Accurate prediction of glaucoma from colour fundus images with a convolutional neural network that relies on active and transfer learning.

Journal: Acta ophthalmologica
Published Date: Jul 25, 2019

Abstract

PURPOSE: To assess the use of deep learning (DL) for computer-assisted glaucoma identification, and the impact of training using images selected by an active learning strategy, which minimizes labelling cost. Additionally, this study focuses on the explainability of the glaucoma classifier.

Authors

  • Ruben Hemelings
    Research Group Ophthalmology, KU Leuven, Kapucijnenvoer 33, 3000 Leuven, Belgium; ESAT-PSI, KU Leuven, Kasteelpark Arenberg 10, 3001 Leuven, Belgium; VITO NV, Boeretang 200, 2400 Mol, Belgium.
  • Bart Elen
    VITO NV, Boeretang 200, 2400 Mol, Belgium.
  • João Barbosa-Breda
    Research Group Ophthalmology, KU Leuven, Leuven, Belgium.
  • Sophie Lemmens
    Research Group Ophthalmology, KU Leuven, Leuven, Belgium.
  • Maarten Meire
    TC CS-ADVISE, KU Leuven, Geel, Belgium.
  • Sayeh Pourjavan
    Department of Ophthalmology, Cliniques Universitaires Saint Luc, UCL, Brussels, Belgium.
  • Evelien Vandewalle
    Research Group Ophthalmology, KU Leuven, Leuven, Belgium.
  • Sara Van de Veire
    AZ Sint-Jan, Brugge, Belgium.
  • Matthew B Blaschko
  • Patrick De Boever
    Hasselt University, Agoralaan building D, 3590 Diepenbeek, Belgium; VITO NV, Boeretang 200, 2400 Mol, Belgium. Electronic address: patrick.deboever@vito.be.
  • Ingeborg Stalmans
    Research Group Ophthalmology, KU Leuven, Kapucijnenvoer 33, 3000 Leuven, Belgium.

Keywords

External Resources

Popular Topics
Recent Journals