Meeting Banner
Abstract #4200

Advances in MRzero – supervised learning of parallel imaging sequences including joint non-Cartesian trajectory and flip angle optimization

Felix Glang1, Alexander Loktyushin1, Kai Herz1,2, Hoai Nam Dang3, Anagha Deshmane1, Simon Weinmüller3, Arnd Doerfler3, Andreas Maier4, Bernhard Schölkopf5, Klaus Scheffler1,2, and Moritz Zaiss1,3
1High-field Magnetic Resonance Center, Max-Planck Institute for Biological Cybernetics, Tübingen, Germany, 2Department of Biomedical Magnetic Resonance, Eberhard Karls University Tübingen, Tübingen, Germany, 3Friedrich-Alexander Universität Erlangen-Nürnberg (FAU), Neuroradiology, University Clinic Erlangen, Erlangen, Germany, 4Pattern Recognition Lab, Friedrich-Alexander-University Erlangen-Nürnberg, Erlangen, Germany, 5Empirical Inference, Max-Planck Institute for Intelligent Systems, Tübingen, Germany

Recently, MRzero has been proposed as a fully differentiable Bloch-equation-based MRI sequence invention framework. In this work, the approach is extended by parallel imaging capability, employing a CG SENSE reconstruction that allows optimizing for non-Cartesian sampling trajectories simultaneously with other sequence parameters like RF pulses and timings. The approach is tested herein by simulations on an in silico brain phantom and is found to yield improved reconstructions compared to regular Cartesian undersampling, and to simultaneously find variable flip angle patterns that compensate for transient signal induced blurring.

This abstract and the presentation materials are available to members only; a login is required.

Join Here