This paper proposes a deep learning-based denoising method for noisy low-dose computerized tomography (CT) images in the absence of paired training data. The proposed method uses a fidelityembedded generative adversarial network (GAN) to learn a denoising function from unpaired training data of low-dose CT (LDCT) and standard-dose CT (SDCT) images, where the denoising function is the optimal generator in the GAN framework. This paper analyzes the f-GAN objective to derive a suitable generator that is optimized by minimizing a weighted sum of two losses: The Kullback-Leibler divergence between an SDCT data distribution and a generated distribution, and the 2 loss between the LDCT image and the corresponding generated images (or denoised image). The computed generator reffects the prior belief about SDCT data distribution through training. We observed that the proposed method allows the preservation of fine anomalous features while eliminating noise. The experimental results show that the proposed deeplearning method with unpaired datasets performs comparably to a method using paired datasets. A clinical experiment was also performed to show the validity of the proposed method for noise arising in the low-dose X-ray CT.
Bibliographical noteFunding Information:
The work of H. S. Park and J. Baek was supported by the National Institute for Mathematical Sciences (NIMS) Grant Funded by the Korean government under Grant NIMS-B19610000. The work of S. K. You was supported in part by the Chungnam National University Hospital Research Fund, in 2018. The work of J. K. Seo was supported by the National Research Foundation of Korea (NRF) under Grant 2015R1A5A1009350.
© 2020 Inderscience Enterprises Ltd.. All rights reserved.
All Science Journal Classification (ASJC) codes
- Computer Science(all)
- Materials Science(all)