Schematic
illustrating a Gauss-Newton scheme for maximizing the posterior probability _{}of the parameters
required to spatially normalize an image.
This scheme is iterative. At each
step the conditional estimate of the parameters is obtained by jointly
minimizing the likelihood and the prior potentials. The former is the difference between a
resampled (*i.e.* warped) version *y *of the image *f *and the best linear combination of some templates *g*.
These parameters are used to mix the templates and resample the image to
progressively reduce both the spatial and intensity differences. After convergence the resampled image can be
considered normalized.