Hi Sabrina > Hi all, > I am doing a VBM analysis, and I would like to extract the significant peak > voxels (or sphere) I found in SPM, in order to do post-hoc analysis. > I tried with the script attached (a script done for extraction in fMRI), > and I tried with the eigenvariate option in SPM. > However, I don't find the same thing with the two... > Does someone have an idea about why? Or a better manner to extract in VBM > analysis? have you had a look at what your function does? help sjf_get_vals_spm8 tells you it computes the mean the eigenvariate option in SPM returns the 1st eigen value so yes it will be different -- if the data values in your ROI are close to normal both should give similar results - I tend to prefer eigen values as it is more stable ie more resistent to outlier values, the breakdown point of the mean is 0% my pwn tool -EasyROI returns both if you fancy trying it cyril -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.