Training immunophenotyping deep learning models with the same-section ground truth cell label derivation method improves virtual staining accuracy.

Journal: Frontiers in immunology
PMID:

Abstract

INTRODUCTION: Deep learning (DL) models predicting biomarker expression in images of hematoxylin and eosin (H&E)-stained tissues can improve access to multi-marker immunophenotyping, crucial for therapeutic monitoring, biomarker discovery, and personalized treatment development. Conventionally, these models are trained on ground truth cell labels derived from IHC-stained tissue sections adjacent to H&E-stained ones, which might be less accurate than labels from the same section. Although many such DL models have been developed, the impact of ground truth cell label derivation methods on their performance has not been studied.

Authors

  • Abu Bakr Azam
    School of Mechanical and Aerospace Engineering, College of Engineering, Nanyang Technological University, Singapore, Singapore.
  • Felicia Wee
    Institute of Molecular and Cell Biology, Agency for Science, Technology and Research, Singapore, Singapore.
  • Juha P Väyrynen
    Department of Medical Oncology, Dana-Farber Cancer Institute and Harvard Medical School, Boston, Massachusetts.
  • Willa Wen-You Yim
    Institute of Molecular and Cell Biology, Agency for Science, Technology and Research, Singapore, Singapore.
  • Yue Zhen Xue
    Institute of Molecular and Cell Biology, Agency for Science, Technology and Research, Singapore, Singapore.
  • Bok Leong Chua
    School of Mechanical and Aerospace Engineering, College of Engineering, Nanyang Technological University, Singapore, Singapore.
  • Jeffrey Chun Tatt Lim
    Institute of Molecular and Cell Biology (IMCB), Agency for Science, Technology and Research (A*STAR), Singapore, 138673, Singapore.
  • Aditya Chidambaram Somasundaram
    School of Electrical and Electronics Engineering, Nanyang Technological University, Singapore, Singapore.
  • Daniel Shao Weng Tan
    Division of Medical Oncology, National Cancer Centre, Singapore, Singapore.
  • Angela Takano
    Department of Anatomical Pathology, Division of Pathology, Singapore General Hospital, Singapore, Singapore.
  • Chun Yuen Chow
    Department of Anatomical Pathology, Division of Pathology, Singapore General Hospital, Singapore, Singapore.
  • Li Yan Khor
    Department of Anatomical Pathology, Division of Pathology, Singapore General Hospital, Singapore, Singapore.
  • Tony Kiat Hon Lim
    Department of Anatomical Pathology, Division of Pathology, Singapore General Hospital, Singapore, Singapore.
  • Joe Yeong
    Institute of Molecular and Cell Biology (IMCB), Agency of Science, Technology and Research (A*STAR), 61 Biopolis Drive, Proteos, Singapore 138673, Singapore; Department of Anatomical Pathology, Singapore General Hospital, 20 College Road, Academia, Level 10 Diagnostic Tower, Singapore 169856, Singapore. Electronic address: yeongps@imcb.a-star.edu.sg.
  • Mai Chan Lau
    Program in MPE Molecular Pathological Epidemiology, Department of Pathology, Brigham and Women's Hospital and Harvard Medical School, Boston, Massachusetts.
  • Yiyu Cai
    School of Mechanical and Aerospace Engineering, College of Engineering, Nanyang Technological University, Singapore, Singapore.