Dear SPMs,
I'm a relatively new user of the SPM and I have tried to explain myself hoe
the numbers from the SPMmip results graphic window are computed. I have get
some answers but I still can not explain me how the FWHM values are
computed . I have tried the formula FWHM_i=\sqrt(f_i^2 + f^2) where f =FWHM
of the kernel that we use for smoothing and f_i the FWHM in i direction of
the original field after reconstruction but I can not get consistent
results.
For example for a bypass smoothing (I call smooth level f=0) I get (6.6 6.8
3.8)
for a smoothing with f=5 then i get (7.6 7.7 6.0)
for a smoothing with f= 8 then i get (9.0 9.1 8.5)
for a smoothing with f=10 then i get (10.1 10.3 10.3)
for a smoothing with f=15 then i get (13.8 14.0 15.5)
for a smoothing with f=20 then i get (17.1 17.5 20.5)
How is possible to have the value 20.5 >20?? for this large smoothing
kernel?
The second questions is about the number of resels within the search volume:
For example S=2384753mm^3=162503 voxels, 1 resel=47.17 voxels but S=3229
resels not 3445 that are expected. Why is this difference? How do you
compute the number of the voxels within aresel and the number of the resels
within the search volume? If I divide S by the product of the FWHM_i i get
3485 resels in S thus another value.
The third question is about which is the explication of the two p-valus
that are in the bottom of the page, near the extend threshold in voxels.
The forth question is how they are find the number of clusters (for example
c=10) in the search volume and the number of voxels within each cluster ,
the k_E?
I know that my questions are very basics but I would like to clarify these
for me.
I hope that you will help me,
Yours sincerely,
Catalina
|