MFCA 2015 - 5th MICCAI workshop on Mathematical Foundations of Computational Anatomy

The goal of computational anatomy is to analyze and to statistically model the anatomy of organs in different subjects. Computational anatomic methods are generally based on the extraction of anatomical features or manifolds which are then statistically analyzed, often through a non-linear registration. There are nowadays a growing number of methods that can faithfully deal with the underlying biomechanical behavior of intra-subject deformations. However, it is more difficult to relate the anatomies of different subjects. In the absence of any justified physical model, diffeomorphisms provide the most general mathematical framework that enforce topological consistency. However, working with this infinite dimensional space raises some deep computational and mathematical problems, in particular for doing statistics. Likewise, modeling the variability of surfaces leads to rely on shape spaces that are much more complex than for curves. To cope with these, different methodological and computational frameworks have been proposed (e.g. smooth left-invariant metrics, focus on well-behaved subspaces of diffeomorphisms, modeling surfaces using currents, etc.) The goal of the workshop is to foster interactions between researchers investigating the combination of geometry and statistics in non-linear image and surface registration in the context of computational anatomy from different points of view. A special emphasis will be put on theoretical developments, applications and results being welcomed as illustrations.
Workshop format and topics
The program will be composed of oral presentations selected by the peer-reviewed contributions of the participants. To foster interactions, a large amount of time will be reserved for discussions after each presentation. Contributions are solicited in (but not limited to) the areas of:
Riemannian sub-Riemannian and group theoretical methods
Statistical models for manifold-valued data, including surfaces, deformations and shapes
Metrics for computational anatomy
Statistics of surfaces
Time-evolving geometric processes
Stratified spaces
Optimal transport in registration problems
Approximation methods in statistical learning (e.g. variational Bayes, importance sampling, Monte Carlo methods)