Hello,

I am trying to create a registration of a template brain, generated from an average of many brains (http://imgur.com/a/CyBSQ#0), to a template brain consisting of just one single brain (http://imgur.com/a/CyBSQ#1). While I have successfully done this before, this particular averaged brain appears rather blurred, and this seems to be messing up the registration near the edges (http://imgur.com/a/CyBSQ#2). Does anyone have any suggestions on how I can improve on this? I have tried blurring the single brain such that it looks, by eye, similar to the averaged brain and, while this did change the end result, the differences were not marked. The non-rigid registration definitely has enough 'play' to correct the displacement, as much larger displacements in other regions of the brain are successfully fixed.

Thanks,
Ashley