3D Cascaded Convolutional Networks for Multi-vertebrae Segmentation.

Journal: Current medical imaging
PMID:

Abstract

BACKGROUND: Automatic approach to vertebrae segmentation from computed tomography (CT) images is very important in clinical applications. As the intricate appearance and variable architecture of vertebrae across the population, cognate constructions in close vicinity, pathology, and the interconnection between vertebrae and ribs, it is a challenge to propose a 3D automatic vertebrae CT image segmentation method.

Authors

  • Liu Xia
    School of Automation, Harbin University of Science and Technology, Harbin 150001, China.
  • Liu Xiao
    School of Automation, Harbin University of Science and Technology, Harbin 150001, China.
  • Gan Quan
    School of Automation, Harbin University of Science and Technology, Harbin 150001, China.
  • Wang Bo
    School of Automation, Harbin University of Science and Technology, Harbin 150001, China.