Hierarchical deep learning system for orbital fracture detection and trap-door classification on CT images.

Journal: Computers in biology and medicine
Published Date: Jul 11, 2025

Abstract

OBJECTIVE: To develop and evaluate a hierarchical deep learning system that detects orbital fractures on computed tomography (CT) images and classifies them as depressed or trap-door types.

Authors

  • Hiroaki Oku
    Department of Ophthalmology, Kyoto Prefectural University of Medicine, Kyoto, Japan.
  • Yuto Nakamura
    Division of Radiology, Yamaguchi University Hospital, 1-1-1 Minamikogushi, Yamaguchi, 755-8505, Japan.
  • Yuma Kanematsu
    Department of Biomedical Engineering, Faculty of Life and Medical Sciences, Doshisha University, Kyotanabe, Japan.
  • Ayumu Akagi
    Department of Biomedical Engineering, Faculty of Life and Medical Sciences, Doshisha University, Kyotanabe, Japan.
  • Shigeru Kinoshita
    Department of Ophthalmology, Kyoto Prefectural University of Medicine, Kyoto, Japan.
  • Chie Sotozono
    Department of Ophthalmology, Kyoto Prefectural University of Medicine, Kyoto, Japan.
  • Noriko Koizumi
    Department of Biomedical Engineering, Faculty of Life and Medical Sciences, Doshisha University, Kyotanabe, Japan.
  • Akihide Watanabe
    Department of Ophthalmology, Kyoto Prefectural University of Medicine, Kyoto, Japan.
  • Naoki Okumura
    Department of Biomedical Engineering, Faculty of Life and Medical Sciences, Doshisha University, Kyotanabe, Japan. Electronic address: nokumura@mail.doshisha.ac.jp.

Keywords

External Resources

Popular Topics
Recent Journals