Dear SPMers,
I am currently analyzing some data from a PET study with the SPM2
programme. The aim is to find out whether there are any relations between
these PET data and neuropsychological data or not. I have a big problem
with the smoothing function I have to describe a little bit in detail.
First I normalized all the PET pictures and then I smoothed them with an
eight millimetre filter. After this I did the simple regression analysis
using the neuropsychological data as covariates and finally the design
looked like this:
simple regression (correlation)
Covariate: 12,13,8,20,15
Covariate name: XYZ
scaling of overall grand mean
scale overall grand mean to: 50
threshold masking: relative and 0.8
implicit mask: yes
explicitly mask images: no
mean voxel value
After this I used the estimation function.
In the result section I defined the contrast „1“, the threshold 0,02 and
the voxel size is 100.
In the result file at the end the named filter size used in the beginning
is completely different from eight millimetres. It ranges from 60.000 to
80.000 millimetres and is changing with the used covariates.
What did I wrong? Maybe my problem is too simple but I am a new SPM user. I
would be very thankful to everyone who maybe can help me.
Best regards, Cathleen
|