[FieldTrip] Connectivity analysis with powcorr_ortho

[Frédéric Roux]

Hi Matthias,

if you want the dimensions of your data to be rpttap_chan_freq
instead of rpttap_chan_freq_time you can try this cfg:

cfg = [];
cfg.channel = 'all';
cfg.method = 'mtmfft';
cfg.output = 'pow';
cfg.taper = 'hanning';
cfg.pad = 'maxperlen';
cfg.keeptapers = 'yes';
cfg.keeptrials = 'yes';
cfg.foi = 0:Fs/2;

Here instead of computing a spectrogram with wavelets, you are 
computing a spectrum (no temporal information) using either
a hanning taper, or any other taper of your choice.
The output of this cfg should give you a data-structure with
the dimensions that you are interested in.

Best,
Fred

Frédéric Roux

----- Original Message -----
From: "Matthias Treder" <matthias.treder at gmail.com>
To: fieldtrip at science.ru.nl
Sent: Thursday, October 16, 2014 8:54:20 PM
Subject: Re: [FieldTrip] Connectivity analysis with powcorr_ortho






Yes I tried cfg.keeptrials = 'yes'. 

I should add that I just have one long trial (resting state). I chopped it into 10 seconds-trials with ft_definetrials and re-did the analysis but it didn't help either. 

As said below, looking into the code the dimensions should be rpttap_chan_freq but I have rpttap_chan_freq_time because I used wavelets (as in Hipp's paper). 


Any other ideas? 



Best, 
Matthias 












Message: 1 
Date: Wed, 15 Oct 2014 19:56:17 +0100 
From: Matthias Treder < matthias.treder at gmail.com > 
To: fieldtrip at science.ru.nl 
Subject: [FieldTrip] Connectivity analysis with powcorr_ortho 
Message-ID: 
< CAAQEB5rODY2bTwOkQhPPKMeWQG68hozdOB4qiSn1pBP2W7LyHw at mail.gmail.com > 
Content-Type: text/plain; charset="utf-8" 

Hello FieldTrippers, 

I want to do a envelope-based connectivity analysis of resting state MEG 
following the regime of Hipp. The respective method is implemented in 
ft_connectivityanalysis as powcorr_ortho. 

Like in the Hipp paper, I am first doing a wavelet analysis on the sensor 
level (10 Hz for me). 
I eventually want to do source-space analysis - currently I stay in sensor 
space, however. 
I keep the complex Fourier coefficients (cfg.output = 'fourier') and 
the result is: 

label: {204x1 cell} 
dimord: 'rpttap_chan_freq_time' 
freq: [10 20] 
time: [1x57400 double] 
fourierspctrm: [4-D double] 
cumtapcnt: [1 1] 
grad: [1x1 struct] 
cfg: [1x1 struct] 


Looking in ft_connectivityanalysis in line 697 the function requires either 
a field .mom or the structure of the data has to be rpttap_chan_freq, but I 
have sensor-level time-frequency data. 

Any idea how to fix my analysis? 

Thanks a lot ! 
Matthias 
-------------- next part -------------- 
An HTML attachment was scrubbed... 
URL: < http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20141015/e7a48133/attachment-0001.html > 

------------------------------ 

Message: 2 
Date: Wed, 15 Oct 2014 22:09:43 +0200 (CEST) 
From: Fr?d?ric Roux < f.roux at bcbl.eu > 
To: FieldTrip discussion list < fieldtrip at science.ru.nl > 
Subject: Re: [FieldTrip] Connectivity analysis with powcorr_ortho 
Message-ID: < 577666873.3492259.1413403783469.JavaMail.root at bcbl.eu > 
Content-Type: text/plain; charset=utf-8 

Hi Matthias, 

have you tried: 

cfg.keetrials = 'yes' in your cfg? 

Best, 
Fred 

Fr?d?ric Roux 

----- Original Message ----- 
From: "Matthias Treder" < matthias.treder at gmail.com > 
To: fieldtrip at science.ru.nl 
Sent: Wednesday, October 15, 2014 8:56:17 PM 
Subject: [FieldTrip] Connectivity analysis with powcorr_ortho 







Hello FieldTrippers, 


I want to do a envelope-based connectivity analysis of resting state MEG following the regime of Hipp. The respective method is implemented in ft_connectivityanalysis as powcorr_ortho. 



Like in the Hipp paper, I am first doing a wavelet analysis on the sensor level (10 Hz for me). 
I eventually want to do source-space analysis - currently I stay in sensor space, however. 
I keep the complex Fourier coefficients (cfg.output = 'fourier') and the result is: 

label: {204x1 cell} 
dimord: 'rpttap_chan_freq_time' 
freq: [10 20] 
time: [1x57400 double] 
fourierspctrm: [4-D double] 
cumtapcnt: [1 1] 
grad: [1x1 struct] 
cfg: [1x1 struct] 


Looking in ft_connectivityanalysis in line 697 the function requires either a field .mom or the structure of the data has to be rpttap_chan_freq, but I have sensor-level time-frequency data. 

Any idea how to fix my analysis? 

Thanks a lot ! 
Matthias 

_______________________________________________ 
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 

End of fieldtrip Digest, Vol 47, Issue 13 
***************************************** 


_______________________________________________
fieldtrip mailing list
fieldtrip at donders.ru.nl
http://mailman.science.ru.nl/mailman/listinfo/fieldtrip