TY - JOUR
T1 - Assessing the Ability of Generative Adversarial Networks to Learn Canonical Medical Image Statistics
AU - Kelkar, Varun A.
AU - Gotsis, Dimitrios S.
AU - Brooks, Frank J.
AU - Prabhat, K. C.
AU - Myers, Kyle J.
AU - Zeng, Rongping
AU - Anastasio, Mark A.
N1 - Funding Information:
This work was supported in part by the National Institute of Health (NIH) under Award R01EB031585 and Award P41EB031772. The work of Varun A. Kelkar was supported by the Research Participation Program at the Center for Devices and Radiological Health Administered by the Oak Ridge Institute for Science and Education through an Inter-Agency Agreement between the U.S. Department of Energy and U.S. Food and Drug Administration (FDA).
Publisher Copyright:
© 1982-2012 IEEE.
PY - 2023/6/1
Y1 - 2023/6/1
N2 - 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.
AB - 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.
KW - Generative models
KW - generative adversarial networks
KW - objective image quality assessment
KW - stochastic image models
UR - http://www.scopus.com/inward/record.url?scp=85148438236&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85148438236&partnerID=8YFLogxK
U2 - 10.1109/TMI.2023.3241454
DO - 10.1109/TMI.2023.3241454
M3 - Article
C2 - 37022374
AN - SCOPUS:85148438236
SN - 0278-0062
VL - 42
SP - 1799
EP - 1808
JO - IEEE transactions on medical imaging
JF - IEEE transactions on medical imaging
IS - 6
ER -