[FieldTrip] Output dimension of eLORETA

Andy Liang liang134 at mail.chapman.edu
Thu Nov 12 01:38:06 CET 2020


Dear all fieldtrip experts,

Hope all is well. I am a PhD student working on my thesis applying source
localization to some MEG data. I am now stuck at a technical problem that I
am a bit confused and wasn't able to find a straightforward answer. Let me
know if you have a quick answer for this:

In fieldtrip, I can input a time series (the source data) or the
frequency power of a particular frequency band (eg. alpha band after
ft_freqanalysis) to ft_sourceanalysis with eLORETA method and get the
source output. However, the structure of the output is a little bit
confusing. I would assume that if I am inputting a time series with
n_channels x n_timepoints, I will get an output of also a time series, with
dimension n_source_location x n_timepoints. But in reality, I am getting an
output of n_source_location x 1 no matter what the input is. It seems like
eLORETA takes the time series and produces a summarized score for each of
the source points. I have done quite a bit of research on these algorithms
and I think eLORETA is similar to wMNE which takes in a time series and
outputs a time series except that it adopts a different normalization
function.

Am I missing any parameters that I need to specify or I am just
misunderstanding eLORETA? If you can point me to any resource, it will be
much appreciated.

Thank you very much!
Andy Liang
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20201111/41b415ff/attachment.htm>


More information about the fieldtrip mailing list