 |
 |
First of all, I would suggest using only "modulated" data, as these are easier to interpret in terms of the volume of GM within each region. I would suggest looking at the results of the segmentation to see how well the algorithm deals with your MR data. Perhaps you are seeing some non-brain structure classified more strongly as GM in males. With lots of subjects, an SPM analysis will be very sensitive to even the smallest systematic differences between the groups. Best regards, -John On Wed, 2009-10-21 at 21:17 +0900, Yasuhiro Kawasaki wrote: > Dear SPMers, > > Recently I am conducting VBM group analysis using DARTEL. Subjects > consist of 4 groups of female control, male control, female patient, > and male patient. Group comparison between male and female group using > modulated images with ICV covariates and un-modulated images showed > sillier result. As shown in the attached results, male subjects tend > to have more caudal brain mass compared to female subjects. This > finding seems to me a kind of bias or distortion during the > normalization process. How can I get rid of this artifact? > > Yasuhiro, > > ============================================ > > Yasuhiro Kawasaki, M.D., Ph.D. > > Department of Neuropsychiatry > > University of Toyama, School of Medicine > > 2630 Sugitani, Toyama 930-0194, Japan > > TEL: +81-76-434-7322 > > FAX: +81-76-434-5030 > > e-mail: [log in to unmask] > > ============================================ > > > > -- John Ashburner <[log in to unmask]>