Latent3DU-net: Multi-level Latent Shape Space Constrained 3D U-net for Automatic Segmentation of the Proximal Femur from Radial MRI of the Hip

  • Guodong Zeng
  • Qian Wang
  • Till Lerch
  • Florian Schmaranzer
  • Moritz Tannast
  • Klaus Siebenrock
  • Guoyan ZhengEmail author
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 11046)


Radial 2D MRI scans of the hip are routinely used for the diagnosis of the cam-type of femoroacetabular impingement (FAI) and of avascular necrosis (AVN) of the femoral head, which are considered causes of hip joint osteoarthritis in young and active patients. However, for computer assisted planning of surgical treatment, it is highly desired to have 3D models of the proximal femur. In this paper, we propose a novel volumetric convolutional neural network (CNN) based framework to fully automatically extract 3D models of the proximal femur from sparsely hip radial slices. Our framework starts with a spatial transform to interpolate sparse 2D radial MR images to a densely sampled 3D volume data. Automated segmentation of the interpolated 3D volume data is very challenging due to the poor image quality and the interpolation artifact. To tackle these challenges, we introduce a multi-level latent shape space constrained 3D U-net, referred as Latent3DU-net, to incorporate prior shape knowledge into voxelwise semantic segmentation of the interpolated 3D volume. Comprehensive results obtained from 25 patient data demonstrated the effectiveness of the proposed framework.



This study was partially supported by the Swiss National Science Foundation via project 205321_163224/1.


  1. 1.
    Leunig, M., Beaule, P., Ganz, R.: The concept of femoroacetabular impingement: current status and future perspectives. Clin. Orthop. Relat. Res. 467, 616–622 (2009)CrossRefGoogle Scholar
  2. 2.
    Chughtai, M., Piuzzi, N.: An evidence-based guide to the treatment of osteonecrosis of the femoral head. Bone Joint J. 99(10), 1267–1279 (2017)CrossRefGoogle Scholar
  3. 3.
    Tannast, M., Kubiak-Langer, M.: Noninvasive three-dimensional assessment of femoroacetabular impingement. J. Orthop. Res. 25(1), 122–131 (2007)CrossRefGoogle Scholar
  4. 4.
    Xia, Y., Fripp, J.: Automated bone segmentation from large field of view 3d MR images of the hip joint. Phys. Med. Biol. 58(20), 7375–7390 (2013)CrossRefGoogle Scholar
  5. 5.
    Arezoomand, S., Lee, W.: A 3d active model framework for segmentation of proximal femur in MR images. Int. J. CARS 10(1), 55–66 (2015)CrossRefGoogle Scholar
  6. 6.
    Chandra, S.S., Xia, Y., et al.: Focused shape models for hip joint segmentation in 3d magnetic resonance images. Med. Image Anal. 18(3), 567–578 (2014)CrossRefGoogle Scholar
  7. 7.
    Çiçek, Ö., Abdulkadir, A., Lienkamp, S.S., Brox, T., Ronneberger, O.: 3D U-net: learning dense volumetric segmentation from sparse annotation. In: Ourselin, S., Joskowicz, L., Sabuncu, M.R., Unal, G., Wells, W. (eds.) MICCAI 2016. LNCS, vol. 9901, pp. 424–432. Springer, Cham (2016). Scholar
  8. 8.
    Dou, Q., Yu, L.: 3d deeply supervised network for automated segmentation of volumetric medical images. Med. Image Anal. 41, 40–54 (2017)CrossRefGoogle Scholar
  9. 9.
    Zeng, G., Yang, X., Li, J., Yu, L., Heng, P.-A., Zheng, G.: 3D U-net with multi-level deep supervision: fully automatic segmentation of proximal femur in 3D MR images. In: Wang, Q., Shi, Y., Suk, H.-I., Suzuki, K. (eds.) MLMI 2017. LNCS, vol. 10541, pp. 274–282. Springer, Cham (2017). Scholar
  10. 10.
    Van Assen, H., Danilouchkine, M.: Spasm: a 3d-ASM for segmentation of sparse and arbitrarily oriented cardiac MRI data. Med. Image Anal. 10(2), 286–303 (2006)CrossRefGoogle Scholar
  11. 11.
    Litjens, G., Kooi, T.: A survey on deep learning in medical image analysis. Med. Image Anal. 42, 60–88 (2017)CrossRefGoogle Scholar
  12. 12.
    Girdhar, R., Fouhey, D.F., Rodriguez, M., Gupta, A.: Learning a predictable and generative vector representation for objects. In: Leibe, B., Matas, J., Sebe, N., Welling, M. (eds.) ECCV 2016. LNCS, vol. 9910, pp. 484–499. Springer, Cham (2016). Scholar
  13. 13.
    Oktay, O., Kamnisas, K.: Anotomically constrained neural networks (ACNN): application to cardiac image enhancement and segmentation. IEEE Trans. Med. Imaging 37(2), 384–395 (2018)CrossRefGoogle Scholar
  14. 14.
    Sharma, A., Grau, O., Fritz, M.: VConv-DAE: deep volumetric shape learning without object labels. In: Hua, G., Jégou, H. (eds.) ECCV 2016. LNCS, vol. 9915, pp. 236–250. Springer, Cham (2016). Scholar
  15. 15.
    Ioffe, S., Szegedy, C.: Batch normalization: accelerating deep network training by reducing internal covariate shift. Proc. ICML 448–456 (2015)Google Scholar
  16. 16.
    Krizhevsky, A., Ilya, S., Hinton, G.: Imagenet classification with deep convolutional neural networks. Proc. NIPS 1097–1105 (2012)Google Scholar
  17. 17.
    Carr, J., Beatson, R., et al.: Reconstruction and representation of 3d objects with radial basis functions. Computer Graphics (2001) 67–76Google Scholar

Copyright information

© Springer Nature Switzerland AG 2018

Authors and Affiliations

  • Guodong Zeng
    • 1
  • Qian Wang
    • 2
  • Till Lerch
    • 3
  • Florian Schmaranzer
    • 3
  • Moritz Tannast
    • 3
  • Klaus Siebenrock
    • 3
  • Guoyan Zheng
    • 1
    Email author
  1. 1.Institute of Surgical Technology and Biomechanics, University of BernBernSwitzerland
  2. 2.School of Biomedical EngineeringShanghai Jiao Tong UniversityShanghaiChina
  3. 3.Department of Orthopaedic Surgery, InselspitalUniversity of BernBernSwitzerland

Personalised recommendations