 |
 |
Dear Henrike This question is true of any activation, but my guess is that you are concerned because it is relatively small and near an edge. I suggest 1) look at the quality of your overall data. Do you have a lot of activations around edges? 2) look at the activations in individual subjects. Is this activation consistent. 3) Overlay the activation on your functional images. Is it in the correct place? Also you can look at the SNR of your functional images to see if the activation is in a location where there is sufficient signal to expect it to be real. For example see this paper: Small, D.M., Gregory, M.D., Mak, Y.E., Gitelman, D., Mesulam, M.M., Parrish, T., 2003. Dissociation of neural representation of intensity and affective valuation in human gustation. Neuron 39, 701-711. take care, ----- Darren Gitelman On Fri, Sep 5, 2008 at 2:34 AM, Henrike Hemingway <[log in to unmask]> wrote: > Dear Guillaume > > thank you for your answer. > > One question remains, maybe you can help as well: I wonder whether the > activated cluster we find (that is, amygdala activation) is "real" or not. > It seems to be right at the border of the brain mask, but that was expected. > Is there anything specific I have additionally to look for when reporting > such activation clusters, just to be convincing? > > Best wishes > Henrike > > > -------- Original-Nachricht -------- > > Datum: Thu, 04 Sep 2008 17:06:02 +0100 > > Von: Guillaume Flandin <[log in to unmask]> > > An: Henrike Hemingway <[log in to unmask]> > > CC: [log in to unmask] > > Betreff: Re: [SPM] NaNs > > > Dear Henrike, > > > > Perseverance is good. > > What you observe is very likely due to the interpolation used when > > displaying images. If you change it to Nearest Neighbours > > (Interpolation>NN) instead of Bilinear, then you should only observe NaN > > values outside the brain mask. > > > > (You get a different behaviour between the beta/con/ResMS and the T > > images because the former use NaNs outside the mask (implicit masking) > > while the latter use 0). > > > > Best regards, > > Guillaume. > > > > -- > GMX startet ShortView.de. Hier findest Du Leute mit Deinen Interessen! > Jetzt dabei sein: > http://www.shortview.de/wasistshortview.php?mc=sv_ext_mf@gmx >