Incorporating human and learned domain knowledge into training deep neural networks: A differentiable dose-volume histogram and adversarial inspired framework for generating Pareto optimal dose distributions in radiation therapy.

Journal: Medical physics
Published Date:

Abstract

PURPOSE: We propose a novel domain-specific loss, which is a differentiable loss function based on the dose-volume histogram (DVH), and combine it with an adversarial loss for the training of deep neural networks. In this study, we trained a neural network for generating Pareto optimal dose distributions, and evaluate the effects of the domain-specific loss on the model performance.

Authors

  • Dan Nguyen
    University of Massachusetts Chan Medical School, Worcester, Massachusetts.
  • Rafe McBeth
    Medical Artificial Intelligence and Automation (MAIA) Laboratory, Department of Radiation Oncology, UT Southwestern Medical Center, Dallas, TX, 75390, USA.
  • Azar Sadeghnejad Barkousaraie
    Medical Artificial Intelligence and Automation (MAIA) Laboratory, Department of Radiation Oncology, UT Southwestern Medical Center, Dallas, TX, 75390, USA.
  • Gyanendra Bohara
    Medical Artificial Intelligence and Automation (MAIA) Laboratory, Department of Radiation Oncology, UT Southwestern Medical Center, Dallas, TX, 75390, USA.
  • Chenyang Shen
    Department of Radiation Oncology, University of Texas Southwestern Medical Center, Dallas, TX 75002, USA.
  • Xun Jia
    Department of Radiation Oncology, University of Texas Southwestern Medical Center, Dallas, Texas 75235.
  • Steve Jiang