Clusterplot problem

[Roemer van der Meij]

Hi Matteo,

How do your ft_topoplot's look if you average over frequency and average
over time (averaging before baseline correcting)? That could be one of
the causes for your different outputs. I do not expect something to be
wrong with the stat.mask field as you mentioned, as it does fit really
well to the plotted t-value spectrum. Could you also post your
ft_topoplot and ft_clusterplot settings, just to be sure?

Best,
Roemer



On 5/21/2010 1:09 PM, Matteo Fraschini wrote:
> Hi Roemer,
> I tried using the last FT version with same results.
>
> This is the code for statistic:
>
> function [stat]=withinSubjs(GA_act,GA_bl)
> cfg = [];
> cfg.channel = {'MEG','-MLT22','-MLT27','-MLO42','-MLO43','-MRT22','-MRT27','-MRO42','-MRO43'};
> cfg.frequency        = [4 7];
> cfg.avgovertime = 'yes';
> cfg.avgoverfreq = 'yes';
> cfg.method           = 'montecarlo';
> cfg.statistic        = 'actvsblT';
> cfg.correctm         = 'cluster';
> cfg.clusteralpha     = 0.05;
> cfg.clusterstatistic = 'maxsum';
> cfg.minnbchan        = 2;
> cfg.tail             = 0;
> cfg.clustertail      = 0;
> cfg.alpha            = 0.05;
> cfg.numrandomization = 500;
> n_subjs=size(GA_act.powspctrm);
> n_subjs=n_subjs(1);
> design = zeros(2,2*n_subjs);
> for i = 1:n_subjs
>    design(1,i) = i;
> end
> for i = 1:n_subjs
>    design(1,n_subjs+i) = i;
> end
> design(2,1:n_subjs)        = 1;
> design(2,n_subjs+1:2*n_subjs) = 2;
> cfg.design   = design;
> cfg.uvar     = 1;
> cfg.ivar     = 2;
> stat = ft_freqstatistics(cfg,GA_act,GA_bl)
>
> You can see the ft_topoplot (with absolute baseline correction) and ft_clusterplot (same time) figures here:
> http://192.167.154.14/ft/topo.jpg
> http://192.167.154.14/ft/cluster.jpg
>
> I also tried using 'ctf_old' specification when preprocessing my data but an error occurs in ft_megplanar:
>
> ??? Index exceeds matrix dimensions.
> Error in ==>  channelposition at 53
> sens.pnt = sens.pnt(used,:);
> Error in ==>  megplanar_sincos at 26
> [pnt, ori, lab] = channelposition(grad);
>
> Any suggestion?
> Thank you very much,
> Matteo
>
>
> Il giorno 20/mag/2010, alle ore 10.24, Roemer van der Meij ha scritto:
>
>
>> Hi Matteo,
>>
>> What FieldTrip version are you using? There have been some changes in the past few months in the functions you use.
>> Could you post all the cfg-settings you used (from statistical analysis onwards)?
>>
>> Best,
>> Roemer
>>
>>
>> On 5/19/2010 3:51 PM,<Matteo Fraschini>  wrote:
>>
>>> Dear all,
>>> I am currently working on a study involving Activation vs Baseline paradigm on MEG data.
>>> By using FT preprocessing, analysis and plotting functions I really can see what I expected
>>> to. But, I have a problem with statistical analysis. I tried to use cluster-based permutation
>>> tests on time-frequency data following the web tutorial. The analysis shows significative
>>> cluster but both ft_clusterplot and stat.mask seem corrupted/incorrect. What I see (with
>>> ft_clusterplot) is significative cluster but in region/channels with no frequency differences
>>> from baseline to activation. Viceversa, I can't see any cluster where the act vs bl difference
>>> is more evident. Looks like channels are not correctly placed in the plot.
>>> Any suggestions?
>>> Thank you very much,
>>> Matteo
>>>
>>> ----------------------------------
>>> 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.
>>>
>>>
>>>
>> --
>> Roemer van der Meij MSc
>> Scientific Programmer&   Data-Analyst
>> Donders Institute for Brain, Cognition and Behaviour
>> Centre for Cognition
>> P.O. Box 9104
>> 6500 HE Nijmegen
>> The Netherlands
>> Tel: +31(0)24 3612631
>> E-mail: r.vandermeij at donders.ru.nl
>>
>> ----------------------------------
>> 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.
>>
>>
> ----------------------------------
> 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.
>
>

--
Roemer van der Meij MSc
Scientific Programmer&  Data-Analyst
Donders Institute for Brain, Cognition and Behaviour
Centre for Cognition
P.O. Box 9104
6500 HE Nijmegen
The Netherlands
Tel: +31(0)24 3612631
E-mail: r.vandermeij at donders.ru.nl

----------------------------------
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.