An investigation into the risk of population bias in deep learning autocontouring.

Journal: Radiotherapy and oncology : journal of the European Society for Therapeutic Radiology and Oncology
Published Date: Jun 16, 2023

Abstract

BACKGROUND AND PURPOSE: To date, data used in the development of Deep Learning-based automatic contouring (DLC) algorithms have been largely sourced from single geographic populations. This study aimed to evaluate the risk of population-based bias by determining whether the performance of an autocontouring system is impacted by geographic population.

Authors

  • Yasmin McQuinlan
    DeepMind, London, United Kingdom.
  • Charlotte L Brouwer
    Department of Radiation Oncology, University of Groningen, University Medical Center Groningen, The Netherlands.
  • Zhixiong Lin
    Department of Psychiatry, Affiliated Hospital of Guangdong Medical University, Zhanjiang, China.
  • Yong Gan
    School of Computer and Communication Engineering, Zhengzhou University of Light Industry, Zhengzhou 450002, China.
  • Jin Sung Kim
    Yonsei University Health System, Seoul, Republic of Korea. Electronic address: jinsung@yuhs.ac.
  • Wouter van Elmpt
    Department of Radiation Oncology (Maastro), GROW School for Oncology, Maastricht University Medical Centre+, The Netherlands. Electronic address: wouter.vanelmpt@maastro.nl.
  • Mark J Gooding
    2 Mirada Medical Ltd, Oxford, UK.

Keywords

External Resources

Popular Topics
Recent Journals