specifying individual lateralization in sourcestatistic
a.stolk@fcdonders.ru.nl
a.stolk at FCDONDERS.RU.NL
Fri Nov 5 17:48:17 CET 2010
Hi Tony,
Appreciated. The answer to your question depends on what your data looks like and the conclusions you'd like to draw.
1) A baseline-correction might come in handy if you have more task trials than baseline trials. What happens, is that for every every task trial the relative difference is taken against the average of the baselines. At the second level of your analysis, only the average per subject of the relative differences is taken.
Here, the procedure is:
-frequency analysis
-baseline correction
-source analysis
-sourcegrandaverage
-sourcestatistics
2) When you do T-statistics on tasks vs baselines, the standard deviation plays a role in the computation of your T values which then is a more robust estimate of the difference at the subject level. Preferably you use common filters when doing sourceanalysis as the average spatial filter then is based on more trials (greater SNR). Have a look at our page;
http://fieldtrip.fcdonders.nl/example/common_filters_in_beamforming
Here, the procedure is:
-source analysis (common filter)
-split the source data into task and baseline
-sourcestatistics
-sourcegrandaverage
-sourcestatistics
Best regards,
Arjen
----- Original Message -----
From: "Antony D Passaro" <Antony.Passaro at UTH.TMC.EDU>
To: FIELDTRIP at NIC.SURFNET.NL
Sent: Friday, November 5, 2010 5:30:10 PM GMT +01:00 Amsterdam / Berlin / Bern / Rome / Stockholm / Vienna
Subject: Re: [FIELDTRIP] specifying individual lateralization in sourcestatistic
Hi Arjen,
Thank you very much for your reply, that was very helpful. I was wondering would source_task and source_baseline come straight from ft_sourceanalysis or would I first redefine avg.pow in terms of task - baseline / baseline (as in the sourceanalysis example on the Fieldtrip website)?
Thank you,
-Tony
-----Original Message-----
From: FieldTrip discussion list [mailto:FIELDTRIP at NIC.SURFNET.NL] On Behalf Of a.stolk at fcdonders.ru.nl
Sent: Friday, November 05, 2010 11:12 AM
To: FIELDTRIP at NIC.SURFNET.NL
Subject: Re: [FIELDTRIP] specifying individual lateralization in sourcestatistic
Hi Tony,
Basically, you'd only need to change these lines.
cfg.design = [(ones(1,Ntrials) ones(1,Ntrials)*2];
cfg.ivar = 1;
stat = ft_sourcestatistics(cfg, source_task, source_baseline);
Note that you want the SNR of your baseline segments and the task segments to be the same. You'll have to make sure you'll have as many baseline cross-spectral-density matrices (output from your fourieranalysis) as task CSD matrices.
Best regards,
Arjen
----- Original Message -----
From: "Antony D Passaro" <Antony.Passaro at UTH.TMC.EDU>
To: FIELDTRIP at NIC.SURFNET.NL
Sent: Friday, November 5, 2010 4:47:32 PM GMT +01:00 Amsterdam / Berlin / Bern / Rome / Stockholm / Vienna
Subject: Re: [FIELDTRIP] specifying individual lateralization in sourcestatistic
Hi Michael (or anyone else),
I had a question in regards to your comment in this email, specifically where you mention obtaining t-values from first level statistics. In the example on the Fieldtrip website, ft_sourcestatistics describes a method of obtaining the t-values when comparing conditions directly but I was wondering how it might be possible to obtain the t-values by using source statistics for a task-vs-basline comparison for one condition? More specifically, what would have to be changed in the example from the website (posted below)?
cfg = [];
cfg.dim = source.dim;
cfg.method = 'montecarlo';
cfg.statistic = 'indepsamplesT';
cfg.parameter = 'pow';
cfg.correctm = 'cluster';
cfg.numrandomization = 1000;
cfg.alpha = 0.05;
cfg.tail = 0;
cfg.design(1,:) = [1:length(find(design==1)) 1:length(find(design==2))];
cfg.design(2,:) = design;
cfg.uvar = 1; % row of design matrix that contains unit variable (in this case: trials)
cfg.ivar = 2; % row of design matrix that contains independent variable (the conditions)
stat = ft_sourcestatistics(cfg, source);
Thank you very much for your help,
-Tony
-----Original Message-----
From: FieldTrip discussion list [mailto:FIELDTRIP at NIC.SURFNET.NL] On Behalf Of Michael Wibral
Sent: Friday, October 15, 2010 9:08 AM
To: FIELDTRIP at NIC.SURFNET.NL
Subject: Re: [FIELDTRIP] specifying individual lateralization in sourcestatistic
Hi Thomas,
I suggest you simply extract the power values from the results of sourceanalysis (or t-values if you already did a fisrt level statistics) and average them by hand in MATLAB - as ROI is known for each subject this should be easy (you have to extract the values by the voxel indices contained in your ROI. These in turn you can find in interactive mode source plotting if everything else fails). This way you end up with a single value per subject and condition. Afterwards you do a simple permutation test by hand in MATLAB.
((The test for two conditions for example can be done this way:
1. concatenate all data in a vector D first data in one condition, then in the other) 2. compute your test metric between first and second half of D.
3. shuffle the entries of D: DS=D(randperm(length(D)); % creates a new D with shuffled entries by shuffling the indexes of the old one 4. compute your test metric for DS and remember it 5. Do 3+4 N times (e.g. >1901 times for accurate testing at alpha<0.05) 6. check where your original test metric obtained from D is in the distribution of the mtrices obtained from the DS.))
Hope this helps,
Michael
-----Ursprüngliche Nachricht-----
Von: "Thomas Hartmann" <thomas.hartmann at UNI-KONSTANZ.DE>
Gesendet: Oct 15, 2010 3:30:40 PM
An: FIELDTRIP at NIC.SURFNET.NL
Betreff: [FIELDTRIP] specifying individual lateralization in sourcestatistic
> hi,
>i want to do a roi analysis on source data and average over the all the
>voxels. the problem is that i am dealing with a lateralized effect that
>i expect to show up either on the left or the right side of the same
>structure. this lateralization is individual for each subject and known
>a priori.
>is there any possibility to individually define, which side to take
>into account for the statistics?
>
>thanks in advance,
>thomas
>
>--
>Dipl. Psych. Thomas Hartmann
>
>OBOB-Lab
>University of Konstanz
>Department of Psychology
>P.O. Box D25
>78457 Konstanz
>Germany
>
>Tel.: +49 (0)7531 88 4612
>Fax: +49 (0)7531-88 4601
>Email: thomas.hartmann at uni-konstanz.de
>Homepage: http://www.uni-konstanz.de/obob
>
>"I am a brain, Watson. The rest of me is a mere appendix. " (Arthur
>Conan Doyle)
>
>-----------------------------------------------------------------------
>---- You are receiving this message because you are subscribed to the
>FieldTrip list. The aim of this list is to facilitate the discussion
>between users of the FieldTrip toolbox, to share experiences and to
>discuss new ideas for MEG and EEG analysis.
>See also http://listserv.surfnet.nl/archives/fieldtrip.html
>and http://www.ru.nl/neuroimaging/fieldtrip.
>-----------------------------------------------------------------------
>----
---------------------------------------------------------------------------
You are receiving this message because you are subscribed to the FieldTrip list. The aim of this list is to facilitate the discussion between users of the FieldTrip toolbox, to share experiences and to discuss new ideas for MEG and EEG analysis.
See also http://listserv.surfnet.nl/archives/fieldtrip.html
and http://www.ru.nl/neuroimaging/fieldtrip.
---------------------------------------------------------------------------
---------------------------------------------------------------------------
You are receiving this message because you are subscribed to
the FieldTrip list. The aim of this list is to facilitate the discussion
between users of the FieldTrip toolbox, to share experiences
and to discuss new ideas for MEG and EEG analysis.
See also http://listserv.surfnet.nl/archives/fieldtrip.html
and http://www.ru.nl/neuroimaging/fieldtrip.
---------------------------------------------------------------------------
---------------------------------------------------------------------------
You are receiving this message because you are subscribed to
the FieldTrip list. The aim of this list is to facilitate the discussion
between users of the FieldTrip toolbox, to share experiences
and to discuss new ideas for MEG and EEG analysis.
See also http://listserv.surfnet.nl/archives/fieldtrip.html
and http://www.ru.nl/neuroimaging/fieldtrip.
---------------------------------------------------------------------------
More information about the fieldtrip
mailing list