[FieldTrip] UPDATE: Artifacts on plotting grand-averaged BEMs: Normalization issue?

[Gregor Volberg]

Dear list members, 

I think I fixed the problem described in my last mail. I manually averaged the avg.pow fields of the single subjects and put the average on the corresponding field of the grandaverage-structure. This worked fine, without artifactual borders at the outline of the brain compartement. 

Best regards, 
Gregor 



-- 
Dr. rer. nat. Gregor Volberg <gregor.volberg at psychologie.uni-regensburg.de> ( mailto:gregor.volberg at psychologie.uni-regensburg.de )
University of Regensburg
Institute for Experimental Psychology
93040 Regensburg, Germany
Tel: +49 941 943 3862 
Fax: +49 941 943 3233
http://www.psychologie.uni-regensburg.de/Greenlee/team/volberg/volberg.html


>>> "Gregor Volberg" <Gregor.Volberg at psychologie.uni-regensburg.de> 2/15/2011 5:16 PM >>>

Dear list members, 


I encountered  a problem when plotting the results of a beamformer analysis with individual BEM headmodels. For each subject, I did an ft_sourceanalysis (DICS) and then interpolated the individual volume on the individual anatomy with ft_sourceinterpolate. The result was normalised onto the SPM8 template brain with ft_volumenormalise; then I computed the grand average with ft_sourcegrandaverage.  


When I plot the grand average (with cfg.funparameter = 'pow'), I see nice results at expected brain regions, but also artifacts around the outline of the volume  - see my attached Figure "source_power.png". It appears as if normalization was imperfect so that a border remains, possibly showing idiosyncratic activations from single subjects. The border does not show up if, with the same data, I do ft_sourcestatistics and then plot the results with cfg.funparameter = 'stat' (see attached "source_stat.png"). One might expect this if the voxels showed activation of only one subject so that a t-test would return NANs. 
  
I then had a look into the "*.inside" fields of the normalized volumes and was surprised to see that number of voxels inside the brain was different for each subject (ranging between 23 and 27 % roughly). Should that not be the same number of voxels for each subject after normalization? I am now unsure whether the border is a known behavior when plottig the grand average over individual heads,or whether I made an error with normalization. I would be very happy if any someone had an advice for me. 


Many thanks and best regards, 
Gregor 

-- 
Dr. rer. nat. Gregor Volberg <gregor.volberg at psychologie.uni-regensburg.de> ( mailto:gregor.volberg at psychologie.uni-regensburg.de )
University of Regensburg
Institute for Experimental Psychology
93040 Regensburg, Germany
Tel: +49 941 943 3862 
Fax: +49 941 943 3233
http://www.psychologie.uni-regensburg.de/Greenlee/team/volberg/volberg.html


 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20110216/9c205778/attachment-0002.html>