[FieldTrip] Problems time-frequency analysis of Local Field Potential

[jan-mathijs schoffelen]

Dear Zuyue,

I agree with Eelke. One thing you need to check is whether the time axis (in data.time{}) is in seconds or milliseconds.
In the cfg to ft_freqanalysis you specify it in seconds. FieldTrip does not care about the exact units, so you need to ensure that the units in the time-axis correspond.

BW,

Jan-Mathijs
 
On Dec 16, 2011, at 1:44 PM, Eelke Spaak wrote:

> Dear Zuyue,
> 
> When you invoke ft_singleplotTFR, do you specify a means of baseline
> correction and/or explicit z-limits (i.e., limits of the color axis)?
> When looking at raw power (i.e., not contrasted), it is advisable to
> use a baseline correction in order to see something. Also, specifying
> z-limits of a different order of magnitude than your data will of
> course lead to all blue plots.
> 
> You should take a look at the time-frequency analysis tutorial,
> particularly the part about visualization:
> http://fieldtrip.fcdonders.nl/tutorial/timefrequencyanalysis#visualization
> 
> If you still don't see anything after you use baseline correction (or
> plot a contrast instead) and have proper z-limits, there might be
> something else wrong.
> 
> Best,
> Eelke
> 
> 2011/12/16 chenzuyue <chenzuyue at live.cn>:
>> Hi, everyone,
>> 
>> I am a new user of FieldTrip. In my study, I want to computes the difference
>> of time-frequency representations of event-related LFP in two conditions.
>> 
>> The data in the first condition is a 4 X 1100 X 103 matrix (Nchan X Nsamples
>> X Ntrials). The data in the second condition is a 4 X 1100 X 107 matrix. The
>> data reformatted are as follows:
>> 
>> dataset1.label (4 X 1 cell );
>> 
>> dataset1.fsample 1000;
>> 
>> dataset1.trial (1 X 103 cell );
>> 
>> dataset1.time (1X 103 cell); % time is from -100ms to 999ms.
>> 
>> 
>> dataset2.label (4 X 1 cell );
>> 
>> dataset2.fsample 1000;
>> 
>> dataset2.trial (1 X 107 cell );
>> 
>> dataset2.time (1X 107 cell ); % time is from -100ms to 999ms.
>> 
>> 
>> The frequency interested is 0-100Hz. The method to calculate TFRs is the
>> multitaper method. The configuration is as below:
>> 
>> cfg = [];
>> cfg.output     = 'pow';
>> cfg.method     = 'mtmconvol';
>> cfg.foi        = 1:2:100;
>> cfg.t_ftimwin  = 5./cfg.foi;
>> cfg.tapsmofrq  = 0.4 *cfg.foi;
>> cfg.toi        = -0.1:0.05:1;
>> cfg.pad        = 'maxperlen';
>> TFRmult1 = ft_freqanalysis(cfg, dataset1);
>> 
>> But after running, I can not get the right results. When I use
>> ft_singleplotTFR to plot graph, there is only blue color. I wonder whether
>> the process and configuration is right.
>> 
>> The method is from one paper of E Maris and R Oostenvels published on the
>> Journal of Neuroscience Methods.164(2007), 177-190; Figure 3.
>> http://www.ncbi.nlm.nih.gov/pubmed/17517438
>> 
>> Any help is much appreciated.
>> 
>> Yours Sincerely,
>> 
>> Zuyue
>> 
>> 
>> 
>> _______________________________________________
>> 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

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20111216/fd5aeb06/attachment-0002.html>