A deep learning-based interactive medical image segmentation framework with sequential memory.

Journal: Computer methods and programs in biomedicine
Published Date:

Abstract

BACKGROUND AND OBJECTIVE: Image segmentation is an essential component in medical image analysis. The case of 3D images such as MRI is particularly challenging and time consuming. Interactive or semi-automatic methods are thus highly desirable. However, existing methods do not exploit the typical sequentiality of real user interactions. This is due to the interaction memory used in these systems, which discards ordering. In contrast, we argue that the order of the user corrections should be used for training and lead to performance improvements.

Authors

  • Ivan Mikhailov
    EnCoV, Institut Pascal, Université Clermont Auvergne, Clermont-Ferrand, 63000, France; SurgAR, 22 All. Alan Turing, Clermont-Ferrand, 63000, France. Electronic address: ivanmikhailov.mail@gmail.com.
  • Benoit Chauveau
    SurgAR, 22 All. Alan Turing, Clermont-Ferrand, 63000, France; CHU de Clermont-Ferrand, Clermont-Ferrand, 63000, France.
  • Nicolas Bourdel
    Department of Gynaecological Surgery, CHU Clermont-Ferrand, 1 Place Lucie et Raymond Aubrac, 63000, Clermont-Ferrand, France. nicolas.bourdel@gmail.com.
  • Adrien Bartoli
    EnCoV, Institut Pascal, CNRS, Université Clermont Auvergne, Clermont-Ferrand, France.