[FieldTrip] Problem averaging sources ("vary in the field 'inside'")

[Tzvetan Popov]

Peace T,
this is due to the unnecessary interpolating and normalizing back and forth where due to some edge irregularities you end up with slightly different inside fields. Using mni aligned grids is actually meant to circumvent the need for normalizing etc.
In general you should use ft_sourceinterpolate in order to visualize 1). single subject data (this worked for you) 2). output data from ft_sourcegrandaverage or 3) output from ft_sourcestatistics. Because you are using the mni template approach you’ll never use ft_volumenormalize in your pipeline. So in order to make 2) and 3) work you should exchange the single subject source.dim and source.pos fields with template_grid.dim and template_grid.pos respectively. Subsequently you can call ft_sourcegrandaverage with all your subjects as an input and use the output to interpolate onto the mni template brain. 

good luck
tzvetan



> Dear fieldtrippers,
> 
> We are doing beamformer with ERF acquired with an Elekta-Neuromag. Regardless our different approaches, we always end with the same problem at the end of the pipeline - the 'inside' field slightly varies between subjects. I attach a plot of the difference between the inside field of two different subjects where one can see that there are truly minor differences.
> 
> We follow the standard pipeline, I think. We use a template grid and template mri for all the subjects, which we realign using individual headpoints. The resulting leadfield (every subject has the same numbers of dipoles) is used in the sourceanalysis ('lcmv'), and the difference between pre and post periods is first interpolated (using the individual mri aligned) and finally normalized. We also tried using volumereslice before interpolate.
> 
> Individual results seem coherent but not matter what we end up with different 'inside' fields and sourcegrandaverage complains about that. We are using the fieldtrip-20141201 version and if we use the fieldtrip-20111101 sourcegrandaverage does not complain about that.
> 
> I don't want to make a really complex mail so I leave it here looking forward to hear input from anybody and willing to give more precise information,
> 
> Peace!
> T
> 
> 
> 
> -- 
> Constantino Méndez-Bértolo
> Laboratorio de Neurociencia Clínica, Centro de Tecnología Biomédica (CTB)
> Parque Científico y Tecnológico de la UPM, Campus de Montegancedo
> 28223 Pozuelo de Alarcón, Madrid, SPAIN
> 
> 
> <s5.inside minus s6.inside.jpg>_______________________________________________
> 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/20141203/ca222376/attachment-0002.html>