OpenLB
Open Library of Bioscience
Advanced Search
Radiological physics and technology
Jun, 2024;17 (2): 367-374.

Subjective and objective image quality of low-dose CT images processed using a self-supervised denoising algorithm.

Yuya Kimura, Takeru Q Suyama, Yasuteru Shimamura, Jun Suzuki, Masato Watanabe, Hiroshi Igei, Yuya Otera, Takayuki Kaneko, Maho Suzukawa, Hirotoshi Matsui, Hiroyuki Kudo
Author Information
  1. Yuya Kimura: Clinical Research Center, National Hospital Organization Tokyo National Hospital, Tokyo, Japan. yuk.close.to.wrd.34@gmail.com. ORCID
  2. Takeru Q Suyama: Nadogaya Research Institute, Nadogaya Hospital, Chiba, Japan.
  3. Yasuteru Shimamura: Department of Diagnostic Radiology, Kasumi Clinic, Hiroshima, Japan.
  4. Jun Suzuki: Department of Respiratory Medicine, National Hospital Organization Tokyo Hospital, Tokyo, Japan.
  5. Masato Watanabe: Department of Respiratory Medicine, National Hospital Organization Tokyo Hospital, Tokyo, Japan.
  6. Hiroshi Igei: Department of Respiratory Medicine, National Hospital Organization Tokyo Hospital, Tokyo, Japan.
  7. Yuya Otera: Department of Radiology, National Hospital Organization Tokyo Hospital, Tokyo, Japan.
  8. Takayuki Kaneko: Radiological Physics and Technology Department, National Center for Global Health and Medicine, Tokyo, Japan.
  9. Maho Suzukawa: Clinical Research Center, National Hospital Organization Tokyo National Hospital, Tokyo, Japan.
  10. Hirotoshi Matsui: Department of Respiratory Medicine, National Hospital Organization Tokyo Hospital, Tokyo, Japan.
  11. Hiroyuki Kudo: Institute of Systems and Information Engineering, University of Tsukuba, Ibaraki, Japan.
PMID: 38413510 DOI: 10.1007/s12194-024-00786-x

Abstract

This study aimed to assess the subjective and objective image quality of low-dose computed tomography (CT) images processed using a self-supervised denoising algorithm with deep learning. We trained the self-supervised denoising model using low-dose CT images of 40 patients and applied this model to CT images of another 30 patients. Image quality, in terms of noise and edge sharpness, was rated on a 5-point scale by two radiologists. The coefficient of variation, contrast-to-noise ratio (CNR), and signal-to-noise ratio (SNR) were calculated. The values for the self-supervised denoising model were compared with those for the original low-dose CT images and CT images processed using other conventional denoising algorithms (non-local means, block-matching and 3D filtering, and total variation minimization-based algorithms). The mean (standard deviation) scores of local and overall noise levels for the self-supervised denoising algorithm were 3.90 (0.40) and 3.93 (0.51), respectively, outperforming the original image and other algorithms. Similarly, the mean scores of local and overall edge sharpness for the self-supervised denoising algorithm were 3.90 (0.40) and 3.75 (0.47), respectively, surpassing the scores of the original image and other algorithms. The CNR and SNR for the self-supervised denoising algorithm were higher than those for the original images but slightly lower than those for the other algorithms. Our findings indicate the potential clinical applicability of the self-supervised denoising algorithm for low-dose CT images in clinical settings.

Keywords

Artificial intelligence Deep learning Low-dose X-ray computed tomography Self-supervised denoising algorithm

References

  1. Brenner D, Elliston C, Hall E, Berdon W. Estimated risks of radiation-induced fatal cancer from pediatric CT. AJR Am J Roentgenol. 2001;176:289–96. [DOI: 10.2214/ajr.176.2.1760289]
  2. Berrington de González A, Darby S. Risk of cancer from diagnostic X-rays: estimates for the UK and 14 other countries. Lancet. 2004;363:345–51. [DOI: 10.1016/S0140-6736(04)15433-0]
  3. Brenner DJ, Hall EJ. Computed tomography–an increasing source of radiation exposure. N Engl J Med. 2007;357:2277–84. [DOI: 10.1056/NEJMra072149]
  4. Lauer MS. Elements of danger–the case of medical imaging. N Engl J Med. 2009;361:841–3. [DOI: 10.1056/NEJMp0904735]
  5. Smith-Bindman R. Is computed tomography safe? N Engl J Med. 2010;363:1–4. [DOI: 10.1056/NEJMp1002530]
  6. Pearce MS, Salotti JA, Little MP, et al. Radiation exposure from CT scans in childhood and subsequent risk of leukaemia and brain tumours: a retrospective cohort study. Lancet. 2012;380:499–505. [DOI: 10.1016/S0140-6736(12)60815-0]
  7. Mathews JD, Forsythe AV, Brady Z, et al. Cancer risk in 680,000 people exposed to computed tomography scans in childhood or adolescence: data linkage study of 11 million Australians. BMJ. 2013;346: f2360. [DOI: 10.1136/bmj.f2360]
  8. LeCun Y, Bengio Y, Hinton G. Deep learning. Nature. 2015;521:436–44. [DOI: 10.1038/nature14539]
  9. Lecun Y, Bottou L, Bengio Y, Haffner P. Gradient-based learning applied to document recognition. Proc IEEE. 1998;86:2278–324. [DOI: 10.1109/5.726791]
  10. Chen H, Zhang Y, Kalra MK, et al. Low-dose CT with a residual encoder-decoder convolutional neural network. IEEE Trans Med Imaging. 2017;36:2524–35. [DOI: 10.1109/TMI.2017.2715284]
  11. Shan H, Padole A, Homayounieh F, et al. Competitive performance of a modularized deep neural network compared to commercial algorithms for low-dose CT image reconstruction. Nat Mach Intell. 2019;1:269–76. [DOI: 10.1038/s42256-019-0057-9]
  12. You C, Yang Q, Shan H, et al. Structurally-sensitive multi-scale deep neural network for low-dose CT denoising. IEEE Access. 2018;6:41839–55. [DOI: 10.1109/ACCESS.2018.2858196]
  13. Yang Q, Yan P, Zhang Y, et al. Low-dose CT image denoising using a generative adversarial network with wasserstein distance and perceptual loss. IEEE Trans Med Imaging. 2018;37:1348–57. [DOI: 10.1109/TMI.2018.2827462]
  14. Tang C, Li J, Wang L, et al. Unpaired low-dose CT denoising network based on cycle-consistent generative adversarial network with prior image information. Comput Math Methods Med. 2019;2019:8639825. [DOI: 10.1155/2019/8639825]
  15. Hendriksen AA, Pelt DM, Batenburg KJ. Noise2Inverse: self-supervised deep convolutional denoising for tomography. IEEE Trans Comput Imaging. 2020;6:1320–35. [DOI: 10.1109/TCI.2020.3019647]
  16. Wu D, Kim K, Li Q. Low-dose CT reconstruction with Noise2Noise network and testing-time fine-tuning. Med Phys. 2021;48:7657–72. [DOI: 10.1002/mp.15101]
  17. Krull A, Buchholz TO, Jug F. Noise2Void—Learning denoising from single noisy images. Proc IEEE Comput Soc Conf Comput Vis Pattern Recognit. 2019;2019:2124–32.
  18. Laine S, Karras T, Lehtinen J, Aila T. High-quality self-supervised deep image denoising. Comput Vis Pattern Recognit.
  19. Sidky EY, Pan X. Image reconstruction in circular cone-beam computed tomography by constrained, total-variation minimization. Phys Med Biol. 2008;53:4777–807. [DOI: 10.1088/0031-9155/53/17/021]
  20. Buades A, Coll B, Morel J-M. A non-local algorithm for image denoising. Proc IEEE Comput Soc Conf Comput Vis Pattern Recognit. 2005;2:60–5.
  21. Dabov K, Foi A, Katkovnik V, Egiazarian K. Image denoising by sparse 3-D transform-domain collaborative filtering. IEEE Trans Image Process. 2007;16:2080–95. [DOI: 10.1109/TIP.2007.901238]
  22. Svanholm H, Starklint H, Gundersen HJ, Fabricius J, Barlebo H, Olsen S. Reproducibility of histomorphologic diagnoses with special reference to the kappa statistic. APMIS. 1989;97:689–98. [DOI: 10.1111/j.1699-0463.1989.tb00464.x]
  23. Heyer CM, Mohr PS, Lemburg SP, Peters SA, Nicolas V. Image quality and radiation exposure at pulmonary CT angiography with 100- or 120-kVp protocol: prospective randomized study. Radiology. 2007;245:577–83. [DOI: 10.1148/radiol.2452061919]
  24. Dunn OJ. Multiple comparisons among means. J Am Stat Assoc. 1961;56:52–64. [DOI: 10.1080/01621459.1961.10482090]
  25. Ledig C, Theis L, Huszar F, et al. Photo-realistic single image super-resolution using a generative adversarial network. Proc IEEE Comput Soc Conf Comput Vis Pattern Recognit. 2017. https://doi.org/10.1109/CVPR.2017.19 .
  26. Johnson J, Alahi A, Fei-Fei L. Perceptual losses for real-time style transfer and super-resolution. Comput Vis ECCV 2016. https://doi.org/10.1007/978-3-319-46475-6_43 .
  27. Slavcheva M, Baust M, Cremers D, Ilic S. KillingFusion: Non-rigid 3D reconstruction without correspondences. In: 2017 Proc IEEE Comput Soc Conf Comput Vis Pattern Recognit. 2017. https://doi.org/10.1109/ICCV.2017.581 .
  28. Sajjadi MSM, Schölkopf B, Hirsch M. EnhanceNet: Single image super-resolution through automated texture synthesis. 2017 Proc IEEE Int Conf Comput Vis. 2017. https://doi.org/10.1109/ICCV.2017.481 .
  29. Yeh RA, Chen C, Lim TY, Schwing AG, Hasegawa-Johnson M, Do MA. Semantic image inpainting with deep generative models. 2017 Proc IEEE Comput Soc Conf Comput Vis Pattern Recognit. 2017. https://doi.org/10.1109/CVPR.2017.728 .
  30. Yang CY, Ma C, Yang MH. Single-image super-resolution: A benchmark. Comput Vis ECCV 2014. https://doi.org/10.1007/978-3-319-10593-2_25 .
  31. Shaham TR, Dekel T, Michaeli T. SinGAN: Learning a generative model from a single natural image. 2019 IEEE/CVF Int Conf Comput Vis (ICCV).
  32. Radpour A, Bahrami-Motlagh H, Taaghi MT, et al. COVID-19 evaluation by low-dose high resolution CT scans protocol. Acad Radiol. 2020;27:901. [DOI: 10.1016/j.acra.2020.04.016]
  33. Agostini A, Floridi C, Borgheresi A, et al. Proposal of a low-dose, long-pitch, dual-source chest CT protocol on third-generation dual-source CT using a tin filter for spectral shaping at 100 kVp for CoronaVirus Disease 2019 (COVID-19) patients: a feasibility study. Radiol Med. 2020;125:365–73. [DOI: 10.1007/s11547-020-01179-x]
  34. Kang Z, Li X, Zhou S. Recommendation of low-dose CT in the detection and management of COVID-2019. Eur Radiol. 2020;30:4356–7. [DOI: 10.1007/s00330-020-06809-6]
  35. Dangis A, Gieraerts C, De Bruecker Y, et al. Accuracy and reproducibility of low-dose submillisievert chest CT for the diagnosis of COVID-19. Radiol Cardiothorac Imaging. 2020;2: e200196. [DOI: 10.1148/ryct.2020200196]
  36. Li M, Hsu W, Xie X, Cong J, Gao W. SACNN: self-attention Convolutional Neural Network for low-dose CT denoising with self-supervised perceptual loss network. IEEE Trans Med Imaging. 2020;39:2289–301. [DOI: 10.1109/TMI.2020.2968472]

MeSH Term

Tomography, X-Ray Computed
Humans
Algorithms
Signal-To-Noise Ratio
Radiation Dosage
Image Processing, Computer-Assisted
Female
Male
Middle Aged
Aged
Adult
Journal Article

Word Cloud

Created with Highcharts 10.0.0denoisingimagesself-supervisedCTalgorithmlow-dosealgorithmsimageusingoriginal30qualityprocessedmodel40scoresobjectivecomputedtomographylearningpatientsnoiseedgesharpnessvariationratioCNRSNRmeanlocaloverall90respectivelyclinicalstudyaimedassesssubjectivedeeptrainedappliedanother30Imagetermsrated5-pointscaletworadiologistscoefficientcontrast-to-noisesignal-to-noisecalculatedvaluescomparedconventionalnon-localmeansblock-matching3Dfilteringtotalminimization-basedstandarddeviationlevels9351outperformingSimilarly7547surpassinghigherslightlylowerfindingsindicatepotentialapplicabilitysettingsSubjectiveArtificialintelligenceDeepLow-doseX-raySelf-supervised

Similar Articles

Cited By