Pathologic Image Classification of Flat Urothelial Lesions Using Pathologic Criteria-Based Deep Learning.

Journal: American journal of clinical pathology
PMID:

Abstract

OBJECTIVES: Pathologic diagnosis of flat urothelial lesions is subject to high interobserver variability. We expected that deep learning could improve the accuracy and consistency of such pathologic diagnosis, although the learning process is a black box. We therefore propose a new approach for pathologic image classification incorporating the diagnostic process of the pathologist into a deep learning method.

Authors

  • Toui Nishikawa
    Department of Human Pathology, Wakayama Medical University, Wakayama, Japan.
  • Ryuta Iwamoto
    Department of Human Pathology, Wakayama Medical University, Wakayama, Japan.
  • Ibu Matsuzaki
    Department of Human Pathology, Wakayama Medical University, Wakayama, Japan.
  • Fidele Yambayamba Musangile
    Department of Human Pathology, Wakayama Medical University, Wakayama, Japan.
  • Ayata Takahashi
    Department of Human Pathology, Wakayama Medical University, Wakayama, Japan.
  • Yurina Mikasa
    Department of Human Pathology, Wakayama Medical University, Wakayama, Japan.
  • Yuichi Takahashi
    Department of Human Pathology, Wakayama Medical University, Wakayama, Japan.
  • Fumiyoshi Kojima
    Department of Human Pathology, Wakayama Medical University, Wakayama, Japan.
  • Shin-Ichi Murata
    Department of Human Pathology, Wakayama Medical University, Wakayama, Japan.