Predictive uncertainty in deep learning-based MR image reconstruction using deep ensembles: Evaluation on the fastMRI data set.

Journal: Magnetic resonance in medicine
Published Date: Jan 28, 2024

Abstract

PURPOSE: To estimate pixel-wise predictive uncertainty for deep learning-based MR image reconstruction and to examine the impact of domain shifts and architecture robustness.

Authors

  • Thomas Küstner
    Department of Radiology, Diagnostic and Interventional Radiology, Eberhard Karls University Tübingen, Germany.
  • Kerstin Hammernik
    Institute of Computer Graphics and Vision, Graz University of Technology, Graz, Austria.
  • Daniel Rueckert
    Biomedical Image Analysis (BioMedIA) Group, Department of Computing, Imperial College London, UK. Electronic address: d.rueckert@imperial.ac.uk.
  • Tobias Hepp
    University of Tübingen, Department of Radiology, Tübingen, Germany.
  • Sergios Gatidis
    Department of Radiology, Diagnostic and Interventional Radiology, Eberhard Karls University Tübingen, Germany.

Keywords

External Resources

Popular Topics
Recent Journals