Advertisement

Convolutional Neural Network and In-Painting Techniques for the Automatic Assessment of Scoliotic Spine Surgery from Biplanar Radiographs

  • B. AubertEmail author
  • P. A. Vidal
  • S. Parent
  • T. Cresson
  • C. Vazquez
  • J. De Guise
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 10434)

Abstract

Assessing the effectiveness of scoliosis surgery requires the quantification of 3D spinal deformities from pre- and post-operative radiographs. This can be achieved from 3D reconstructed models of the spine but a fast-automatic method to recover this model from pre- and post-operative radiographs remains a challenge. For example, the vertebrae’s visibility varies considerably and large metallic objects occlude important landmarks in postoperative radiographs. This paper presents a method for automatic 3D spine reconstruction from pre- and post-operative calibrated biplanar radiographs. We fitted a statistical shape model of the spine to images by using a 3D/2D registration based on convolutional neural networks. The metallic structures in postoperative radiographs were detected and removed using an image in-painting method to improve the performance of vertebrae registration. We applied the method to a set of 38 operated patients and clinical parameters were computed (such as the Cobb and kyphosis/lordosis angles, and vertebral axial rotations) from the pre- and post-operative 3D reconstructions. Compared to manual annotations, the proposed automatic method provided values with a mean absolute error <5.6° and <6.8° for clinical angles; <1.5 mm and <2.3 mm for vertebra locations; and <4.5° and <3.7° for vertebra orientations, respectively for pre- and post-operative times. The fast-automatic 3D reconstruction from pre- and post in-painted images provided a relevant set of parameters to assess the spine surgery without any human intervention.

References

  1. 1.
    Ilharreborde, B., et al.: Use of EOS imaging for the assessment of scoliosis deformities: application to postoperative 3D quantitative analysis of the trunk. Eur. Spine J. 23(Suppl. 4), 97–405 (2014)Google Scholar
  2. 2.
    Humbert, L., et al.: 3D reconstruction of the spine from biplanar X-rays using parametric models based on transversal and longitudinal inferences. Med. Eng. Phys. 31, 681–687 (2009)CrossRefGoogle Scholar
  3. 3.
    Ilharreborde, B., et al.: Angle measurement reproducibility using EOS three-dimensional reconstructions in adolescent idiopathic scoliosis treated by posterior instrumentation. Spine 36(20), E1306–E1313 (2011)CrossRefGoogle Scholar
  4. 4.
    Lecron, F., et al.: Fast 3D spine reconstruction of postoperative patients using a multilevel statistical model. MICCAI 15, 446–453 (2012)Google Scholar
  5. 5.
    Aubert, B., et al.: Automatic spine and pelvis detection in frontal X-rays using deep neural networks for patch displacement learning. In: IEEE ISBI, June 2016, pp. 1426–1429 (2016)Google Scholar
  6. 6.
    Kadoury, S., et al.: Postoperative 3D spine reconstruction by navigating partitioning manifolds. Med. Phys. 43(3), 1045–1056 (2016)CrossRefGoogle Scholar
  7. 7.
    Korez, R., et al.: Sparse and multi-object pose + shape modeling of the three-dimensional scoliotic spine. In: 2016 IEEE 13th ISBI, pp. 225–228 (2016)Google Scholar
  8. 8.
    Cresson, T., et al.: Surface reconstruction from planar X-ray images using moving least squares. In: Conference Proceedings IEEE EMBS, pp. 3967–3970 (2008)Google Scholar
  9. 9.
    Chen, C., Xie, W., et al.: Fully automatic X-ray image segmentation via joint estimation of image displacements. MICCAI 16, 227–234 (2013)Google Scholar
  10. 10.
    Guillemot, C., Le Meur, O.: Image inpainting [overview and advances]. IEEE Signal Process. Mag. 31, 127–144 (2014)CrossRefGoogle Scholar
  11. 11.
    Vidal, P.A.: Image processing tools development to help the 3D reconstruction of the spine from postoperative radiographic images. Master thesis, Montréal, École de technologie supérieure (2017)Google Scholar

Copyright information

© Springer International Publishing AG 2017

Authors and Affiliations

  • B. Aubert
    • 1
    • 2
    Email author
  • P. A. Vidal
    • 1
  • S. Parent
    • 2
  • T. Cresson
    • 1
  • C. Vazquez
    • 1
  • J. De Guise
    • 1
    • 2
  1. 1.Laboratoire de recherche en imagerie et orthopédie (LIO), École de technologie supérieureCentre de recherche du CHUMMontréalCanada
  2. 2.Sainte-Justine Hospital Research CenterMontréalCanada

Personalised recommendations