Testing the Ability of Convolutional Neural Networks to Learn Radiomic Features.

Journal: Computer methods and programs in biomedicine
Published Date: Mar 17, 2022

Abstract

BACKGROUND AND OBJECTIVE: Radiomics and deep learning have emerged as two distinct approaches to medical image analysis. However, their relative expressive power remains largely unknown. Theoretically, hand-crafted radiomic features represent a mere subset of features that neural networks can approximate, thus making deep learning a more powerful approach. On the other hand, automated learning of hand-crafted features may require a prohibitively large number of training samples. Here we directly test the ability of convolutional neural networks (CNNs) to learn and predict the intensity, shape, and texture properties of tumors as defined by standardized radiomic features.

Authors

  • Ivan S Klyuzhin
    Department of Medicine, University of British Columbia, Vancouver, BC, Canada.
  • Yixi Xu
    AI for Health, Microsoft, Redmond, WA, USA.
  • Anthony Ortiz
    AI for Good Research Lab, Microsoft, Redmond, WA, United States of America.
  • Juan Lavista Ferres
    AI for Good Research Lab, Microsoft, Redmond, WA, 98052, USA.
  • Ghassan Hamarneh
    Medical Image Analysis Lab, Simon Fraser University, Burnaby, BC, Canada. Electronic address: hamarneh@sfu.ca.
  • Arman Rahmim

Keywords

External Resources

Popular Topics
Recent Journals