Advertisement

Deep Segmentation Refinement with Result-Dependent Learning

A Double U-Net for Hip Joint Segmentation in MRI
  • Duc Duy PhamEmail author
  • Gurbandurdy Dovletov
  • Sebastian Warwas
  • Stefan Landgraeber
  • Marcus Jäger
  • Josef Pauli
Conference paper
Part of the Informatik aktuell book series (INFORMAT)

Zusammenfassung

In this contribution, we propose a 2D deep segmentation refinement approach, that is inspired by the U-Net architecture and incorporates result-dependent loss adaptation. The performance of our method regarding segmentation quality is evaluated on the example of hip joint segmentation in T1-weighted MRI data sets. The results are compared to an ordinary U-Net implementation. While the segmentation quality of the proximal femur does not significantly change, our proposed method shows promising improvements for the segmentation of the pelvic bone complex, which shows more shape variability in the 2D image slices along the longitudinal axis.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Literatur

  1. 1.
    Chu C, Chen C, Liu L, et al. FACTS: fully automatic CT segmentation of a hip joint. Ann Biomed Eng. 2015;43(5):1247–1259.CrossRefGoogle Scholar
  2. 2.
    Xia Y, Fripp J, Chandra SS, et al. Automated bone segmentation from large field of view 3D MR images of the hip joint. Phys Med Biol. 2013;58(20):7375.CrossRefGoogle Scholar
  3. 3.
    Chandra SS, Xia Y, Engstrom C, et al. Focused shape models for hip joint segmentation in 3D magnetic resonance images. Med Image Anal. 2014;18(3):567–578.CrossRefGoogle Scholar
  4. 4.
    Klein A, Warszawski J, Hillengaß, et al. Towards whole-body CT bone segmentation. Proc BVM. 2018; p. 204–209.Google Scholar
  5. 5.
    Ronneberger O, Fischer P, Brox T; Springer. U-net: convolutional networks for biomedical image segmentation. Proc MICCAI. 2015; p. 234–241.Google Scholar
  6. 6.
    Ravishankar H, Venkataramani R, Thiruvenkadam S, et al.; Springer. Learning and incorporating shape models for semantic segmentation. Proc MICCAI. 2017; p. 203–211.Google Scholar
  7. 7.
    Newell A, Yang K, Deng J. Stacked hourglass networks for human pose estimation. Proc ECCV. 2016; p. 483–499.Google Scholar

Copyright information

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

Authors and Affiliations

  • Duc Duy Pham
    • 1
    Email author
  • Gurbandurdy Dovletov
    • 1
  • Sebastian Warwas
    • 2
  • Stefan Landgraeber
    • 2
  • Marcus Jäger
    • 2
  • Josef Pauli
    • 1
  1. 1.Intelligent Systems, Faculty of EngineeringUniversity of Duisburg-EssenDuisburgDeutschland
  2. 2.Department of Orthopedics and Trauma SurgeryUniversity Hospital Essen, University of Duisburg-EssenDuisburgDeutschland

Personalised recommendations