Dear Volkmar,
Thanks for your answer.
N=5 is indeed not a very large number. I'm performing a test fMRI experiment in songbirds (no GM + WM masks possible to rerun the analysis with). We have done similar pilot experiments with a small number of animals in the past and never encountered this problem. In addition I scanned the animals twice and in the paired test (10 images are loaded) I have the same problem.
How can I calculate the FWHM and Resel count that I should expect after preprocessing? Currently estimated FWHM and Resel count at single subject 1st Level and 2nd Lever are fairly similar to eachother and to what I usually have in experiments with more animals where similar analysis seem to work fine.
Do you have any other ideas about what might cause this problem?
Can someone help me with my second issue mentioned in my previous email?:
2- I always acquire multi-slice sagittal images with adjacent slices which I normalize to a 3D atlas and smooth (2voxels). In the result display, my SPM’s show white lines in
between the slices. As if there is a slice gap (see picture). In contrast, the CON and beta files look fine. Does this point to a problem with the data?
Best regards
Lisbeth Van Ruijssevelt
Lisbeth Van Ruijssevelt
PhD student
Faculty of FBD /Department of Biomedical Sciences
Campus Drie Eiken– Building Uc 1.18
Universiteitsplein 1 – B-2610 Wilrijk
[log in to unmask]
P +32 3 265 27 96
https://www.uantwerpen.be/en/rg/bio-imaging-lab/
-----Original Message-----
From: SPM (Statistical Parametric Mapping) [mailto:[log in to unmask]] On Behalf Of Volkmar Glauche
Sent: donderdag 3 september 2015 16:03
To: [log in to unmask]
Subject: Re: [SPM] visualization issues
Dear Lisbeth,
N=5 is a very small number for RFX group analysis, and you may run into problems estimating the smoothness of your residuals or the error variance. This would explain why the con image looks fine, but the stats image does not. Have a look at the bottom of your results table. Estimated FWHM and Resel count should be reasonably close to what you would expect after preprocessing. If the voxels are outside the brain it would also help to include an explicit GM+WM mask in your model specification and rerun model estimation.
Hope this helps,
Volkmar
|