Magnetic particle imaging is a novel imaging technique based on a linear reconstruction problem consisting of the computation of the magnetic particle distribution given the measured induced voltage. We formulate the reconstruction problem in a Bayesian framework and apply approximation error modeling in order to incorporate the uncertainties of the model in the reconstruction scheme. Numerical results will be presented to show the image quality improvement.

This presentation is part of Minisymposium “MS54 - Hybrid Approaches that Combine Deterministic and Statistical Regularization for Applied Inverse Problems (4 parts)”
organized by: Cristiana Sebu (University of Malta) , Taufiquar Khan (Clemson University) .

Authors:

Christina Brandt (University of Hamburg)

Keywords: