English

Towards generative adversarial networks as a new paradigm for radiology education

Computer Vision and Pattern Recognition 2018-12-05 v1

Abstract

Medical students and radiology trainees typically view thousands of images in order to "train their eye" to detect the subtle visual patterns necessary for diagnosis. Nevertheless, infrastructural and legal constraints often make it difficult to access and quickly query an abundance of images with a user-specified feature set. In this paper, we use a conditional generative adversarial network (GAN) to synthesize 1024×10241024\times1024 pixel pelvic radiographs that can be queried with conditioning on fracture status. We demonstrate that the conditional GAN learns features that distinguish fractures from non-fractures by training a convolutional neural network exclusively on images sampled from the GAN and achieving an AUC of >0.95>0.95 on a held-out set of real images. We conduct additional analysis of the images sampled from the GAN and describe ongoing work to validate educational efficacy.

Keywords

Cite

@article{arxiv.1812.01547,
  title  = {Towards generative adversarial networks as a new paradigm for radiology education},
  author = {Samuel G. Finlayson and Hyunkwang Lee and Isaac S. Kohane and Luke Oakden-Rayner},
  journal= {arXiv preprint arXiv:1812.01547},
  year   = {2018}
}

Comments

Machine Learning for Health (ML4H) Workshop at NeurIPS 2018 arXiv:cs/0101200