Biomedical Engineering Reference
In-Depth Information
registration methods on functional data is still unknown. As a matter of fact,
since these methods are dedicated to anatomical and functional normalization,
it would be interesting to know how much of the intersubject functional vari-
ability can be understood and compensated by anatomical non-rigid registra-
tion. This is a challenging research subject that requires a better knowledge
about the relationship between the anatomy of the brain and its functional
organization.
Questions
1.
What are photometric and geometric registration methods? How can these
methods be compared?
2.
What is optical-flow?
3.
What is the aperture problem? How can it be solved?
4.
What are the advantages and drawbacks of optical-flow based registration?
5.
What are robust estimators?
6.
What are the advantages and drawbacks of robust M-estimators?
7.
How useful is a multiresolution scheme?
8.
How should the Gaussian standard deviation be chosen for building the
multiresolution pyramid?
9.
What is a multigrid optimization scheme?
10.
What are the different options to regularize the deformation field?
