[FieldTrip] Inverse warping vs interpolating & normalizing for across-subject source space analysis

[Brian Maniscalco]

When it comes to conducting across-subject analysis on source space data,
I've seen two approaches described on the Field Trip web site and mailing
list.

1) for each subject, interpolate the source space data onto the subject's
brain anatomy and then normalize the result to a standard template before
conducting across-subject analysis

2) for each subject, perform an inverse warp to a template as described in
the following link, which then allows across-subject analysis without
having to interpolate and normalize

http://www.fieldtriptoolbox.org/example/create_single-subject_grids_in_individual_head_space_that_are_all_aligned_in_mni_space

I have seen approach (2) recommended for Field Trip users but I have not
been able to find an in depth discussion of the pros and cons of doing (1)
vs (2). I gather that (2) is computationally simpler and may save
computation time. But aside from computational considerations, is there any
theoretical or statistical benefit to using approach (2) over approach (1)
or vice versa?

thanks,
Brian
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20150528/b9e8759e/attachment.html>