Minisymposia Abstracts

Speaker: Martin Benning
University of Munster, Germany
Title: A Linearized Model for Dynamic Positron Emission Tomography
Abstract: Dynamic Positron Emission Tomography (PET) allows monitoring physiological processes within the body that can be described by kinetic parameters. However, recovery of these parameters often requires the solution of complex and nonlinear operator equations. In this talk, a method for the quantification of myocardial perfusion is introduced. The proposed method is a two-step procedure consisting of a nonlinear 4D PET reconstruction and a subsequent post-processing of the reconstructed PET sequence, in order to quantify the desired parameters.

Speaker: George Biros
Georgia Institute of Technology
Title: Adaptive discretizations for the time harmonic inverse medium problem
Abstract: TBA

Speaker: Herbert Egger
Rice University
Title: On the multilevel iterative solution of inverse problems
Abstract: In this talk, we discuss multilevel concepts for the efficient numerical solution of inverse problems.
We start with a short review of some basic approaches, including cascadic multilevel methods and additive respectively multiplicative Schwarz methods, which have been proposed and analyzed in the last couple of years.
Based on a simple model problem, different algorithms, their theoretical promises and limitations, are compared. The basic ingredients for a successful application of multilevel ideas are elaborated, and comments on generalizations and possible improvements are given.
In the third part of the presentation, the performance of multilevel algorithms is demonstrated numerically by application to tomographic imaging problems.

Speaker: Thorsten Hohage
University of Gottingen, Germany
Title: Large-scale inverse problems in photonic imaging
Abstract: TBA

Speaker: Youssef Marzouk
Title: High-dimensional statistical inverse problems without Markov chains
Abstract: TBA

Speaker: John Schotland
University of Michigan, USA
Title: Inverse transport with large data sets
Abstract: TBA

Speaker: Christoph Schwarzbach
University of British Columbia, Canada
Title: Inversion of electromagnetic data collected using a large number of sources
Abstract: First, in order to reduce the number of degrees of freedom of the forward problem we use a finite element formulation that allows the use of semi-structured (OcTree) and unstructured grids. Second, in order to solve the problem with many sources/receivers we use stochastic optimization techniques that allow us to work only with a subset of source/receiver combinations at each iteration. This reduces the number of forward problems to be solved. Combining these techniques we are able to solve electromagnetic inverse problems that are traditionally solved using intensive computational platforms on modest ones.

Speaker: Georg Stadler
University of Texas at Austin, USA
Title: Global seismic full waveform inversion with quantified uncertainties
Abstract: TBA

Please address administrative questions to Scientific questions should be addressed to the chair of the Scientific Program Committee: rundell AT

Copyright © 2010, Texas A&M University, Department of Mathematics, All Rights Reserved.