thank you for your prompt response. 

I have a question regarding the function: I am confused about what vol_orig, vol_dest and vol_ind0 represent.

For instance, after calling the function on different transformations, vol_orig and vol_dest resulted to be precisely the same matrix (compared with the '==' Matlab operator). Moreover, the two volumes have values that increment in a linear way, as reported in the figure attached.

So they don't seem to represent the transformation itself, even if they have the right dimension 182x218x182x3. 

On the other hand, if vol_orig and vol_dest need to be subtracted in order to extract the warp field (i.e. they represent a kind of moving and fixed volumes), I would have a transformation with all zero values, since vol_orig==vol_dest. I wouldn't expect a warp field full of zero values, since the mri_cvs_register call successfully aligned the images in all the cases. In addition, vol_ind0 resulted to be a matrix of zeros. 

For completeness, I attached the log files for one mri_cvs_register execution.

Thank you very much for your attention and I look forward to hearing from you.

Best wishes,

Stefano Zappala'.