[FieldTrip] How to do statistical analysis of connectivity data with FieldTrip?

Jan Ostrowski j.ostrowski at uke.de
Mon May 30 11:10:02 CEST 2022

Dear all,

First of all, I'd like to thank the community for its support with 
previous issues and the swift replies. It's been helping a lot!

Now, the actual problem at hand:

I'd like to statistically compare connectivity data corresponding to two 
conditions from my paradigm. For that, I have already computed the 
coherence from the preprocessed EEG data for every participant using 
ft_connectivityanalysis. This has been done separately for both 
conditions. The resulting structures look like this:

labelcmb: {1770×2 cell}
        dimord: 'chancmb_freq_time'
     cohspctrm: [1770×40×36 double]
          time: [1×36 double]
          freq: [1×40 double]
           dof: [1×40 double]
           cfg: [1×1 struct]

Now it seems that FieldTrip does not provide specialized functions to 
handle statistical comparisons of connectivity data (please correct me 
if I'm wrong!). So my approach so far consisted in defining a source 
channel, extracting the data from the corresponding combinations, and 
modifying the resulting structures in a way so that ft_freqgrandaverage 
and ft_freqstatistics could handle them (I used cluster-based 
permutation for statistical testing). However, this approach is not very 
elegant and quite prone to errors along the way, as one has to keep 
track of the modifications meticulously.

Is there any way of doing this kind of analysis with FieldTrip, that I 
have not found yet? If not, have any of you done this kind of analysis 
and can offer some advice on how to approach this? My goal is to 
statistically compare the connectivity for every channel combination in 
the data. I have been running that analysis on the whole time-frequency 
spectrum of my data. Also, how would one go about correcting for 
multiple testing in this instance?

Any advice or pointers are greatly appreciated!

Have a nice week!

Jan Ostrowski, M.Sc.
Department of Systems Neuroscience
University Medical Center Hamburg-Eppendorf
Martinistr. 52
Building W34, Room 320b
D-20246 Hamburg, Germany



Universitätsklinikum Hamburg-Eppendorf; Körperschaft des öffentlichen Rechts; Gerichtsstand: Hamburg | www.uke.de
Vorstandsmitglieder: Prof. Dr. Burkhard Göke (Vorsitzender), Joachim Prölß, Prof. Dr. Blanche Schwappach-Pignataro, Marya Verdel


More information about the fieldtrip mailing list