Inter-trial phase coherence (ITC)

Discussions
1

Hello all,

I’m very new to brainstorm, & I’m loving it so far.

I’m curious what methods people are using for ITC.

  • Exporting to different software?
  • Saving the complex values for all trials then using matlab functions from within brainstorm?
  • Some standardly used, but undocumented, plugin?

Thanks!
Matt

2

We talked previously (@MartinC) about having this easily available through Brainstorm, either as standalone process, or an option to save intertrial coherence instead of average power from the Time-frequency process. You could do it for now by calculating the complex wavelet coefficients through the Time-frequency process (save complex values). Then loading it into Matlab and computing the metric you want as described here:

http://www.fieldtriptoolbox.org/faq/itc (the sum is always taken over trials)

There’s two different metrics that people use, one that uses only phase information (inter-trial phase coherence, like phase-locking-value) and one that keeps also amplitude information (inter-trial linear coherence, like coherence). But I’ve read different papers referring to either of the two when talking about ITC.

3

Ok cool. Until now all my TF computations have been either with EEGLab or just manually in matlab, so that will be straightforward. Is there a decent way to load the ITC values back into Brainstorm for stats &/or visualization?

4

@MartinC: Can I let you handle this with @peterd and @Sylvain ?

5

I just saw @Francois’s comment from yesterday on the thread Import time-frequency data from fieldtrip that one solution would be to compute the equivalent file in Brainstorm & replace the data matrix. I should be able to do that with some tinkering (though of course if someone has already figured it out, I wouldn’t turn them away).

Suppose I computed a TF matrix of magnitude averaged over all trials in Brainstorm, then replaced the magnitude values with ITC. Am I correct in thinking Brainstorm would treat the new data in the same way? Is there a different data format that I should “tell Brainstorm” it is that would work better?

Thanks again.

6

This could work as well.
You have to pay attention to the additional processing that might be done on time-frequency files when you load them. Setting the field “Measure” to ‘none’ should prevent any transformation of the values in the TF matrix.

7

I just saw this old post about computing ITC. I saw the link to the sample code using fieldtrip, but I was just curious if you have implemented something similar in Brainstorm?

Thanks,
Ann

8

@Sylvain @leahy @peterd ?

9

I don’t have a working BST process for this, sorry.

10

Hey,
It’s my turn to compute ITC :). Are there some news about a BST process for this, please? Or some advices from someone who exported data from BST to fieldtrip to compute ITC with fieldtrip, please?
Thanks,
Emmanuelle