Artificially-generated consolidations and balanced augmentation increase performance of U-net for lung parenchyma segmentation on MR images.

Journal: PloS one
PMID: 37159468

Abstract

PURPOSE: To improve automated lung segmentation on 2D lung MR images using balanced augmentation and artificially-generated consolidations for training of a convolutional neural network (CNN).

Authors

  • Cristian Crisosto
    Institute for Diagnostic and Interventional Radiology, Hannover Medical School (MHH), Hannover, Germany.
  • Andreas Voskrebenzev
    Institute for Diagnostic and Interventional Radiology, Hannover Medical School (MHH), Hannover, Germany.
  • Marcel Gutberlet
    Institute for Diagnostic and Interventional Radiology, Hannover Medical School (MHH), Hannover, Germany.
  • Filip Klimeš
    Institute for Diagnostic and Interventional Radiology, Hannover Medical School (MHH), Hannover, Germany.
  • Till F Kaireit
    Institute for Diagnostic and Interventional Radiology, Hannover Medical School (MHH), Hannover, Germany.
  • Gesa Pöhler
    Institute for Diagnostic and Interventional Radiology, Hannover Medical School (MHH), Hannover, Germany.
  • Tawfik Moher
    Institute for Diagnostic and Interventional Radiology, Hannover Medical School (MHH), Hannover, Germany.
  • Lea Behrendt
    Institute for Diagnostic and Interventional Radiology, Hannover Medical School (MHH), Hannover, Germany.
  • Robin Müller
    Institute for Diagnostic and Interventional Radiology, Hannover Medical School (MHH), Hannover, Germany.
  • Maximilian Zubke
    Institute for Diagnostic and Interventional Radiology, Hannover Medical School (MHH), Hannover, Germany.
  • Frank Wacker
  • Jens Vogel-Claussen
    Institute of Diagnostic and Interventional Radiology Hannover Medical School Hannover Germany.

Keywords

External Resources

Popular Topics
Recent Journals