[FieldTrip] Problem with Geodesic 129 layout!!
KatrinH Heimann
katrinheimann at gmail.com
Mon Sep 8 21:38:01 CEST 2014
Dear Joseph, thanks so much, I really appreciate your help. I was also
wandering, if maybe there is another bug in my code.
Our nets are 128 channels, Hydrocel, but as a colleague of me adviced me to
do so, I also tried the 129 layout.
My code you find below (without selection of bad channels etc.)
If you need also the data of two subjects to have a look, let me know!!!
Best and thanks really!
Katrin
subject=input('Which subject do you want to analyse? ','s');
name = strcat(subject,'.raw');
subjectnumber=input('Which is the subjectnumber?', 's');
sb=subjectnumber
%subjectnumber = strcat(subjectnumber, '.txt')
%%
cfg = [];
cfg.dataset = name;
cfg.trialfun = 'ft_trialfun_general'; % this is the default
cfg.trialdef.eventtype = 'trigger';
cfg.trialdef.eventvalue = 'stim'; % the value of the stimulus trigger
cfg.trialdef.prestim = 0.234; % in seconds
cfg.trialdef.poststim = 7.266; % in seconds (whole time of video
observation, but maybe we need less)
cfg = ft_definetrial(cfg);
cfg.trl([1:7],:) = [];
%change timeline according to constant offset of 16 ms = 8 samples (because
recorded with 500 hz)in
%structure trial
cfg.trl(:,3)=cfg.trl(:,3)-8
%change timeline to make the cut the zeropoint
cfg.trl(:,3)=cfg.trl(:,3)- 1000;
%% preprocess data
cfg.channel = 'all';
cfg.preproc.detrend = 'yes';
cfg.preproc.demean = 'yes';
cfg.preproc.baselinewindow = [-2.1 -2.0]
%cfg.preproc.bpfilter = 'yes'; - the data here has to be already filtered
%(0.1-45) and bad channels have to be replaced!!!!
%cfg.preproc.bpfreq = [6 32];
%
%
obs_data = ft_preprocessing(cfg);
%
save (strcat(sb,'obs_data') , 'obs_data')
%% determining channel neighbours (necessary for Artifact detection)
cfg = [];
cfg.channel = obs_data.label;
cfg.layout = 'GSN128.sfp';
cfg.feedback = 'yes';
lay = ft_prepare_layout(cfg);
cfg_neighb = [];
cfg_neighb.feedback = 'yes';
cfg_neighb.method = 'triangulation';
cfg_neighb.layout = 'GSN128.sfp';
neighbours = ft_prepare_neighbours(cfg_neighb, obs_data);
obs_data.elec = ft_read_sens('GSN128.sfp');
%% Artifacts - to detect bad channels - is not saved!!!
cfg.method = 'summary';
cfg.latency=[0 1];
cfg.layout = 'GSN128.sfp'; % this allows for plotting
obs_data_try = ft_rejectvisual(cfg, obs_data);
cfg = [];
cfg.viewmode = 'vertical';
cfg.latency=[0 1];
cfg.continuous = 'no';
cfg.plotlabels='yes'
cfg = ft_databrowser(cfg,obs_data);
%cfg.artfctdef.reject = 'complete';
obs_data_try = ft_rejectartifact (cfg,obs_data);
%% Preparing neighbours for channel repair - but bad channel info in!!!
% cfg_neighb = [];
% cfg_neighb.feedback = 'yes';
% cfg_neighb.method = 'triangulation';
% cfg_neighb.layout = 'GSN-HydroCel-128.sfp';
% neighbours = ft_prepare_neighbours(cfg_neighb, obs_data);
% Interpolate and put into new data structure
cfg = [];
cfg.badchannel = {};
cfg.layout = 'GSN128.sfp';
cfg.method = 'nearest';
cfg.neighbours = neighbours;
obs_data_channelrepaired = ft_channelrepair(cfg,obs_data)
% Check reconstruction
cfg = [];
cfg.viewmode = 'vertical';
cfg.continuous = 'no';
cfg.plotlabels='yes'
cfg = ft_databrowser(cfg,obs_data_channelrepaired);
%cfg.artfctdef.reject = 'complete';
obs_data_clean1 = ft_rejectartifact (cfg,obs_data_channelrepaired);
% dat1=obs_data;
% dat2=obs_data_channelrepaired;
%
% x=dat1.trial{1}(62,:); % 68 is channel index of E68
% y=dat2.trial{1}(62,:);
% plot(x);hold on;plot(y,'r');
%
% x=dat1.trial{1}(72,:);
% y=dat2.trial{1}(75,:);
% figure;
% plot(x);hold on;plot(y,'r')
%
% x=dat1.trial{1}(75,:);
% y=dat2.trial{1}(75,:);
% figure;
% plot(x);hold on;plot(y,'r')
%% artifact rejection/trial inspection - throw out electrode jumps etc.
cfg = [];
cfg.viewmode = 'vertical';
cfg.continuous = 'no';
cfg.plotlabels='yes'
cfg = ft_databrowser(cfg,obs_data_channelrepaired);
%cfg.artfctdef.reject = 'complete';
obs_data_clean1 = ft_rejectartifact (cfg,obs_data_channelrepaired);
%% ICA - Anzahl der Komponenten anpassen!- adapt numcomponent according to
channels interpolated %% 128-number of interp. channels)
cfg = [];
cfg.channel = {'all'};
cfg.numcomponent = 128
comp = ft_componentanalysis(cfg,obs_data_clean1);
save (strcat(sb,'comp_all') , 'comp')
%%
cfg = [];
cfg.viewmode = 'component';
cfg.continuous = 'no';
cfg.plotlabels='some';
cfg.layout = 'GSN128.sfp';
ft_databrowser(cfg,comp);
%% poweranalysis components
cfg = [];
cfg.output = 'pow';
cfg.channel = 'all';%compute the power spectrum in all ICs
cfg.method = 'mtmfft';
cfg.taper = 'hanning';
cfg.foi = 0:0.2:50;
obs_freq = ft_freqanalysis(cfg, comp);
%And you can plot the spectra:
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
nsubplots = 16;
nbyn = sqrt(nsubplots);% sqrt(nsubplots) should not contain decimals, type
doc subplot
Nfigs = ceil(size(comp.topo,1)/nsubplots);
tot = Nfigs*nsubplots;
rptvect = 1:size(comp.topo,1);
rptvect = padarray(rptvect, [0 tot-size(comp.topo,1)], 0,'post');
rptvect = reshape(rptvect,nsubplots,Nfigs)';
for r=1:size(rptvect,1);
figure;set(gcf,'units','normalized','outerposition',[0 0 1 1]);%full
screen
k=0;
for j=1:size(rptvect,2);
if~(rptvect(r,j)==0);
k=k+1;
cfg=[];
cfg.channel = rptvect(r,j);
cfg.ylim =[0 500]
cfg.xlim =[0 50]
subplot(nbyn,nbyn,k);ft_singleplotER(cfg,obs_freq);
end
end
end
%For the IC topos you'll follow the same logic as above but with:
figure
cfg = [];
cfg.component = [1:20]; % specify the component(s) that should be
plotted
cfg.layout = 'GSN128.sfp';
cfg.comment = 'no';
ft_topoplotIC(cfg, comp)
figure
cfg = [];
cfg.component = [21:40]; % specify the component(s) that should be
plotted
cfg.layout = 'GSN128.sfp';
cfg.comment = 'no';
ft_topoplotIC(cfg, comp)
figure
cfg = [];
cfg.component = [41:60]; % specify the component(s) that should be
plotted
cfg.layout = 'GSN128.sfp';
cfg.comment = 'no';
ft_topoplotIC(cfg, comp)
figure
cfg = [];
cfg.component = [61:80]; % specify the component(s) that should be
plotted
cfg.layout = 'GSN128.sfp';
cfg.comment = 'no';
ft_topoplotIC(cfg, comp)
%% Seperate observation conditions
name1= strcat(num2str(sb),'_90.xls');
list1=xlsread (name1)
name2= strcat(num2str(sb),'_180.xls');
list2=xlsread (name2)
%
cfg = []
cfg.trials = [list1]
obs90_data = ft_redefinetrial(cfg,obs_data_channelrepaired)
cfg = []
cfg.trials = [list2]
obs180_data = ft_redefinetrial(cfg,obs_data_channelrepaired)
%%PIPELINE FOR obs90
%% Decompose original data according to the components found before
load (strcat(sb,'comp_all'))
cfg = [];
cfg.numcomponent = 128;
cfg.unmixing = comp.unmixing;
cfg.topolabel = comp.topolabel;
comp_1 = ft_componentanalysis(cfg, obs90_data);
%% Reject component
cfg = [];
cfg.component = [];
obs90_data_ica_cleaned = ft_rejectcomponent(cfg, comp_1, obs90_data);
save (strcat(sb,'obs90_ica_cleaned') , 'obs90_data_ica_cleaned')
%% Artifacts - final detection
cfg.method = 'summary';
cfg.latency=[0 1];
cfg.layout = 'GSN128.sfp'; % this allows for plotting
obs90_data_clean3 = ft_rejectvisual(cfg, obs90_data_ica_cleaned);
cfg = [];
cfg.viewmode = 'vertical';
cfg.latency=[0 1];
cfg.continuous = 'no';
cfg.plotlabels='yes'
cfg = ft_databrowser(cfg,obs90_data_clean3);
%cfg.artfctdef.reject = 'complete';
obs90_data_clean2 = ft_rejectartifact (cfg,obs90_data_clean3);
% Save clean data
save (strcat(subject,'obs90_clean') , 'obs90_data_clean2')
%% Rereferencing data
cfg = [];
cfg.channel = 'all';
cfg.preproc.reref = 'yes';
cfg.preproc.refchannel = 'all';
obs90_data_ref= ft_preprocessing(cfg,obs90_data_clean2)
save (strcat(subject,'obs90_ref') , 'obs90_data_ref')
%% Snap out smaller pieces (the third second)
cfg = []
cfg.toilim = [0 1]
obs90_data_small = ft_redefinetrial(cfg,obs90_data_ref)
%% TIMELOCK ERP
obs90_data_ERP = ft_timelockanalysis(cfg, obs90_data_small);
save (strcat(subject,'obs90_ERP') , 'obs90_data_ERP')
%% plot it
cfg = [];
cfg.layout = lay ;
cfg.interactive = 'yes';
ft_multiplotER(cfg, obs90_data_ERP)
%% PIPELINE FOR obs180
%% Decompose original data according to the components found before
load (strcat(sb,'comp_all'))
cfg = [];
cfg.numcomponent = 128
cfg.unmixing = comp.unmixing;
cfg.topolabel = comp.topolabel;
comp_2 = ft_componentanalysis(cfg, obs180_data);
%% Reject component
cfg = [];
cfg.component = [];
obs180_data_ica_cleaned = ft_rejectcomponent(cfg, comp_2, obs180_data);
save (strcat(sb,'obs180_ica_cleaned') , 'obs180_data_ica_cleaned')
%% Artifacts final 180
cfg.method = 'summary';
cfg.latency=[0 1];
cfg.layout = 'GSN128.sfp'; % this allows for plotting
obs180_data_clean3 = ft_rejectvisual(cfg, obs180_data_ica_cleaned);
cfg = [];
cfg.viewmode = 'vertical';
cfg.latency=[0 1];
cfg.continuous = 'no';
cfg.plotlabels='yes'
cfg = ft_databrowser(cfg,obs180_data_clean3);
%cfg.artfctdef.reject = 'complete';
obs180_data_clean2 = ft_rejectartifact (cfg,obs180_data_clean3);
% Save clean data
save (strcat(subject,'obs180_clean') , 'obs180_data_clean2')
%% Rereferencing data
cfg = [];
cfg.channel = 'all';
cfg.preproc.reref = 'yes';
cfg.preproc.refchannel = 'all';
obs180_data_ref= ft_preprocessing(cfg,obs180_data_clean2)
save (strcat(subject,'obs180_ref') , 'obs180_data_ref')
%% Snap out smaller pieces (the third second)
cfg = []
cfg.toilim = [0 1]
obs180_data_small = ft_redefinetrial(cfg,obs180_data_ref)
%% TIMELOCK ERP
obs180_data_ERP = ft_timelockanalysis(cfg, obs180_data_small);
save (strcat(subject,'obs180_ERP') , 'obs180_data_ERP')
%% plot 180 ERP
cfg = [];
cfg.layout = lay ;
cfg.interactive = 'yes';
ft_multiplotER(cfg, obs180_data_ERP)
%% plot both ERPs
cfg = [];
cfg.layout = lay ;
cfg.interactive = 'yes';
cfg.showlabels = 'yes';
ft_multiplotER(cfg, obs90_data_ERP, obs180_data_ERP)
%% plot difference wave
difference = obs180_data_ERP; % copy one of the
structures
difference.avg = obs180_data_ERP.avg - obs90_data_ERP.avg; % compute the
difference ERP
cfg = [];
cfg.layout = lay;
cfg.interactive = 'yes';
ft_multiplotER(cfg, difference)
2014-09-07 7:53 GMT+02:00 Joseph Dien <jdien07 at mac.com>:
> Hi,
> I’m one of the main Fieldtrip contributors working on supporting EGI
> data. A couple things. First of all, exactly which net do you use? If
> they are 129-channel, there is still the question of whether they are the
> Hydrocel or the older geodesic nets. Regardless, that shouldn't cause an
> error like this. Could you send me the file you are trying to process and
> the script you are using and I’ll try troubleshooting it.
>
> Joe
>
>
> On Aug 26, 2014, at 12:53 AM, Ana Laura Diez Martini <
> diezmartini at gmail.com> wrote:
>
> Which nets do you use? I use EGI. I tried both the ones you mention and
> they didn't work. It was hard to find that exact one online but it was the
> only file that actually worked.
>
>
> On Mon, Aug 25, 2014 at 7:52 AM, KatrinH Heimann <katrinheimann at gmail.com>
> wrote:
>
>> Dear Ana, dear all,
>> the layout you used does not correspond to our nets. I tried the
>> GSN-HydroCel-128.sfp and the GSN-HydroCel-129.sfp. Both don't work.
>> Please can anybody help?
>> Cheers
>> Katrin
>>
>>
>> 2014-08-24 5:40 GMT+02:00 Ana Laura Diez Martini <diezmartini at gmail.com>:
>>
>> Try this one <https://www.dropbox.com/s/58zv658rtdant2c/GSN129.sfp?dl=0>
>>> I use
>>>
>>>
>>> On Sun, Aug 24, 2014 at 4:08 AM, KatrinH Heimann <
>>> katrinheimann at gmail.com> wrote:
>>>
>>>> Dear all, my problems seem neverending. This time however i really need
>>>> help.
>>>> I implemented a pipeline for ERPs. All works now, trialdefinition,
>>>> preprocessing, channelreplacement, ica, componentrejection, final
>>>> artifactdetection, timelock of the single subject data. However, when I
>>>> wanna compute the grandaverage of the single subjects I get the following
>>>> error message:
>>>>
>>>> computing average of avg over 19 subjects
>>>> Warning: discarding electrode information because it cannot be averaged
>>>> > In ft_timelockgrandaverage at 249
>>>> the call to "ft_timelockgrandaverage" took 0 seconds and required the
>>>> additional allocation of an estimated 3 MB
>>>> computing average of avg over 2 subjects
>>>> Warning: discarding electrode information because it cannot be averaged
>>>> > In ft_timelockgrandaverage at 249
>>>> the call to "ft_timelockgrandaverage" took 0 seconds and required the
>>>> additional allocation of an estimated 0 MB
>>>> the call to "ft_prepare_layout" took 0 seconds and required the
>>>> additional allocation of an estimated 0 MB
>>>> the call to "ft_topoplotER" took 0 seconds and required the additional
>>>> allocation of an estimated 0 MB
>>>>
>>>> Furthermore in the plot of the significant clusters, the channelnames
>>>> are mixed up (do not correspond to my net)
>>>>
>>>>
>>>> I guess that there is a problem with the layout I use.
>>>> Here the code that I use
>>>>
>>>> %For generating the layout (also in the single subjects):
>>>>
>>>> cfg = [];
>>>>
>>>> cfg.channel = obs_data.label;
>>>>
>>>> cfg.layout = 'GSN-HydroCel-129.sfp';
>>>>
>>>> cfg.feedback = 'yes';
>>>>
>>>> lay = ft_prepare_layout(cfg);
>>>>
>>>>
>>>> cfg_neighb = [];
>>>>
>>>> cfg_neighb.feedback = 'yes';
>>>>
>>>> cfg_neighb.method = 'triangulation';
>>>>
>>>> cfg_neighb.layout = lay;
>>>>
>>>> neighbours = ft_prepare_neighbours(cfg_neighb, obs_data);
>>>>
>>>>
>>>> obs_data.elec = ft_read_sens('GSN-HydroCel-129.sfp');
>>>>
>>>>
>>>> %For computing the grand average
>>>>
>>>> cfg = [];
>>>>
>>>> cfg.channel = 'all';
>>>>
>>>> cfg.latency = 'all';
>>>>
>>>> cfg.parameter = 'avg';
>>>>
>>>> cfg.keepindividual = 'no'
>>>>
>>>> GA_90 = ft_timelockgrandaverage(cfg,all90{:});
>>>>
>>>> GA_180 = ft_timelockgrandaverage(cfg,all180{:});
>>>>
>>>> % "{:}" means to use data from all elements of the variable
>>>>
>>>>
>>>> For plotting the significant clusters
>>>>
>>>> cfg = [];
>>>>
>>>> cfg.style = 'blank';
>>>>
>>>> cfg.layout = lay;
>>>>
>>>> cfg.channellabels = 'yes';
>>>>
>>>> cfg.highlight = 'labels';
>>>>
>>>> cfg.highlightchannel = find(stat.mask);
>>>>
>>>> cfg.comment = 'no';
>>>>
>>>> figure; ft_topoplotER(cfg, GA_90)
>>>>
>>>> title('Nonparametric: significant with cluster multiple comparison
>>>> correction')
>>>>
>>>>
>>>> Do you have ANY idea to this? I am really completely helpless....
>>>>
>>>> thanks and best
>>>>
>>>> Katrin
>>>>
>>>> _______________________________________________
>>>> 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
>
>
>
> --------------------------------------------------------------------------------
>
> Joseph Dien, PhD
> Research Associate
> Cognitive Neurology
> The Johns Hopkins University School of Medicine
>
> Lab E-mail: jdien1 at jhmi.edu
> Private E-mail: jdien07 at mac.com
> Office Phone: 410-614-3115
> Cell Phone: 202-297-8117
> Fax: 410-955-0188
> http://joedien.com
>
>
> _______________________________________________
> fieldtrip mailing list
> fieldtrip at donders.ru.nl
> http://mailman.science.ru.nl/mailman/listinfo/fieldtrip
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20140908/e8846b61/attachment-0002.html>
More information about the fieldtrip
mailing list