[FieldTrip] EEG connectivity without averaging over the trials.

Gabriel Costa gcosta18 at gmail.com
Wed May 18 19:50:56 CEST 2016

Dear Maggie,

If I'm not mistaken, connectivity estimates can be obtained at single trial if one estimates phase correlation over time (see Cohen, 2014 Analyzing Neural Time Series).

If you are not interested in the time dimension you could just obtain a single value of coherence for each trial, estimating phase coherence over time, which you can then average over trials.
If you still want to maintain time dimension and get a time(x)-frequency(y) plot of coherence, than you should use a sliding window.

I believe this method (coherence over time) is not integrated in either Fieldtrip or EEGLAB, but it's not that difficult to implement this from the matrix with the complex fourier spectra (ft_freqanalysis with cfg.output = 'fourier').

Good luck,
Gabriel Costa


On Wed, May 18, 2016 at 5:33 PM, Mengqi Xing <xmq1990 at gmail.com<mailto:xmq1990 at gmail.com>> wrote:
Dear all,

I'm using fieldtrip to do dynamic EEG connectivity analyses and I have a few questions about trial averaging in ft_connectivityanalysis. I realized regardless which method I'm using, the ft_connectivityanalysis tends to averaging over trials (dimension rpt), which sacrifices the dynamic of  the data. Is it possible not let the function averaging over trials while still getting a meaningful result? ( I tried to duplicate trial and let it average itself and it reported some errors.)  Or is there a connectivity method that allows non-trial-averaging to run?  Thank you very much!

Maggie Xing

fieldtrip mailing list
fieldtrip at donders.ru.nl<mailto:fieldtrip at donders.ru.nl>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20160518/686f70f7/attachment-0002.html>

More information about the fieldtrip mailing list