[FieldTrip] denoising data
Frédéric Roux
f.roux at bcbl.eu
Thu Apr 3 12:07:11 CEST 2014
Hi Ana, Haiteng and everyone else,
I am joining in on this discussion as I am thinking about a way of combining
ft_componentanalysis with ft_denoise_pca to remove EOG, ECG and EMG activity.
The pipeline that I would like to try is the following:
%% ICA decomposition
cfg = [];
cfg.method = 'runica';
ICs = ft_componentanalysis(cfg,meg_data); % here meg_data has undergone basic preprocessing including bp-filtering as well as
& ft_artifact_jump and ft_artifact_clip
%% Substract artifacts from ICA components
cfg = [];
cfg.refchannel = eeg_data.label(1); % eeg_data contains the EOG (1-2), ECG(3) and EMG(4) ref-channels
cfg.channel = {'runica'};
cfg.zscore = 'yes';
cfg.pertrial = 'yes';
cfg.trials = 'all';
[ICs] = ft_denoise_pca(cfg,ICs,eeg_data); % substract VEOG activity from all ICs
%%
The idea is to repeat this step for all refchannels in order to substract the artifacts.
Then backproject ICs to original data:
%% Backproject data
cfg = [];
meg_data = ft_rejectcomponent(cfg,ICs,meg_data);
The advantage is that you don't have to "throw away" the entire component as sometimes ICs
contain a mix of artifacts and brain activity. So by regressing out the artifacts, you could
avoid throwing away the good stuff.
However, I don't know if this is a good approach. There may be pitfalls that I am not thinking
of right now.
Therefore any suggestions or comments about whether this is a good idea would be very welcome!
I am hoping that some of the ft_guru's could drop a line or two about this!
Fred
----- Original Message -----
From: "Haiteng Jiang" <haiteng.jiang at gmail.com>
To: fieldtrip at science.ru.nl
Sent: Thursday, April 3, 2014 8:27:30 AM
Subject: Re: [FieldTrip] denoising data
Dear Ana,
If I understand correctly , we use ft_denoise_synthetic and ft_denoise_pca to reduce noise due to recording (e.g. environment noise , thermal noise... ) not the brain signal artefact (e.g eye blink). Normally ,I use these two functions in the very early stage .
Best,
Haiteng
Message: 5
Date: Thu, 3 Apr 2014 00:54:42 +0200 (CEST)
From: "Todorovic, A." < a.todorovic at fcdonders.ru.nl >
To: FieldTrip List < fieldtrip at donders.ru.nl >
Subject: [FieldTrip] denoising data
Message-ID:
< 124252612.215995.1396479282885.JavaMail.root at monoceros.zimbra.ru.nl >
Content-Type: text/plain; charset=utf-8
Dear 'trippers,
I'm curious to hear your thoughts on how to best denoise data. What I'm doing is working, but I'd like to hear whether it's logical or reasonable.
In particular, I have two issues. One is that I use ICA to remove blinks, so I have to fit denoising into a pipeline that incorporates ICA (which, I am guessing, works well only without including reference channels).
The other is that I see there are two functions for denoising, ft_denoise_synthetic and ft_denoise_pca.
In sum (1) I'm not really sure WHEN the best moment for denoising is, given that I use ICA, and (2) I don't understand when it's better to use ft_denoise_synthetic and when ft_denoise_pca. I do see that ft_denoise_pca has the option of preprocessing reference channels separately, which makes it easier to use after ICA.
I used to know only about the first function, and my solution was to do ft_denoise_synthetic (using the 'G3BR' option) prior to the ICA. Then I would remove components which contain artifacts, and continue with making ERF/TFRs/whatever. This produces data that is somewhat cleaner than when I skip the denoising step.
I was curious about the ft_denoise_pca when I saw it, so I tried running it on filtered, preprocessed data that I had after I rejected ICA components. [In the process of doing ICA, I used the ft_denoise_synthetic option, as above.] This produced a different TFR at the end of the road, which again looked cleaner. Significantly cleaner, actually, but N=1.
Now I'm not sure if it was a logical step to use both denoising functions, and if it would have been a better idea to do things differently. I'd like to hear both whether something in the logic is wrong, and whether it's inelegant.
Cheers,
Ana
------------------------------
Message: 6
Date: Thu, 3 Apr 2014 07:53:36 +0200
From: Haiteng Jiang < haiteng.jiang at gmail.com >
To: fieldtrip at science.ru.nl
Subject: Re: [FieldTrip] Question on Cluster-based permutation tests
on time-frequency data
Message-ID:
<CAHSK_TQ5v_dOD6iOaqiu= 2+knQ6iHyZqSfyu0r2xvLessob65w at mail.gmail.com >
Content-Type: text/plain; charset="iso-8859-1"
Dear Nithin,
Cluster statistic works both for MEG and EEG data since the machinery
is the same . You can organize the Non-FT in the Fieldtrip style.
Please not that FT have a data structure (channel*frequency*time ), so you
need to transpose your data matrix first. Besides , you also need to
define the layout of your channels , then you know the neighbors of
channels (see ft_prepare_neighbours) because clusters are formed on the
basis of temporal, spatial and spectral adjacency. For more information,
please have a look at the tutorial again
http://fieldtrip.fcdonders.nl/tutorial/cluster_permutation_freq .
All
the best,
Haiteng
>
> Message: 6
> Date: Wed, 2 Apr 2014 10:23:01 -0400
> From: "Nithin Krishna" < nkrishna at mprc.umaryland.edu >
> To: "list, FieldTrip discussion" < fieldtrip at science.ru.nl >
> Subject: [FieldTrip] Question on Cluster-based permutation tests on
> time-frequency data
> Message-ID: < 20140402T102301Z_155300170001 at mprc.umaryland.edu >
> Content-Type: text/plain; charset="utf-8"
>
> Dear All,
> I read the tutorial section on Cluster-based permutation tests on
> time-frequency data as well as the Maris and Oostenveld, Journal of
> Neuroscience Methods, 2007 report. I am interested in setting up a ERP
> experiment on EEG data in this regard I like the procedure and processing
> section of the tutorial, however I would like to know if there is any
> section for EEG, I understand that megplanar is specific for MEG.
> Also is it possible to use 3D matrix (TFR) (time, Freq channels ) non
> feild trip output to run the cluster based permutation tests.
> Looking forward
> Nithin
>
>
>
> --
Haiteng Jiang
PhD candidate
Neuronal Oscillations Group
Donders Institute for Brain, Cognition and Behaviour
Centre for Cognitive Neuroimaging
Radboud University Nijmegen
Visiting address
Room 2.32
Donders Centre for Cognitive Neuroimaging
Kapittelweg 29
6525 EN Nijmegen
the Netherlands
Tel.: +31 (0)243668291
Web: https://sites.google.com/site/haitengjiang/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: < http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20140403/31c1ca38/attachment.html >
------------------------------
_______________________________________________
fieldtrip mailing list
fieldtrip at donders.ru.nl
http://mailman.science.ru.nl/mailman/listinfo/fieldtrip
End of fieldtrip Digest, Vol 41, Issue 8
****************************************
--
Haiteng Jiang
PhD candidate
Neuronal Oscillations Group
Donders Institute for Brain, Cognition and Behaviour
Centre for Cognitive Neuroimaging
Radboud University Nijmegen
Visiting address
Room 2.32
Donders Centre for Cognitive Neuroimaging
Kapittelweg 29
6525 EN Nijmegen
the Netherlands
Tel.: +31 (0)243668291
Web: https://sites.google.com/site/haitengjiang/
_______________________________________________
fieldtrip mailing list
fieldtrip at donders.ru.nl
http://mailman.science.ru.nl/mailman/listinfo/fieldtrip
More information about the fieldtrip
mailing list