I am no expert, but if I may make a small suggestion based on the image you have provided? I don't know exactly what you're comparing in your VBM, but if indeed there is on average a difference in *shape* of the cortex between the groups you are comparing, then it is not at all surprising that the areas of significant VBM t- statistic are relatively far from the average surface.
If you can interpret the VBM results as saying that one group is more likely to have cortex in the significant area than the other group, then purely for the sake of what you're trying to do here, you might want to generate an average surface from only the group which is *more* likely to have cortex in the area of significance. Then you *would* expect that average surface to overlap the area of significance.
Then you could project the volume t-stat map onto *that* average surface for the purpose of getting as you say "a rough approximation" of the MNI space results projected onto a surface.
Hope this is of some small benefit...
On Jun 1, 2007, at 11:47 AM, Doug Greve wrote:
And it does not intersect the other hemi, I assume? I don't think there's really anthing you can do about it.
doug
-- -dave---------------------------------------------------------------- All I ask is that the kind of unsolvable that it turns out to be has respectable precedents. -Jerry Fodor