Have you tried MarsBaR? (Its a toolbox for SPM) http://marsbar.sourceforge.net/ It includes predefined ROIs for a large number of brain structures (with the ability to define your own). I think you can easily make mask images from these. Joel Yu, James CL wrote: >Does anyone know or have detailed brain atlas mask images (maybe in >Talairach space) that can be used directly in SPM? > >What I would need it for is to calculate the ROI value in a spatially >normalized PET image for each brain structure by using these mask images. > >Thanks for any help. > >