CycleGAN denoising of extreme low-dose cardiac CT using wavelet-assisted noise disentanglement

Gu, Jawook; Yang, Tae Seong; Ye, Jong Chul; Yang, Dong Hyun

doi:10.1016/j.media.2021.102209

Cited by 48 publications

(26 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For deep learning methods, we cannot use the supervised learning approaches as there are no ground-truth data available. Moreover, cycleGAN-based approaches [43]- [45] are not feasible in this case, as we do not have access to clean liver data set albeit unmatched. Therefore, we compare with self-supervised learning approaches, such as Noise2Noise (N2N) [13], Neighbor2Neighbor (Nei2Nei) [34], and Noise2Score (N2Score) [15].…”

Section: Comparison Methodsmentioning

confidence: 99%

MR Image Denoising and Super-Resolution Using Regularized Reverse Diffusion

Chung¹,

Lee²,

Ye³

2022

Preprint

Self Cite

View full text Add to dashboard Cite

Patient scans from MRI often suffer from noise, which hampers the diagnostic capability of such images. As a method to mitigate such artifact, denoising is largely studied both within the medical imaging community and beyond the community as a general subject. However, recent deep neural network-based approaches mostly rely on the minimum mean squared error (MMSE) estimates, which tend to produce a blurred output. Moreover, such models suffer when deployed in real-world sitautions: out-of-distribution data, and complex noise distributions that deviate from the usual parametric noise models. In this work, we propose a new denoising method based on score-based reverse diffusion sampling, which overcomes all the aforementioned drawbacks. Our network, trained only with coronal knee scans, excels even on out-of-distribution in vivo liver MRI data, contaminated with complex mixture of noise. Even more, we propose a method to enhance the resolution of the denoised image with the same network. With extensive experiments, we show that our method establishes state-of-theart performance, while having desirable properties which prior MMSE denoisers did not have: flexibly choosing the extent of denoising, and quantifying uncertainty.

show abstract

Section: Comparison Methodsmentioning

confidence: 99%

MR Image Denoising and Super-Resolution Using Regularized Reverse Diffusion

Chung¹,

Lee²,

Ye³

2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…1, we first obtain a high-frequency image by a wavelet transform, and then exploit the patch-wise deep metric learning using the features of the image before and after the denoising process. Here, similar to [8,7], high frequency images are obtained by a wavelet decomposition, putting zeros at the low-frequency band, and then performing wavelet recomposition. This preprocessing preserves the low frequency image of input.…”

Section: Overview Of the Proposed Methodsmentioning

confidence: 99%

“…We split the low-dose data into two sets, a train set with 45 volumes and a test set with 9 volumes. We compare our method with existing denoising methods based on the image translation framework, such as cycleGAN [23], wavCycleGAN [7], GAN-Circle [21]and Cycle-free invertible cycleGAN [11]. We measure PSNR, SSIM to compare the denoising performance for AAPM dataset.…”

Section: Network Architecturementioning

confidence: 99%

“…Furthermore, the difficulty of obtaining matched CT data pairs has led to exploring various unsupervised learning approaches [22,2]. In particular, image translation methodology is successfully employed for the low-dose CT noise suppression by learning the noise patterns through a comparison between the low-dose CT(LDCT) and high-dose CT(HDCT) distributions [20,7,16].…”

Section: Introductionmentioning

confidence: 99%

“…Especially in the temporal bone CT scans which we are interested in this paper, the CT number supports the evaluation of a pathologies of the ear, such as an inflammation within the inner ear, and a cholesteatoma in early stage [5,13,1]. Wavelet-assisted approaches [8,7] partially address this problem by constraining the pixel value variations only in highfrequency subbands. However, even in the high-frequency image, there exists many important structural information, which can be altered during the image translation so that the localized CT number variation is unavoidable.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Patch-wise Deep Metric Learning for Unsupervised Low-Dose CT Denoising

Jung¹,

Lee²,

You³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

The acquisition conditions for low-dose and high-dose CT images are usually different, so that the shifts in the CT numbers often occur. Accordingly, unsupervised deep learning-based approaches, which learn the target image distribution, often introduce CT number distortions and result in detrimental effects in diagnostic performance. To address this, here we propose a novel unsupervised learning approach for lowdose CT reconstruction using patch-wise deep metric learning. The key idea is to learn embedding space by pulling the positive pairs of image patches which shares the same anatomical structure, and pushing the negative pairs which have same noise level each other. Thereby, the network is trained to suppress the noise level, while retaining the original global CT number distributions even after the image translation. Experimental results confirm that our deep metric learning plays a critical role in producing high quality denoised images without CT number shift.

show abstract

CT image denoising methods for image quality improvement and radiation dose reduction

Sadia,

Chen,

Zhang

2024

J Applied Clin Med Phys

View full text Add to dashboard Cite

With the ever‐increasing use of computed tomography (CT), concerns about its radiation dose have become a significant public issue. To address the need for radiation dose reduction, CT denoising methods have been widely investigated and applied in low‐dose CT images. Numerous noise reduction algorithms have emerged, such as iterative reconstruction and most recently, deep learning (DL)‐based approaches. Given the rapid advancements in Artificial Intelligence techniques, we recognize the need for a comprehensive review that emphasizes the most recently developed methods. Hence, we have performed a thorough analysis of existing literature to provide such a review. Beyond directly comparing the performance, we focus on pivotal aspects, including model training, validation, testing, generalizability, vulnerability, and evaluation methods. This review is expected to raise awareness of the various facets involved in CT image denoising and the specific challenges in developing DL‐based models.

show abstract

CycleGAN denoising of extreme low-dose cardiac CT using wavelet-assisted noise disentanglement

Cited by 48 publications

References 18 publications

MR Image Denoising and Super-Resolution Using Regularized Reverse Diffusion

MR Image Denoising and Super-Resolution Using Regularized Reverse Diffusion

Patch-wise Deep Metric Learning for Unsupervised Low-Dose CT Denoising

CT image denoising methods for image quality improvement and radiation dose reduction

Contact Info

Product

Resources

About