Research on Low-Dose CT Image Denoising Methods Based on Self-Supervised Learning

An Ge

doi:10.54097/1924aq22

Authors

An Ge

DOI:

https://doi.org/10.54097/1924aq22

Keywords:

Low-dose CT, Image Denoising, Self-supervised Learning, Guided Image Filtering, Attention Gate, Residual Network

Abstract

Low-dose CT images are affected by noise and artifacts, compromising diagnostic reliability. While deep learning-based denoising methods have shown promise, most rely on paired normal-dose CT images, which are difficult to obtain clinically. To address this, this study proposes a self-supervised denoising method using pseudo-labels generated by guided image filtering, along with an Attention-Gate-enhanced REsidual Denoising network (AG-REDCNN). Experiments demonstrate that the method effectively reduces noise and preserves structural details without requiring normal-dose CT images during training, outperforming several existing approaches.

Downloads

Download data is not yet available.

References

[1] Borsdorf, Anja, Rainer Raupach, Thomas Flohr, and Joachim Hornegger. "Wavelet based noise reduction in CT-images using correlation analysis." IEEE transactions on medical imaging 27, no. 12 (2008): 1685-1703.

[2] Li, Zeheng, Junzhou Huang, Lifeng Yu, Yujie Chi, and Mingwu Jin. "Low-dose CT image denoising using cycle-consistent adversarial networks." In 2019 IEEE nuclear science symposium and medical imaging conference (NSS/MIC), pp. 1-3. IEEE, 2019.

[3] Zhang, Lei, Jianshe Xiong, and Yuezhong Zhou. “Edge-enhanced dense network based on attention for low-dose ct denoising.” In 2023 8th international conference on image, vision and computing (ICIVC), pp. 449-454. IEEE, 2023.

[4] Huang, Jiaxin, Kecheng Chen, Jiayu Sun, Xiaorong Pu, and Yazhou Ren. "Cross domain low-dose CT image denoising with semantic information alignment." In 2022 IEEE international conference on image processing (ICIP), pp. 4228-4232. IEEE, 2022.

[5] Trung, Nguyen Thanh, Dinh-Hoan Trinh, Nguyen Linh Trung, Tran Thi Thuy Quynh, and Manh-Ha Luu. "Dilated residual convolutional neural networks for low-dose CT image denoising." In 2020 IEEE Asia pacific conference on circuits and systems (APCCAS), pp. 189-192. IEEE, 2020.

[6] Koh, Sungho, Nam Kyung Lee, Suk Kim, Seung Baek Hong, Dong Uk Kim, and Sung Yong Han. "The efficacy of low-dose CT with deep learning image reconstruction in the surveillance of incidentally detected pancreatic cystic lesions." Abdominal Radiology 48, no. 8 (2023): 2585-2595.

[7] Ma, Yaoyao, Jing Wang, Chao Xu, Yuling Huang, Minghang Chu, Zhiwei Fan, Yishen Xu, and Di Wu. "CDAF-Net: a contextual contrast detail attention feature fusion network for low-dose CT denoising." IEEE Journal of Biomedical and Health Informatics 29, no. 3 (2024): 2048-2060.

[8] Kim, Seongjun, Byeongjoon Kim, and Jongduk Baek. "Dose-aware denoising diffusion model for low-dose CT." Physics in Medicine & Biology 70, no. 14 (2025): 145019.

[9] Wang, Shubin, Yi Liu, Pengcheng Zhang, Ping Chen, Zhiyuan Li, Rongbiao Yan, Shu Li, Ruifeng Hou, and Zhiguo Gui. "Compound feature attention network with edge enhancement for low-dose CT denoising." Journal of X-Ray Science and Technology 31, no. 5 (2023): 915-933.

[10] Wang, Zhenchuan, Minghui Liu, Xuan Cheng, Jinqi Zhu, Xiaomin Wang, Haigang Gong, Ming Liu, and Lifeng Xu. "Self‐adaption and texture generation: A hybrid loss function for low‐dose CT denoising." Journal of Applied Clinical Medical Physics 24, no. 9 (2023): e14113.

[11] Choi, Kihwan, and Sungwon Kim. "Statistical image restoration for low-dose CT using convolutional neural networks." In 2020 42nd Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC), pp. 1303-1306. IEEE, 2020.