Hello FSL-ers,
I am producing group maps but I think registration problems may be affecting the appearance of functional maps.
In posterior visual cortex, I frequently get a sharp cut-off in the map, i.e. activation cuts off in a sharp diagonal line. This makes for a messy figure, and unfortunately affects precisely those slices in which I have other areas of activation that I want to present. Dropping one or another subject whose orientation appears to be tilted (by eyeballing the "registration" results page) improves the situation somewhat but does not remove the problem. Further, this approach has the disadvantage of not allowing all subjects to contribute to the group activation map, changing the presentation of results.
I understand that FSL only includes those voxels in a group analyses which are shared by ALL subjects - how can I loosen this restriction?
Are there any other ways of remedying the problem of a few subjects' orientation impacting group maps? For instance, how can I identifying the specific subject(s) that are contributing to this problem? What approaches can I take to deal with these subjects w/o removing their data from the group analyses altogether?
Thanks for your help,
Anna-Maria
|