Feasibility of a generalized convolutional neural network for automated identification of vertebral compression fractures: The Manitoba Bone Mineral Density Registry.

Journal: Bone
Published Date: Sep 1, 2021

Abstract

BACKGROUND: Vertebral fracture assessment (VFA) images are acquired in dual-energy (DE) or single-energy (SE) scan modes. Automated identification of vertebral compression fractures, from VFA images acquired using GE Healthcare scanners in DE mode, has achieved high accuracy through the use of convolutional neural networks (CNNs). Due to differences between DE and SE images, it is uncertain whether CNNs trained on one scan mode will generalize to the other.

Authors

  • Barret A Monchka
    Department of Community Health Sciences, University of Manitoba, Winnipeg, Manitoba, Canada.
  • Douglas Kimelman
    From the Department of Radiology, University of Manitoba, 820 Sherbrook St, GA216, Winnipeg, MB, Canada R3T 2N2 (S.D., C.K., D.K., M.J.J., J.M.D., W.D.L.); and St Boniface Hospital Albrechtsen Research Centre, Winnipeg, Canada (C.K., D.K.).
  • Lisa M Lix
    Department of Community Health Sciences, Rady Faculty of Health Sciences, University of Manitoba, Winnipeg, MB, Canada; George and Faye Yee Centre for Healthcare Innovation, University of Manitoba, Winnipeg, MB, Canada. Electronic address: Lisa.Lix@umanitoba.ca.
  • William D Leslie
    From the Department of Radiology, University of Manitoba, 820 Sherbrook St, GA216, Winnipeg, MB, Canada R3T 2N2 (S.D., C.K., D.K., M.J.J., J.M.D., W.D.L.); and St Boniface Hospital Albrechtsen Research Centre, Winnipeg, Canada (C.K., D.K.).

Keywords

External Resources

Popular Topics
Recent Journals