Hi -pt,

Sorry to come back with this old topic but my new problem is actually related to the co-registration of these animals.
By having a deep look at the co-registrated functional images and extracting some signal with the D99 atlas, I finally understand why the correlation coefficients were not great.

It seems that a slight section of voxels in the grey matter (edge) of the functional image is not really superimposed on the anatomical image (see picture).
That obviously decreases the accuracy of the signal extracted.

1. I checked on the raw image that the grey matter of the anatomical image is well superimposed on the functional (except the area of dropout)
2. I try to fix the problem by decreasing the resolution to 0.5 with -volreg_warp_dxyz 0.5: it doesn't work
3. I try lpa+zz and doesn't help too
Any other ideas how to fix that?

Image of the anat on func:
[ibb.co]

thanks a lot!
Clément