Style transfer strategy for developing a generalizable deep learning application in digital pathology.

Journal: Computer methods and programs in biomedicine
Published Date: Oct 25, 2020

Abstract

BACKGROUND AND OBJECTIVES: Despite recent advances in artificial intelligence for medical images, the development of a robust deep learning model for identifying malignancy on pathology slides has been limited by problems related to substantial inter- and intra-institutional heterogeneity attributable to tissue preparation. The paucity of available data aggravates this limitation for relatively rare cancers. Here, using ovarian cancer pathology images, we explored the effect of image-to-image style transfer approaches on diagnostic performance.

Authors

  • Seo Jeong Shin
    Department of Biomedical Sciences, Ajou University Graduate School of Medicine, Suwon, Republic of Korea.
  • Seng Chan You
    Department of Biomedical Informatics, Ajou University School of Medicine, Suwon, South Korea.
  • Hokyun Jeon
    Department of Biomedical Sciences, Ajou University Graduate School of Medicine, Suwon, Republic of Korea.
  • Ji Won Jung
    Department of Pathology, University of Ulsan College of Medicine, Asan Medical Center, Seoul, Republic of Korea; Asan Institute for Life Science, University of Ulsan College of Medicine, Asan Medical Center, Seoul, Republic of Korea.
  • Min Ho An
    So Ahn Public Health Center, Wando-gun, Jeollanam-do, Republic of Korea.
  • Rae Woong Park
    Department of Biomedical Informatics, Ajou University School of Medicine, Suwon, Gyeonggi-do, Republic of Korea.
  • Jin Roh
    Department of Pathology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Republic of Korea.

Keywords

External Resources

Popular Topics
Recent Journals