[FieldTrip] Define a ROI in sourcestatistics

[Marc Recasens]

Hi all.

I've been trying to define a region of interest (ROI) in order to 
restrict my analyses to a particular region on the temporal lobe during 
sourcestatistics.

I have two grandaverage sources like:
pos: [33480x3 double]
dim: [31 36 30]
avg: [1x1 struct]
var: [1x1 struct]
dimord: 'voxel'
trial: [1x13 struct]
inside: [12773x1 double]
outside: [20707x1 double]
df: [33480x1 double]
cfg: [1x1 struct]

And I apply ft_sourcestatistics with the following configuration:
cfg             =[];
cfg.dim         = GA_dev_source.dim;
cfg.method      = 'montecarlo';
cfg.statistic   = 'depsamplesT';
cfg.parameter   = 'pow';
cfg.correctm    = 'bonferoni';
cfg.numrandomization = 1000;
cfg.alpha       = 0.05;
cfg.tail        = 0;
nsubj           = length(GA_dev_source.trial);
cfg.design(1,:) = [1:nsubj 1:nsubj];
cfg.design(2,:) = [ones(1,nsubj) ones(1,nsubj)*2];
cfg.uvar        = 1;
cfg.ivar        = 2;
stat = ft_sourcestatistics(cfg, GA_dev_source, GA_stad_source)

However it's not very straightforward to me to apply a ROI restriction 
to this analysis.
I downloaded the files from http://fmri.wfubmc.edu (TD_lobe.img, 
TD_lobe.mat, TD_lobe_Border.mat, TD_lobe_List.mat) which I am suposed to 
use for the ROI.

I tried to add this to the previous configuration:
cfg.atlas = '.../TD_lobe.img';        (the template brain containing the 
different lobes)
cfg.inputcoord = 'mni';
cfg.roi = ROI(12).Nom_C;            The names of the roi i want to 
extract ('Temporal Lobe')

But then...
stat_roi =
prob: [31x36x30 double] <---- All zeros.
mask: [31x36x30 logical]
stat: [31x36x30 double]
ref: [31x36x30 double]
dim: [31 36 30]
inside: [0x1 double] <---- which means there's nothing inside the volume
outside: [33480x1 double]
pos: [33480x3 double]
cfg: [1x1 struct]

Amb I defining the ROI incorrectly??? Is it because the dimensions of 
the atlas and the source are not equal? Could anyone give a good example 
of how to do it?

Any suggestion will be appreciated. Thanks in advance.

-- 
Marc Recasens
Tel.: +34 639 24 15 98

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20110111/1c888429/attachment.html>