[FieldTrip] computing threshold for PLV

[jan-mathijs schoffelen]

Hi all,

PLV is in principle a perfectly valid method to investigate synchrony, also in continuous data. Jörg (Hipp)'s paper introduces a connectivity metric that is insensitive to the effects of electromagnetic field spread, but it is a measure whose relationship to synchrony is hard to interpret; it quantifies amplitude envelope correlations after explicit removal of the instantaneous mixing, and does not quantify the consistency of phase relationships, as does PLV. If you are after interpreting the connectivity in terms of synchrony, you could look into using the imaginary part of coherency, or the weighted phase lag index, although neither of these methods are easy in their interpretation, either.

Best,
Jan-Mathijs

On May 19, 2013, at 1:55 AM, Sheraz Khan wrote:

> Most of the previois discussion targeted toward event related plv for continous data may be look at Hipp 2012 Nature Neuroscience orthogonal correlation paper
> 
> On May 18, 2013 3:42 PM, "Subramaniam Iyer" <eeguser at hotmail.com> wrote:
> Dear FT Experts
> 
> Thank you for your inputs. I should add here that I have continuous EEG data (so effectively single trial). The way I am computing PLV is by dividing my data into windows of 2 seconds each and computing PLV for each window and then I finally average the PLV values of all the windows.
> 
> So when we talk of shuffling trials, in my case should i consider each of 2 second window as a trial ? Sorry if this sounds stupid as I am new to this kind of analysis.
> 
> From: smoratti at psi.ucm.es
> Date: Sat, 18 May 2013 20:11:11 +0200
> To: fieldtrip at science.ru.nl
> Subject: Re: [FieldTrip] computing threshold for PLV
> 
> Dear Sheraz,
> 
> You might consult the following article: Lachaux, J. P. et al. 1999. Measuring Phase Synchrony in Brain Signal. Human Brain Mapping 8: 194-208. In principle you have to create surrogate data that represent the null hypothesis that there is no phase locking between your electrodes, sources, etc. Then you create hundreds or thousands of surrogate data sets and calculate each time you PLV. That way you create a distribution of your PLV values under the Null hypothesis. Then you check in what percentile your observed PLV lies (e. g. if its beyond the 95% percentile you would reject the Null hypothesis).
> 
> Hope that helps,
> 
> Stephan
> 
> 
> 
> ________________________________________________________
> Stephan Moratti, PhD
> 
> see also: http://web.me.com/smoratti/
> 
> Universidad Complutense de Madrid
> Facultad de Psicología
> Departamento de Psicología Básica I
> Campus de Somosaguas
> 28223 Pozuelo de Alarcón (Madrid)
> Spain
> 
> and
> 
> Center for Biomedical Technology
> Laboratory for Cognitive and Computational Neuroscience
> Parque Científico y Tecnológico de la Universidad Politecnica de Madrid
> Campus Montegancedo
> 28223 Pozuelo de Alarcón (Madrid)
> Spain
> 
> 
> email: smoratti at psi.ucm.es
> Tel.:    +34 679219982
> 
> El 18/05/2013, a las 18:21, Subramaniam Iyer escribió:
> 
> Dear FT Experts,
> 
> I have calculated PLV for a set of EEG data from 5 different patients. Now I want to convert the PLV matrix of each patient into a binary matrix. For this I need a threshold PLV value below which I can assume the phase locking is zero.
> My question is, how do I compute this threshold. I know hard thresholding is one option ( for ex setting 0.1 or 0.2 as threshold), but I guess it is not a very good option.
> 
> Can somebody suggest a better and robust (statistical) way of determining the threshold ? 
> _______________________________________________
> fieldtrip mailing list
> fieldtrip at donders.ru.nl
> http://mailman.science.ru.nl/mailman/listinfo/fieldtrip
> 
> 
> _______________________________________________ fieldtrip mailing list fieldtrip at donders.ru.nl http://mailman.science.ru.nl/mailman/listinfo/fieldtrip
> 
> _______________________________________________
> fieldtrip mailing list
> fieldtrip at donders.ru.nl
> http://mailman.science.ru.nl/mailman/listinfo/fieldtrip
> _______________________________________________
> fieldtrip mailing list
> fieldtrip at donders.ru.nl
> http://mailman.science.ru.nl/mailman/listinfo/fieldtrip

Jan-Mathijs Schoffelen, MD PhD 

Donders Institute for Brain, Cognition and Behaviour, 
Centre for Cognitive Neuroimaging,
Radboud University Nijmegen, The Netherlands

Max Planck Institute for Psycholinguistics,
Nijmegen, The Netherlands

J.Schoffelen at donders.ru.nl
Telephone: +31-24-3614793

http://www.hettaligebrein.nl

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20130521/a136b603/attachment-0002.html>