Machine Learning Model for Chest Radiographs: Using Local Data to Enhance Performance.

Journal: Canadian Association of Radiologists journal = Journal l'Association canadienne des radiologistes
Published Date:

Abstract

PURPOSE: To develop and assess the performance of a machine learning model which screens chest radiographs for 14 labels, and to determine whether fine-tuning the model on local data improves its performance. Generalizability at different institutions has been an obstacle to machine learning model implementation. We hypothesized that the performance of a model trained on an open-source dataset will improve at our local institution after being fine-tuned on local data.

Authors

  • Sarah F Mohn
    University of British Columbia, Vancouver, BC, Canada.
  • Marco Law
    University of British Columbia, Vancouver, BC, Canada.
  • Maria Koleva
    University of British Columbia, Vancouver, BC, Canada.
  • Brian Lee
  • Adam Berg
    Vancouver General Hospital, Vancouver, BC, Canada.
  • Nicolas Murray
    Vancouver General Hospital, Vancouver, BC, Canada.
  • Savvas Nicolaou
    Department of Radiology, Vancouver General Hospital, University of British Columbia, Vancouver, 899 12th Avenue West, British Columbia V5Z 1M9, Canada. Electronic address: savvas.nicolaou@vch.ca.
  • William A Parker
    Stanford University, Stanford, CA, USA.