首页 | 本学科首页   官方微博 | 高级检索  
     


Explaining the Rationale of Deep Learning Glaucoma Decisions with Adversarial Examples
Authors:Jooyoung Chang  Jinho Lee  Ahnul Ha  Young Soo Han  Eunoo Bak  Seulggie Choi  Jae Moon Yun  Uk Kang  Il Hyung Shin  Joo Young Shin  Taehoon Ko  Ye Seul Bae  Baek-Lok Oh  Ki Ho Park  Sang Min Park
Affiliation:1. Department of Biomedical Sciences, Seoul National University Graduate School, Seoul, Republic of Korea;2. Department of Ophthalmology, Seoul National University College of Medicine, Seoul, Republic of Korea;3. Department of Ophthalmology, Hallym University Chuncheon Sacred Heart Hospital, Chuncheon, Republic of Korea;4. Department of Ophthalmology, Seoul National University Hospital, Seoul, Republic of Korea;5. Department of Family Medicine, Seoul National University Hospital, Seoul, Republic of Korea;6. InTheSmart Co., Ltd., Seoul, Republic of Korea;7. Biomedical Research Institute, Seoul National University Hospital, Seoul, Republic of Korea;8. Department of Ophthalmology, Seoul Metropolitan Government Seoul National University Boramae Medical Center, Seoul, Republic of Korea;9. Office of Hospital Information, Seoul National University Hospital, Seoul, Republic of Korea
Abstract:
Keywords:AE"  },{"  #name"  :"  keyword"  ,"  $"  :{"  id"  :"  kwrd0015"  },"  $$"  :[{"  #name"  :"  text"  ,"  _"  :"  adversarial example  negative adversarial example for disc rim narrowing  positive adversarial example for disc rim narrowing  negative adversarial example for referable glaucoma  positive adversarial example for referable glaucoma  negative adversarial example for increased cup-to-disc ratio  positive adversarial example for increased cup-to-disc ratio  negative adversarial example for retinal nerve fiber layer defect  positive adversarial example for retinal nerve fiber layer defect  AUC"  },{"  #name"  :"  keyword"  ,"  $"  :{"  id"  :"  kwrd0035"  },"  $$"  :[{"  #name"  :"  text"  ,"  _"  :"  area under the receiver operating characteristic curve  DLM"  },{"  #name"  :"  keyword"  ,"  $"  :{"  id"  :"  kwrd0045"  },"  $$"  :[{"  #name"  :"  text"  ,"  _"  :"  deep learning model  DRN"  },{"  #name"  :"  keyword"  ,"  $"  :{"  id"  :"  kwrd0055"  },"  $$"  :[{"  #name"  :"  text"  ,"  _"  :"  disc rim narrowing  FOV"  },{"  #name"  :"  keyword"  ,"  $"  :{"  id"  :"  kwrd0065"  },"  $$"  :[{"  #name"  :"  text"  ,"  _"  :"  field of view  GradCAM"  },{"  #name"  :"  keyword"  ,"  $"  :{"  id"  :"  kwrd0075"  },"  $$"  :[{"  #name"  :"  text"  ,"  _"  :"  gradient-weighted class activation mapping  ICDR"  },{"  #name"  :"  keyword"  ,"  $"  :{"  id"  :"  kwrd0085"  },"  $$"  :[{"  #name"  :"  text"  ,"  _"  :"  increased cup-to-disc ratio  RG"  },{"  #name"  :"  keyword"  ,"  $"  :{"  id"  :"  kwrd0095"  },"  $$"  :[{"  #name"  :"  text"  ,"  _"  :"  referable glaucoma  RNFLD"  },{"  #name"  :"  keyword"  ,"  $"  :{"  id"  :"  kwrd0105"  },"  $$"  :[{"  #name"  :"  text"  ,"  _"  :"  retinal nerve fiber layer defect
本文献已被 ScienceDirect 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号