Dear SPM experts,
I've run VBM8 using a customized DARTEL-template. Then, I did sample homogeneity check and identified some poor images (i.e. images with severe motion artefacts, enlarged ventricles, etc.). Among 296 images in total, I found 11 "poor" images so far, but there seem to be some more of them.
Now, my question is which option I should choose between the following.
(-----1-----) Start statistical analysis without those outliers
(-----2-----) Start again from the beginning (i.e. re-run VBM8) without those outliers
I was originally going to go for option 2 because, to my understanding, a customized DARTEL-template is made from my own images, and therefore outliers would cause bias in the template. However, I'm not sure whether this reasoning is correct.
So, could anyone who understands how VBM8 works please advise me on this? I would really appreciate your help.
Best wishes,
Kie Woo Nam
___________________________
Kie Woo Nam, PhD student
Division of Psychological Medicine
Institute of Psychiatry, King's College London
16 De Crespigny Park, Denmark Hill
London, SE5 8AF, UK
|