Article Sidebar

Download
PDF
Statistic
Read Counter : 5 Download : 1

Main Article Content

Abstract

Low-dose computed tomography (LDCT) reduces radiation exposure but increases noise and structural degradation, which may affect diagnostic reliability. This paper presents a quantitative benchmarking framework to assess reinforcement learning (RL)-based LDCT denoising under standardized, reproducible experimental conditions. The proposed pipeline combines dataset splitting with controlled fragments, percentile preprocessing, classical and deep learning baselines, an RL denoising environment modeled as a Markov Decision Process, and multi-metric statistical validation. Experimental results on multi-level LDCT data show that the proposed RL_stageB model provides the highest overall reconstruction fidelity, which can achieve a mean PSNR 22.732 ± 0.947 dB and SSIM 0.929 ± 0.056, which is higher than strong classical baselines such as bilateral filtering 22.618 dB and Gaussian filtering 22.727 dB, while reducing edge distortion, Edge-L1=0.242. Statistically significant improvements (p<1e-300) are found in most comparisons using paired Wilcoxon signed-rank testing. The robustness analysis demonstrates that it maintains stable performance under both noise conditions and a small range of seed variance, with RMSE: 0.0696-0.0713. These findings present RL as an adaptive sequential denoising approach and provide a benchmarking framework for future LDCT restoration studies from a reliability perspective.

Keywords

Low-dose CT Reinforcement learning Image denoising Benchmarking framework

Article Details

How to Cite
Pattar, A. B. ., & N, T. S. . (2026). A quantitative benchmarking framework for reinforcement learning-based low-dose CT image denoising. Future Technology, 5(3), 107–117. Retrieved from https://fupubco.com/futech/article/view/885
Download Citation
Endnote/Zotero/Mendeley (RIS)
BibTeX
Bookmark and Share

References

  1. G. D. Jo, C. Ahn, J. H. Hong et al., “75% radiation dose reduction using deep learning reconstruction on low-dose chest CT,” BMC Med. Imaging, vol. 23, p. 121, 2023, doi: 10.1186/s12880-023-01081-8.
  2. H. Wang, L.-L. Li, J. Shang, J. Song, and B. Liu, “Application of deep learning image reconstruction in low-dose chest CT scan,” British Journal of Radiology, vol. 95, no. 1133, p. 20210380, May 2022, doi: 10.1259/bjr.20210380.
  3. H. Chen, Y. Zhang, M. K. Kalra, F. Lin, Y. Chen, P. Liao, J. Zhou, W. Wang, and G. Wang, “Low-dose CT with a residual encoder-decoder convolutional neural network,” IEEE Trans. Med. Imaging, vol. 36, no. 12, pp. 2524–2535, Dec. 2017, doi: 10.1109/TMI.2017.2715284.
  4. Q. Yang, P. Yan, Y. Zhang, H. Yu, Y. Shi, X. Mou, Y. Kalra, M. K. Kalra, and G. Wang, “Low-dose CT image denoising using a generative adversarial network with Wasserstein distance and perceptual loss,” IEEE Trans. Med. Imaging, vol. 37, no. 6, pp. 1348–1357, Jun. 2018, doi: 10.1109/TMI.2018.2827462.
  5. H. S. Park, K. Jeon, J. Lee, and S. K. You, “Denoising of pediatric low-dose abdominal CT using deep learning-based algorithm,” PLOS ONE, vol. 17, no. 1, p. e0260369, 2022, doi: 10.1371/journal.pone.0260369.
  6. C. Tian, L. Fei, W. Zheng, Y. Xu, W. Zuo, and C.-W. Lin, “Deep learning on image denoising: An overview,” Neural Networks, vol. 131, pp. 251–275, 2020, doi: 10.1016/j.neunet.2020.07.025.
  7. J. Zhang, W. Gong, L. Ye, F. Wang, Z. Shangguan, and Y. Cheng, “A review of deep learning methods for denoising of medical low-dose CT images,” Comput. Biol. Med., vol. 171, p. 108112, 2024, doi: 10.1016/j.compbiomed.2024.108112.
  8. G. Wang and X. Hu, “Low-dose CT denoising using a progressive Wasserstein generative adversarial network,” Comput. Biol. Med., vol. 135, p. 104625, 2021, doi: 10.1016/j.compbiomed.2021.104625.
  9. B. Su, P. Dong, X. Hu, B. Wang, Y. Zha, Z. Wu, and J. Wan, “Fast and detail-preserving low-dose CT denoising with diffusion model,” Biomed. Signal Process. Control, vol. 105, p. 107580, 2025, doi: 10.1016/j.bspc.2025.107580.
  10. Y. Zhang, R. Zhang, R. Cao, F. Xu, F. Jiang, J. Meng, et al., “Unsupervised low-dose CT denoising using bidirectional contrastive network,” Comput. Methods Programs Biomed., vol. 251, p. 108206, 2024, doi: 10.1016/j.cmpb.2024.108206.
  11. N. T. Trung, D. H. Trinh, N. L. Trung, and M. Luong, “Low-dose CT image denoising using deep convolutional neural networks with extended receptive fields,” Signal, Image and Video Processing, vol. 16, pp. 1963–1971, 2022, doi: 10.1007/s11760-022-02157-8.
  12. N. T. Trung, D. H. Trinh, N. L. Trung, et al., “Low-dose CT image denoising using deep convolutional neural networks with extended receptive fields,” Signal Image Video Process., vol. 16, pp. 1963–1971, 2022, doi: 10.1007/s11760-022-02157-8.
  13. D. C. Lepcha, B. Goyal, A. Dogra, K. Vaghela, A. Singh, K. R. Kumar, and D. P. Bavirisetti, “Low-dose computed tomography image denoising using pixel level non-local self-similarity prior with non-local means for healthcare informatics,” Sci. Rep., vol. 15, p. 25095, 2025, doi: 10.1038/s41598-025-10139-2.
  14. W. Kim, J. Lee, M. Kang, J. S. Kim, and J. H. Choi, “Wavelet subband-specific learning for low-dose computed tomography denoising,” PLOS ONE, vol. 17, no. 9, p. e0274308, 2022, doi: 10.1371/journal.pone.0274308.
  15. L. Marcos, J. Alirezaie, and P. Babyn, “Low dose CT denoising by ResNet with fused attention modules and integrated loss functions,” Frontiers in Signal Processing, vol. 1, p. 812193, 2022, doi: 10.3389/frsip.2021.812193.
  16. W. Zhang, A. Salmi, C. Yang, and F. Jiang, “Innovative noise extraction and denoising in low-dose CT using a supervised deep learning framework,” Electronics, vol. 13, no. 16, p. 3184, 2024, doi: 10.3390/electronics13163184.
  17. X. Zhan, “Low-dose CT denoising network combined with attention mechanism,” in Proc. Int. Conf. Image Process. Artif. Intell. (ICIPAI), vol. 13213, pp. 924–928, Jul. 2024, doi: 10.1117/12.3035371.
  18. J. Shi, O. Elkilany, A. Fischer, A. Suppes, D. M. Pelt, and K. J. Batenburg, “LoDoInd: A benchmark low-dose industrial CT dataset – 1 of 3,” Zenodo, 2024. [Online]. Available: https://zenodo.org/records/10356955
  19. E. Eulig, B. Ommer, and M. Kachelrieß, “Benchmarking deep learning-based low-dose CT image denoising algorithms,” Med. Phys., vol. 51, no. 12, pp. 8776–8788, 2024, doi: 10.1002/mp.17379.
  20. W. Kim, S. Y. Jeon, G. Byun, H. Yoo, and J. H. Choi, “A systematic review of deep learning-based denoising for low-dose computed tomography from a perceptual quality perspective,” Biomed. Eng. Lett., vol. 14, no. 6, pp. 1153–1173, 2024.
  21. R. Furuta, N. Inoue, and T. Yamasaki, “Fully convolutional network with multi-step reinforcement learning for image processing,” in Proceedings of the AAAI Conference on Artificial Intelligence, vol. 33, no. 1, pp. 3598–3605, Jul. 2019, doi: 10.1609/aaai.v33i01.33013598.
  22. R. Zhang, J. Zhu, Z. Zha, J. Dauwels, and B. Wen, “R3L: Connecting deep reinforcement learning to recurrent neural networks for image denoising via residual recovery,” in Proc. IEEE Int. Conf. Image Process. (ICIP), Anchorage, AK, USA, 2021, pp. 1624–1628, doi: 10.1109/ICIP42928.2021.9506323.
  23. Y. Guo, Y. Gao, B. Hu, X. Qian, and D. Liang, “CMID: Crossmodal image denoising via pixel-wise deep reinforcement learning,” Sensors, vol. 24, no. 1, p. 42, 2023, doi: 10.3390/s24010042.