Comparison of Deep Learning-Based and Patch-Based Methods for Pseudo-CT Generation in MRI-Based Prostate Dose Planning.

Journal: International journal of radiation oncology, biology, physics
PMID: 31505245

Abstract

PURPOSE: Deep learning methods (DLMs) have recently been proposed to generate pseudo-CT (pCT) for magnetic resonance imaging (MRI) based dose planning. This study aims to evaluate and compare DLMs (U-Net and generative adversarial network [GAN]) using various loss functions (L2, single-scale perceptual loss [PL], multiscale PL, weighted multiscale PL) and a patch-based method (PBM).

Authors

  • Axel Largent
    Univ Rennes, CLCC Eugène Marquis, INSERM, LTSI - UMR 1099, F-35000 Rennes, France. Electronic address: axel.largent@hotmail.fr.
  • Anaïs Barateau
    Univ Rennes, CLCC Eugène Marquis, INSERM, LTSI - UMR 1099, F-35000 Rennes, France.
  • Jean-Claude Nunes
    Univ Rennes, CLCC Eugène Marquis, INSERM, LTSI - UMR 1099, F-35000 Rennes, France.
  • Eugenia Mylona
    Univ Rennes, CLCC Eugène Marquis, INSERM, LTSI - UMR 1099, F-35000 Rennes, France.
  • Joël Castelli
    Univ Rennes, CLCC Eugène Marquis, INSERM, LTSI - UMR 1099, F-35000 Rennes, France.
  • Caroline Lafond
    Univ Rennes, CLCC Eugène Marquis, INSERM, LTSI - UMR 1099, F-35000 Rennes, France.
  • Peter B Greer
    School of Mathematical and Physical Sciences, University of Newcastle, Newcastle, Australia; Department of Radiation Oncology, Calvary Mater, Newcastle, Australia.
  • Jason A Dowling
    Australian e-Health Research Centre, CSIRO, Brisbane, QLD, 4029, Australia.
  • John Baxter
    Univ Rennes, CLCC Eugène Marquis, INSERM, LTSI - UMR 1099, F-35000 Rennes, France.
  • Hervé Saint-Jalmes
    Univ Rennes, CLCC Eugène Marquis, INSERM, LTSI - UMR 1099, F-35000 Rennes, France.
  • Oscar Acosta
    Univ Rennes, CLCC Eugène Marquis, INSERM, LTSI - UMR 1099, F-35000 Rennes, France.
  • Renaud de Crevoisier
    Univ Rennes, CLCC Eugène Marquis, INSERM, LTSI - UMR 1099, F-35000 Rennes, France.

Keywords

External Resources

Popular Topics
Recent Journals