Advertisement

Improved X-Ray Bone Segmentation by Normalization and Augmentation Strategies

  • Florian KordonEmail author
  • Ruxandra Lasowski
  • Benedict Swartman
  • Jochen Franke
  • Peter Fischer
  • Holger Kunze
Conference paper
Part of the Informatik aktuell book series (INFORMAT)

Zusammenfassung

X-ray images can show great variation in contrast and noise levels. In addition, important subject structures might be superimposed with surgical tools and implants. As medical image datasets tend to be of small size, these image characteristics are often under-represented. For the task of automated, learning-based segmentation of bone structures, this may lead to poor generalization towards unseen images and consequently limits practical application. In this work, we employ various data augmentation techniques that address X-ray-specific image characteristics and evaluate them on lateral projections of the femur bone. We combine those with data and feature normalization strategies that could prove beneficial to this domain. We show that instance normalization is a viable alternative to batch normalization and demonstrate that contrast scaling and the overlay of surgical tools and implants in the image domain can boost the representational capacity of available image data. By employing our best strategy, we can improve the average symmetric surface distance measure by 36:22 %.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Literatur

  1. 1.
    Huda W, Abrahams RB. Radiographic techniques, contrast, and noise in x-ray imaging. Am J Roentgenol. 2015;204(2):W126-W131.CrossRefGoogle Scholar
  2. 2.
    Garcia-Garcia A, Orts-Escolano S, Oprea S, et al. A review on deep learning techniques applied to semantic segmentation. arXiv:170406857. 2017;.
  3. 3.
    Unberath M, Zaech JN, Lee SC, et al. DeepDRR: a catalyst for machine learning in uoroscopy-guided procedures. arXiv:180308606. 2018;.
  4. 4.
    Ronneberger O, Fischer P, Brox T. U-Net: convolutional networks for biomedical image segmentation. Proc MICCAI. 2015;9351:234-241.Google Scholar
  5. 5.
    Ioe S, Szegedy C. Batch normalization: accelerating deep network training by reducing internal covariate shift. Proc Int Conf Mach Learn. 2015;37:448-456.Google Scholar
  6. 6.
    Ulyanov D, Vedaldi A, Lempitsky VS. Instance normalization: the missing ingredient for fast stylization. arXiv:160708022. 2016;.
  7. 7.
    Ba JL, Kiros JR, Hinton GE. Layer normalization. arXiv:160706450. 2016;.
  8. 8.
    Wu Y, He K. Group normalization. arXiv:180308494. 2018;.
  9. 9.
    Luo P, Ren J, Peng Z. Differentiable learning-to-normalize via switchable normalization. arXiv:180610779. 2018;.

Copyright information

© Springer Fachmedien Wiesbaden GmbH, ein Teil von Springer Nature 2019

Authors and Affiliations

  • Florian Kordon
    • 1
    • 3
    Email author
  • Ruxandra Lasowski
    • 1
  • Benedict Swartman
    • 2
  • Jochen Franke
    • 2
  • Peter Fischer
    • 3
  • Holger Kunze
    • 3
  1. 1.Faculty of Digital MediaFurtwangen University (HFU)SchwarzwaldDeutschland
  2. 2.Department for Trauma and Orthopaedic SurgeryBG Trauma Center LudwigshafenLudwigshafenDeutschland
  3. 3.Advanced TherapiesSiemens Healthineers ForchheimForchheimDeutschland

Personalised recommendations