TY - JOUR
T1 - Optical coherence tomography image denoising using a generative adversarial network with speckle modulation
AU - Dong, Zhao
AU - Liu, Guoyan
AU - Ni, Guangming
AU - Jerwick, Jason
AU - Duan, Lian
AU - Zhou, Chao
N1 - Funding Information:
The authors would like to thank Jinyun Zou, Yongyang Huang, Jing Men and Zhiwen Yang for helpful discussions. This work was supported by NSF grants IDBR (DBI-1455613), PFI: AIR-TT (IIP-1640707), NIH grants R15EB019704 and R01EB025209.
Funding Information:
The authors would like to thank Jinyun Zou, Yongyang Huang, Jing Men and Zhiwen Yang for helpful discussions. This work was supported by NSF grants IDBR (DBI‐1455613), PFI: AIR‐TT (IIP‐1640707), NIH grants R15EB019704 and R01EB025209.
Funding Information:
Foundation for the National Institutes of Health, Grant/Award Numbers: R01EB025209, R15EB019704; National Science Foundation, Grant/Award Numbers: DBI‐1455613, IIP‐1640707 Funding information
Publisher Copyright:
© 2020 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
PY - 2020/4/1
Y1 - 2020/4/1
N2 - Optical coherence tomography (OCT) is widely used for biomedical imaging and clinical diagnosis. However, speckle noise is a key factor affecting OCT image quality. Here, we developed a custom generative adversarial network (GAN) to denoise OCT images. A speckle-modulating OCT (SM-OCT) was built to generate low speckle images to be used as the ground truth. In total, 210 000 SM-OCT images were used for training and validating the neural network model, which we call SM-GAN. The performance of the SM-GAN method was further demonstrated using online benchmark retinal images, 3D OCT images acquired from human fingers and OCT videos of a beating fruit fly heart. The denoise performance of the SM-GAN model was compared to traditional OCT denoising methods and other state-of-the-art deep learning based denoise networks. We conclude that the SM-GAN model presented here can effectively reduce speckle noise in OCT images and videos while maintaining spatial and temporal resolutions.
AB - Optical coherence tomography (OCT) is widely used for biomedical imaging and clinical diagnosis. However, speckle noise is a key factor affecting OCT image quality. Here, we developed a custom generative adversarial network (GAN) to denoise OCT images. A speckle-modulating OCT (SM-OCT) was built to generate low speckle images to be used as the ground truth. In total, 210 000 SM-OCT images were used for training and validating the neural network model, which we call SM-GAN. The performance of the SM-GAN method was further demonstrated using online benchmark retinal images, 3D OCT images acquired from human fingers and OCT videos of a beating fruit fly heart. The denoise performance of the SM-GAN model was compared to traditional OCT denoising methods and other state-of-the-art deep learning based denoise networks. We conclude that the SM-GAN model presented here can effectively reduce speckle noise in OCT images and videos while maintaining spatial and temporal resolutions.
KW - de-noise
KW - deep learning
KW - generative adversarial network
KW - optical coherence tomography
UR - http://www.scopus.com/inward/record.url?scp=85078892187&partnerID=8YFLogxK
U2 - 10.1002/jbio.201960135
DO - 10.1002/jbio.201960135
M3 - Article
C2 - 31970879
AN - SCOPUS:85078892187
SN - 1864-063X
VL - 13
JO - Journal of Biophotonics
JF - Journal of Biophotonics
IS - 4
M1 - e201960135
ER -