[FieldTrip] Estimation of the phase locking value

[Daniel Matthes]

Hi,

In our upcoming studies we want to investigate inter-brain couplings on 
source level. Therefore, I have at present some questions about the 
calculation of the phase locking value (PLV), mainly about it's 
implementation in field trip. Since I'm a very beginner both with 
analysing eeg data and with field trip, I possibly ask already answerd 
question, so you've my apologies at the beginnig.

Based on the paper of Dumas et al. (08/2010) I initially tried to 
compute the PLV in field trip by using the hilbert transformation to 
determine the instantaneous phase (ft_preprocessing) and subsequent with 
ft_connectivityanalysis. I rapidly recognized that this is not possible, 
since only freq data is a valid input for ft_connectivityanalysis in 
connection with the parameter 'plv'. Thus, I realized the PLV 
calculation on my own with matlab from scratch and tried to get a 
similar results using the ft_connectivityanalysis function. I've solved 
this issue by now, but on this way several question came up.

The first one is about the result of ft_connectivityanalysis in 
connection with the parameter 'plv'. The function returns the phase 
difference of the compared components and not the phase locking value , 
as defined in Lachaux et al. (1999). What's the reason for this 
implementation? Existing plans for closing this gap?

The second question is related to my initial problem. Why is it not 
possible to use the instananeous phase as input data of 
ft_connectivityanalysis in connection with the parameter 'plv'. I think 
this would make the calculation less complex.

At last I wonder, why the configuration of cfg.channel and 
cfg.channelcmb has no effect in ft_connectivityanalysis in connection 
with the parameter 'plv'. It is only possible to focus solely on two 
channels if these definitions are made during the previous ft_freqanalysis.

I would be thankful for some advice.

Alle the best,

Daniel