Dear SPMers
I'm doing a research related to brain asymmetry.
However, I've got non-significant results which
were different from the previous literatures.
I hope someone would kindly have a look for the
parameters and give some comment to me
if there is anything wrong:
17 subjects (omit the gender and handedness factors)
flipped images were compared with unflipped one
optimized VBM protocol
smoothing filter 2, 4, 8 and 12mm
In spm analysis,
PET/SPECT model
Compare-populations: 1scan/subject (two sample t-test)
group 1 select all the flipped images
group 2 select all the unflipped images
Proportional Scaling
ProSca global mean: 100
Threshold masking: proportional
analysis threshold: 0.8
global scale: user specfied and use the results from
the previous global gray matter volume
Results,
t-contrast
flip>unflip 1 -1 0 as contrast
mask with other contrast(s): no
correct with height threshold’ no
threshold (T or p value): 0.001
& extent threshold {voxel}: 0
The resulting graphs include very small significant clusters
which may not related to common areas of brain asymmetry.
S.W. LAM
|