatial resolution is probably not whopping anyhow.

Best wishes,
Jan-Mathijs


On 30 Jul 2024, at 16:18, Parker via fieldtrip <fieldtrip at science.ru.nl> wrote:

Hello,

I want to compare a handful of ROIs between subjects, so I: 1) did my source analysis on a subject-specific grid that was inverse-warped to a template grid 2) used the AAL atlas to interpolate onto my sourcemodel to have ROIs.

When I interpolated the atlas onto the source model of one of my subjects, I was able to get a list of indices for the sources that are in my ROIs. I planned to use those same indices to index sources for every subject, as I assume those indices point to the same positions in the brain across subjects.

I was wondering if it is recommended/custom to use that approach when determining which sources to look in my ROIs, or rather, interpolating the atlas for each subject individually to determine which sources to look at.

(The reason I ask is because when I index those sources on a different subject, I notice generally the sources do appear in the same region of the brain - but some sources, understandably, are in a slightly different position (ex. Just outside of the headmodel).

Thanks and I appreciate the help
_______________________________________________
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/20240731/7d2bb3e8/attachment.htm>