Assessing the Ability of Generative Adversarial Networks to Learn Canonical Medical Image Statistics

Varun A. Kelkar, Dimitrios S. Gotsis, Frank J. Brooks, K. C. Prabhat, Kyle J. Myers, Rongping Zeng, Mark A. Anastasio

Research output: Contribution to journalArticlepeer-review


In recent years, generative adversarial networks (GANs) have gained tremendous popularity for potential applications in medical imaging, such as medical image synthesis, restoration, reconstruction, translation, as well as objective image quality assessment. Despite the impressive progress in generating high-resolution, perceptually realistic images, it is not clear if modern GANs reliably learn the statistics that are meaningful to a downstream medical imaging application. In this work, the ability of a state-of-the-art GAN to learn the statistics of canonical stochastic image models (SIMs) that are relevant to objective assessment of image quality is investigated. It is shown that although the employed GAN successfully learned several basic first- and second-order statistics of the specific medical SIMs under consideration and generated images with high perceptual quality, it failed to correctly learn several per-image statistics pertinent to the these SIMs, highlighting the urgent need to assess medical image GANs in terms of objective measures of image quality.

Original languageEnglish (US)
Pages (from-to)1799-1808
Number of pages10
JournalIEEE transactions on medical imaging
Issue number6
StatePublished - Jun 1 2023
Externally publishedYes


  • Generative models
  • generative adversarial networks
  • objective image quality assessment
  • stochastic image models

ASJC Scopus subject areas

  • Software
  • Radiological and Ultrasound Technology
  • Electrical and Electronic Engineering
  • Computer Science Applications


Dive into the research topics of 'Assessing the Ability of Generative Adversarial Networks to Learn Canonical Medical Image Statistics'. Together they form a unique fingerprint.

Cite this