Dear Shail,
as patients vs. controls seems to be unaffected this points to a problem specific to the young healthy, for whatever reason. What about patients vs. young healthy in VBM5 and VBM8?
What exactly do you mean by "truncated", maybe you can upload some pictures showing a representative "truncated" GM volume and a "normal" one, or maybe an unsmoothed mean of the 20 young subjects and the 20 elderly, once in VBM5 and once in VBM8 (you can create means of e.g. the 20 m0wp1* volumes easily with ImCalc, use mean(X) as expression, set "Data Matrix" to yes). Should help to get an idea of your specific problem (whether algorithms went wrong or not...).
Best,
Helmut
|