The Diffeo Microscope: Looking Through Scales in Computational Anatomy

     Current descriptions of brain diseases usually need to put together several order of magnitude ranging from the millimeter scale for tissues in standard imaging devices to the micron or even nano scale for neural cells and molecules.

     Organizing these representations within a given patient or between a population for statistical modelling and understanding could be quite helpful but very challenging. In this talk, I will present an on-going attempt to make a small step in that direction in the context of computational anatomy based on two key assets: 1- Layered coarse-to-fine diffeomorphic transport based on idea coming from optimal control and riemannian geometry 2- Varifold based representations of information and layer to layer reduction. Work in collaboration with M.I. Miller and D. Tward.