For VBM, I would suggest just following the instructions in Section 1.1 of
http://www.fil.ion.ucl.ac.uk/~john/misc/dartel_guide.pdf . If you want the
results in MNI space, then the easiest way to do this is according to the
suggestion in
http://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind0808&L=SPM&P=R16058
Best regards,
-John
On Monday 22 September 2008 15:42, Wenjing Li wrote:
> Dear All,
> During my data processing, I encountered some problems.
> I used VBM to analyse MRI scans of 111 subjects, which were
> divided into three groups: 26 healthy individuals, 54 individuals with
> left-TLE and 31 individuals with right-TLE.
> £Æ£é£ò£ó£ô of all£¬ £É £òeoriented the images to make the segmentation
> more accurate. After segmentation, I used the DARTEL tool for VBM. I
> created the template with the 26 healthy individuals, and then I warped
> every TLE individual to the existing template. In addition, I used DARTEL
> to spatialy normalize all the images to £Í£Î£É space. At last, I smoothed the
> processed images with a 8mm Gaussian smoothing kernels.
> In the statistical analysis, I tried to find the differences
> between the left-TLE individuals and healthy ones, or the right-TLE
> individuals and healthy ones. Here, I only take the former for example. I
> used the two-sample t-test, one group is the individuals with left-TLE, and
> the other group is the healthy individuals. I set the contrast as [1 0],
> and the p value as 0.05. Then, I got the result as seen in the enclosure. I
> find the result a little weird. From the upper picture, I thought there
> were lots of volumns superior to the threshold. In contrast, I saw only a
> few volumns from a whole brain as shown in the table below. What's more,
> when I set the contrst as [-1 0] , [1 -1] or other values, I cannot find
> any volumns superior to the threshold. I cannot explain why and I
> appreciate if you can help me!
>
> With regards!
>
>
> Wenjing.Li
|